Hello! If you are using an ad blocker but find something useful here and want to support me please consider disabling your ad blocker for this site.

Thank you,
Serafeim

A simple OpenID connect tutorial

Wed 29 November 2023

Introduction

OpenID Connect is a simple identity layer on top of the OAuth 2.0 protocol. It allows clients to verify the identity of the user based on the authentication performed by an Authorization Server, as well as to obtain basic profile information about the user.

To learn more about the OpenID Connect protocol, you can read how it works or, even better if you want to get a deeper understanding, you can read the specification.

The main difference between OpenID Connect and OAuth 2.0 is that OpenID Connect is an authentication protocol, while OAuth 2.0 is an authorization protocol; this means that OpenID Connect is used to verify the identity of the user, while OAuth 2.0 is used to verify if the user has access to some resources (and if so also very his identity).

In the following tutorial we’ll try to authenticate a user using OpenID Connect without any external libraries so we can understand how the authentication works. We’ll mainly work from the client side (i.e the web application) but we’ll also be able to understand how the authentication server should work.

We’ll use python and the requests library to make the HTTP requests however you can use any language you want, even command line with curl. To decode the JWT tokens we’ll use the python cryptography library.

Authentication flow

Some terminology

user: The user that tries to log in to the client application
client / client application: The application that the user tries to log in to; this is be a server side web application (however acts as a client for the OpenID Connect protocol)
server / authorization server: The server that authenticates the user and returns an id token and optionally an access token. This is the server that implements the OpenID Connect protocol.

The authentication flow for OpenID connect is more or less the following:

The user tries to “log in” on the client application
The client application generates a URL that points to the authorization endpoint of the authorization server adding some parameters to that url query and redirects the user’s browser there
The user will get a log in screen and try to log in to the authorization server (or if he’s already logged in he doesn’t need to do anything)
The authorization server redirects the user user back to the client application using a pre-agreed redirect url and passing it an authorization code
The client application retrieves the authorization code through the redirect url (callback) and sends a request to the token endpoint of the authorization server passing the authorization code and the client secret
The authorization server responds with an id token and optionally an access token
The client application can now use the id token to read the user information and optionally the access token to access more resources (if available)

This is also explained in the section 1.3. Overview of the OpenID Connect specification.

Authentication server

For the Authentication server for our tutorial we’ll use Keycloak, an open source Identity and Access Management server that implements OpenID Connect. I won’t go into details on how to install and configure Keycloak, the thing is that you need to setup a new client for your realm, enable authentication and add a client secret that will be used later.

Let’s suppose that you have created a realm with the name sample-realm and your keycloak server is hosted on https://keycloak.example.com. This realm will have a base url with the value: https://kc.example.gr/realms/sample-realm/. This base url will be used to build other urls and will be stored as OIDC_BASE_PROVIDER_URL.

If you are using a different OpenID Connect server, you’ll need to have the client id and client secret for your server and the token and authorization endpoints of your server.

Automatic Discovery

Some OpenID Connect servers support automatic discovery of various information from authorization server (including the needed endpoints). This is called WebFinger and is described in detail in the discovery specification.

To use the automatic discovery you can send a GET request to the /.well-known/openid-configuration endpoint of the authorization server. In our keycloak case, the endpoint will be https://kc.example.gr/realms/sample-realm/.well-known/openid-configuration. The response is a JSON object that contains the endpoints of the authorization server. For example, using python requests:

finger = requests.get(
    settings.OIDC_BASE_PROVIDER_URL + ".well-known/openid-configuration"
).json()

This may return a lot of info but you are basically interested in the following:

authorization_endpoint = finger["authorization_endpoint"]
token_endpoint = finger["token_endpoint"]

The Authentication url

The first step of the authentication flow is to generate an authentication url and redirect the user to that. This means that when the user tries to log in to your app (by clicking a button or visiting some url etc) your web application will redirect the user to the authentication url.

The authentication url is the authorization endpoint of the authorization server with some parameters. The parameters are (see 3.1.2.1. Authentication Request of the OpenID Connect specification) for more info if you want:

client_id: The client id of your application. This will be used by the authorization server to identify your application
response_type: The response type, in our case it’s code. There are various response types that can be used however code is the one that should be used to initiate the Authorization Code Flow which is used for server-side applications (there are two more flows described in the sections 3.2 and 3.3 but they aren’t used for traditional web applications).
scope: The scope of the request, in our case it has to contain openid (it can also have more scopes but openid is the minimum)
redirect_uri: The redirect uri of your application, in our case it’s http://localhost:8000/auth/callback. This is the url that the authorization server will redirect the user after the authentication is complete. Please notice that the authentication server will only redirect the user to this url if it’s in the list of allowed redirect uris of your application so make sure that you have added it to the list of allowed redirect uris.
state: A random string that will be used to verify the response from the authorization server

So, to create the url from python we’ll do something like:

from secrets import choice

def rndstr(size=16):
    # Pick a random string of size `size` from the alphabet
    alphabet = string.ascii_letters + string.digits
    return "".join([choice(alphabet) for _ in range(size)])

state = rndstr()

authorization_url = f"{authorization_endpoint}?client_id={settings.OIDC_CLIENT_ID}&response_type=code&scope=openid&redirect_uri={settings.OIDC_REDIRECT_URIS[0]}&state={state}"

and we’ll get something like:

https://kc.example.gr/realms/sample-realm/protocol/openid-connect/auth?client_id=sample-client&response_type=code&scope=openid&redirect_uri=http://localhost:8000/auth/callback&state=skldfj98sdfjio12

User Authentication and redirect

Our application will now redirect the user to the authentication url. To test for this tutorial, we’ll just copy paste the authentication url on our browser; the user will need to log in to the authorization server and then he’ll be redirected back to our redirect_uri we provided with an authorization code.

If we don’t have anything running on localhost:8000 we’ll get an error but we’ll be able to see the redict URL on our browser’s bar(!). So, we’ll see something like http://localhost:8000/auth/callback?state=KpT23RpwimxzXzHa&session_state=d54e38e1-8dfa-4658-ab4e-817ef6d4029b&code=b4938b28-12b4-463b-b2bc-ffc91b29e79e.d54e38e1-8dfa-4658-ab4e-817ef6d4029b.f9302f16-4765-4f2a-b7bc-381208ec71d6. There are three parameters that are returned:

code: The authorization code. This is the code that we’ll use to get the id token.
state: The state we provided earlier. We need to check that this is the same as the one we provided earlier to make sure that the response is valid.
session_state: The session state. This is optional and can be used to check if the user is still logged in to the authorization server. We won’t use this in our tutorial but if you want to learn more you can read the session specification.

query = urlparse(next_url).query
query_dict = {k: v[0] for k, v in parse_qs(query).items()}

Now query_dict will contain all the parameters of the url as a dictionary. So we can check that we got the correct state:

assert query_dict["state"] == state

Getting the token

The next step is for our web application to retrieve the id token. For that it’ll need to send a POST request to the token endpoint of the authorization server passing it the following parameters (see 3.1.3.1. Token Request of the OpenID Connect specification):

client_id: The client id of your application
client_secret: The client secret of your application
grant_type: The grant type, in our case it must be authorization_code .
code: The authorization code we got from the authorization server
redirect_uri: The redirect uri we used before

Here’s the request we need to do. Please notice that this request will be done by your server-side application and not by the user’s browser (so the client secret will be safe).

resp = requests.post(
    token_endpoint,
    data={
        "client_id": settings.OIDC_CLIENT_ID,
        "client_secret": settings.OIDC_CLIENT_SECRET,
        "grant_type": "authorization_code",
        "code": query_dict["code"],
        "redirect_uri": settings.OIDC_REDIRECT_URIS[0],
    },
)

The above response will return a bunch of parameters depending on the authentication server. The most important of these parameters are:

id_token: The id token that contains information about the user. This is a JWT token and needs to be decoded and verified. We’ll see how to do that later.
access_token: The access token that can be used to access the user information.
token_type: This has the Bearer value

Decoding a JWT (JSON Web Token)

Although there are libraries that can be used for decoding and verifying the token (pyjwt), we’ll do it using only python (and the cryptography library) to understand how it works.

The JWT token is a string that contains three parts separated by a dot (.). The first part is the header, the second part is the payload and the third part is the signature. Each part is base64 encoded. To decode the token we can use the following function:

def decode_jwt(jwt_token):
    header, payload, signature = jwt_token.split(".")

    decoded_header = base64.urlsafe_b64decode(header + "=" * (-len(header) % 4))
    decoded_payload = base64.urlsafe_b64decode(payload + "=" * (-len(payload) % 4))

    return json.loads(decoded_header), json.loads(decoded_payload)

The signature will be used later to verify the token.

The decoded header will be similar to this

{'alg': 'RS256',
'kid': 'NjVBRjY5MDlCMUIwNzU4RTA2QzZFMDQ4QzQ2MDAyQjVDNjk1RTM2Qg',
'typ': 'JWT'}

defining the type as JWT, the algorithm used to sing the JWT (RS256) and the public key that was used for the signing (we’ll see later how we can retrieve that public key).

The decoded payload may have various fields depending on the authentication server. The most important fields are (see 2. ID Token of the OpenID Connect specification):

iss: The issuer of the token. This must be the same as the OIDC_BASE_PROVIDER_URL we defined earlier.
sub: The subject of the token. This is more or less a unique id for that particular user that identified. Please notice that this usually is not the username of the user but some internal and unique key of that user.
aud: The audience of the token. This must contain the OIDC_CLIENT_ID we defined earlier (i.e sample-client).
exp: Expiration time of the JWT (in seconds since epoch)
iat: The time the JWT was issued (in seconds since epoch)
auth_time: The time the user was authenticated (in seconds since epoch)

For example the decoded payload may be something like:

{
'aud': 'sample-client',
'auth_time': 1701168800,
'exp': 1701169100,
'iat': 1701168800,
'iss': 'https://kc.example.gr/realms/sample-realm',
'email': '[email protected]',
'preferred_username': 'sample',
'name': 'Sample',
'sub': 'e22f5a0d-e5ac-472d-b41b-06ecb9e4b3f6',
'typ': 'ID'
}

As you see the payload may contain some extra information about the user (like the email, name etc). This information is not guaranteed to be there and it depends on the authentication server. If the information you want is there you can verify the JWT and finish the authentication flow here.

If not you can call the userinfo endpoint with the help of the access_token to get more information about the user.

Retrieving the public key

The JWT token is signed by the authentication server using a private key. To verify the token we need to get the public key of the authentication server and use it to verify the signature of the token.

There are multiple ways to retrieve the public key. The simplest is to get it directly from the authentication server (i.e keycloak has an option to export the RS256 public key as text from the realm settings - keys of your realm). Also, if you visit the base url we defined later you’ll get a JSON with the public key of the server i.e

base_info = requests.get(settings.OIDC_BASE_PROVIDER_URL).json()
public_key = base_info["public_key"]

I’m not sure if this is supported by other server beyond keycloak though.

Finally you can use the jwks_uri of the authentication server to get the public key. The jwks_uri would be returned from the WebFinger response we described earlier. There you’ll see a JSON object named keys that contains JSON array with the list of the keys that the server has.

Each key will have the following fields:

kid: The key id; this can be used to correlate the key with the one in the header of the JWT token
kty: The key type; this should be RSA
alg: The algorithm used to sign the JWT token; this should be RS256
use: The use of the key; this should be sig (for signing)
n: The modulus of the RSA key
e: The exponent of the RSA key
x5c: The x509 certificate chain of the key
x5t: The x509 certificate SHA-1 thumbprint of the key
x5t#S256: The x509 certificate SHA-256 thumbprint of the key

Although the public key isn’t there it is rather straightforward to produce it either from the n and e or by reading the x509 certificate through the x5c field.

Since we have now entered the cryptography fields we need to use the python cryptography library.

If we have the public key as a string we can use the following function to get the public key:

from cryptography.hazmat.backends import default_backend
from cryptography.hazmat.primitives import serialization

public_key_str = "-----BEGIN PUBLIC KEY-----\n{0}\n-----END PUBLIC KEY-----".format(
    settings.OIDC_PUBLIC_KEY_STR
)
public_key = default_backend().load_pem_public_key(public_key_str.encode("utf-8"))
public_key_pem = public_key.public_bytes(
    encoding=serialization.Encoding.PEM,
    format=serialization.PublicFormat.SubjectPublicKeyInfo
)
print(public_key_pem.decode())

If we want to produce the public key using the n and e fields we can use the following function:

import base64
from cryptography.hazmat.primitives.asymmetric import rsa
from cryptography.hazmat.backends import default_backend
from cryptography.hazmat.primitives import serialization

n_bytes = base64.urlsafe_b64decode(key['n'] + "===")
n = int.from_bytes(n_bytes, 'big')

e_bytes = base64.urlsafe_b64decode(key['e'] + "===")
e = int.from_bytes(e_bytes, 'big')

public_key = rsa.RSAPublicNumbers(e, n).public_key(default_backend())

public_key_pem = public_key.public_bytes(
    encoding=serialization.Encoding.PEM,
    format=serialization.PublicFormat.SubjectPublicKeyInfo
)
print(public_key_pem.decode())

Or, to get the public key from the x509 certificate:

from cryptography import x509
from cryptography.hazmat.backends import default_backend
from cryptography.hazmat.primitives import serialization

certificate_str = key["x5c"][0]
decoded_certificate_str = base64.b64decode(certificate_str)
certificate = x509.load_der_x509_certificate(decoded_certificate_str, default_backend())
pub_key = certificate.public_key()
public_key_pem = pub_key.public_bytes(
    encoding=serialization.Encoding.PEM,
    format=serialization.PublicFormat.SubjectPublicKeyInfo
)
print(public_key_pem.decode())

In all three above cases we should see the same public key.

Verifying the JWT token

Finally after retrieving the public key (using any of the above methods) we can verify the JWT token.

from cryptography.hazmat.primitives import hashes
from cryptography.hazmat.primitives.asymmetric import padding

def verify_rs256_signature(jwt_token, pub_key):
    header, payload, signature = jwt_token.split(".")

    data = (header + "." + payload).encode("utf-8")

    decoded_signature = base64.urlsafe_b64decode(
        signature + "=" * (-len(signature) % 4)
    )

    try:
        pub_key.verify(
            signature, data, padding.PKCS1v15(), hashes.SHA256()
        )
        print("Signature verified")
    except Exception as e:
        print("Signature verification failed!", e)

And call it like:

verify_rs256_signature(id_token, pub_key)

Where the id_token is the JWT token we got earlier from the token endpoint (i.e id_token = resp.json()["id_token"]) and the pub_key is the public key object.

Notice that beyond the signature verification we also need to verify that the token has valid values the iss and aud fields and also that it has valid times (i.e not expired, not issued in the future etc). We can do that using the exp, iat and auth_time fields of the payload. These times are in UTC and in seconds since epoch. So we can do something like:

import datetime

exp_datetime = datetime.datetime.utcfromtimestamp(decoded_payload['exp'])

to convert them to datetime objects and then compare them with the current time (in UTC).

After the id token is verified we can safely use the information contained in the payload to authenticate the user. If this information is enough we can finish the authentication flow here. If not we can use the access token to call the userinfo endpoint.

Calling the userinfo endpoint

Calling the userinfo endpoint is rather straightforward. We just need to send a GET request to the userinfo endpoint of the authentication server passing it the access token as a bearer token. The userinfo endpoint is returned from the WebFinger response we described earlier.

userinfo_endpoint = finger["userinfo_endpoint"]
resp = requests.get(
    userinfo_endpoint,
    headers={"Authorization": f"Bearer {access_token}"},
)

The resp may be also a JWT token so you’ll need to decode and verify it as we did before. If it’s not a JWT token it’ll be a JSON object with the user information. Notice that it may not have more information than the id token so could avoid calling the userinfo endpoint. However this depends on the implementation so you probably should check it yourself.

What about the authorization server?

What we would need to do if we wanted to implement the authorization server? It isn’t so complicated:

We’ll need to store the users with their passwords and the clients with their secrets and redirect urls. Then we’d need to implement the following urls:

/login: A user facing login page. The client application will redirect the user’s browser to that page. After the user connects the server would redirect the user’s browser to the redirect url of the client passing it an authorization code.
/token: The client would issue a post request to that url passing the authorization code and the client secret. The server now would return the id token encoded as a jwt token. The id token would contain the user information.

These two are more or less enough for a simple authentication server. Of course there are more things that can be implemented like the userinfo endpoint, the fingering etc but these are optional.

Conclusion

In this tutorial we saw how we can authenticate a user using OpenID Connect without any external libraries. We saw how we can generate the authentication url, how we can get the authorization code and how we can use the authorization code to get the id token and access token. We also saw how we can decode and verify the id token, how we can retrieve the public key of the authorization server and finally and how we can use the access token to call the userinfo endpoint.

Simple Django - DataTables integration

Tue 23 May 2023

In this small post I’ll show a simple and quick way to integrate the jquery DataTables library with Django.

The DataTables library has a lot of features however in this article we’ll only take advantage of the basic features that should be enough for most use cases:

Ajax loading of data
Ajax pagination
Ajax search/filtering

If you want to use more features you can try a django package like django-ajax-datatable that supports most of the DataTables features; however because of the DataTables complexity you’ll see that it will need a lot of work even for a simple integration.

The following model will be used for our example:

class Item(models.Model):
    code = models.CharField(max_length=16, unique=True, )
    name = models.CharField(max_length=512, )
    name_en = models.CharField(max_length=512, )

We’ll create a Django class based view that will return a template with an empty <table> element that we’ll then fill with Ajax. The same view will check if it receives datatables requests and return the correct data in the format datatables expects:

from django.db.models import Q
from django.http import JsonResponse
from django.views.generic import ListView

class ItemListView(ListView):
    model = models.Item
    template_name = "item_datatable.html"

    def render_to_response(self, context, **response_kwargs):
        if self.request.GET.get("datatables"):
            draw = int(self.request.GET.get("draw", "1"))
            length = int(self.request.GET.get("length", "10"))
            start = int(self.request.GET.get("start", "0"))
            sv = self.request.GET.get("search[value]", None)
            qs = self.get_queryset().order_by("code")
            if sv:
                qs = qs.filter(
                    Q(name__icontains=sv)
                    | Q(code__icontains=sv)
                    | Q(name_en__icontains=sv)
                )
            filtered_count = qs.count()
            qs = qs[start : start + length]

            return JsonResponse(
                {
                    "recordsTotal": self.get_queryset().count(),
                    "recordsFiltered": filtered_count,
                    "draw": draw,
                    "data": list(qs.values()),
                },
                safe=False,
            )
        return super().render_to_response(context, **response_kwargs)

The above is a simple ListView for our Item model. It overrides the render_to_response method to return the Ajax json data if the request is a datatables request. To do that, it first checks to see if there’s a datatables parameter in the request.GET. If this isn’t a datatables request it will return the normal template response.

However, if it is a datatables request it will pick the draw, length, start and search[value] parameters from the request.GET (with default values if they aren’t there) and use them to prepare the response. Notice that:

for the filter we need to do an OR (|) because of how the datatables default fildering works (one single filter field for all columns)
we can select any of the fields we want to filter with by adding them to the OR expression
we’ll choose the correct page using qs[start:start+length], the length will be changed if the user uses the page-size field of the datatables
we need to count the filtered results before taking the slice or else Django will throw an error
the draw parameter should be converted to integer and passed back to the response (it is used in case there are multiple pending datatable ajax requests).

Finally, we return a JsonResponse with the correct data. The safe=False is needed because we are returning a list of dictionaries and not a single dictionary. Notice the recordsTotal and recordsFiltered keys; these are needed by datatables to know how many records there are in total and how many records are returned after filtering.

The item_datatable.html template for this view is the following:

{% extends "site_base.html" %}
{% load static %}

{% block extra_style %}
    <link rel="stylesheet" type="text/css" href="{% static 'datatables.min.css' %}"/>
{% endblock %}

{% block page_content %}

<div class="row">
    <div class="col-md-12">
        <table id='table' class='table'></table>
    </div>
</div>

{% endblock %}

{% block extra_script %}
<script type="text/javascript" src="{% static 'jquery.min.js' %}"></script>
<script type="text/javascript" src="{% static 'datatables.min.js' %}"></script>

<script>
$(function() {

  $('#table').DataTable( {
        "ordering": false,
        ajax: '{{ request.path }}?datatables=1',
        serverSide: true,
        columns: [
            { data: 'code', title: 'Κωδικός' },
            { data: 'name', title: 'Περιγραφή' },
            { data: 'name_en', title: 'Περιγραφή (αγγλικά)' },
        ],
        language: {
          "sDecimal":           ",",
          "sEmptyTable":        "Δεν υπάρχουν δεδομένα στον πίνακα",
          "sInfo":              "Εμφανίζονται _START_ έως _END_ από _TOTAL_ εγγραφές",
          "sInfoEmpty":         "Εμφανίζονται 0 έως 0 από 0 εγγραφές",
          "sInfoFiltered":      "(φιλτραρισμένες από _MAX_ συνολικά εγγραφές)",
          "sInfoPostFix":       "",
          "sInfoThousands":     ".",
          "sLengthMenu":        "Δείξε _MENU_ εγγραφές",
          "sLoadingRecords":    "Φόρτωση...",
          "sProcessing":        "Επεξεργασία...",
          "sSearch":            "Αναζήτηση:",
          "sSearchPlaceholder": "Αναζήτηση",
          "sThousands":         ".",
          "sUrl":               "",
          "sZeroRecords":       "Δεν βρέθηκαν εγγραφές που να ταιριάζουν",
          "oPaginate": {
              "sFirst":    "Πρώτη",
              "sPrevious": "Προηγούμενη",
              "sNext":     "Επόμενη",
              "sLast":     "Τελευταία"
          },
          "oAria": {
              "sSortAscending":  ": ενεργοποιήστε για αύξουσα ταξινόμηση της στήλης",
              "sSortDescending": ": ενεργοποιήστε για φθίνουσα ταξινόμηση της στήλης"
          }
      }
    } );
})
</script>
{% endblock %}

(Please ignore the language setting this is needed to translate the datatables messages to greek.)

The important part is that we add the jquery and datatable dependencies (remeber that datatables also has a css) and then add an empty table (<table class='table'></table>). Finally, after the page is loaded the table is initialized as datatable using $('table.table').DataTable(options).

The options we pass to enable the ajax functionality are:

{
    ordering: false,
    ajax: '{{ request.path }}?datatables=1',
    serverSide: true,
    columns: [
        { data: 'code', title: 'Κωδικός' },
        { data: 'name', title: 'Περιγραφή' },
        { data: 'name_en', title: 'Περιγραφή (αγγλικά)' },
    ]
}

I didn’t need ordering so I haven’t implemented it here however it would be possible to implement it by picking the order-related parameters from the request similar to the filtering and using them as order_by parameters to the queryset, see the order[i][column] and order[i][dir] here.

For the response, we use the current request url passing it the datatables=1 parameter as discussed before. We define the datatable columns using the columns attr; the data key is the name of the field in the json data returned by the server and the title is the title of the column. These columns must exist in the json data returned by the server.

Finally, we need to add the view to our urls.py:

    path(
        "item_datatable/",
        ItemListListView.as_view(),
        name="item_datatable",
    ),

The above is enough to have a working datatable with ajax loading and filtering in your Django list views.

AI auto-subtitling

Mon 22 May 2023

Introduction

As English is not my native language, I rely on subtitles to fully enjoy and comprehend most movies. Unfortunately there are a lot of movies that don’t include subtitles or the subtitles that I am able to find are not synchronized with the movie.

In this small post I’ll give you some instructions on how to use the newest “AI” trends to automatically generate subtitles for a movie. The process does not achive 100% accuracy but it is very good and should definitely allow you to understand what’s being said.

Also, the described process will be very useful to people that do the actual subtitling for new movies since it should save them a lot of manual labour. Instead of adding subtitles and timings manually they can use this to generate an automatic subtitle draft and then edit it by hand and ear.

Finally, please notice that all tools described here are free and open source and you should be able to run everything on your PC even if it is very slow and doesn’t have a GPU.

Everything described here is for educational purposes only. Please don’t use it to generate subtitles for movies that you don’t own or have the rights to do so.

Whisper.cpp

For the auto-subtitling we’ll use the whisper.cpp library. This library can be used to auto-transcribe audio files i.e it gets audio as input and outputs the text that is being said in the audio.

This library can be compiled on your PC however to avoid complex workflows you can download the binaries for your system. I’d also recommend to download the BLAS binaries since they should work faster if your system supports them.

After you download the whisper.cpp extract it in a folder of your choice.

Beyond whisper.cpp, you need to download a model that will be used for the transcription. The simplest way is to go to the hugging face whisper.cpp model repository and download the model from there. You only need 1 model file. The largest models will give you better results but will be slower and require more resources. I think that the base or small models should be good and fast enough.

I’ll give you some test results later to see the differences.

Notice that if your movie is in english you should download the .en models.

To continue copy the downloaded models in the whisper.cpp folder.

Extract audio from the movie

The whisper.cpp library requires uncompressed audio (a .wav file) with specific characteristics (a sample rate of 16khz) to work.

To be able to extract that audio from our movie will use ffmpeg. Download the win64 binaries from here (get the ffmpeg-master-latest-win64-gpl.zip file) and copy over ffmpeg.exe to the whisper.cpp folder.

Then, to extract the audio from the movie you can use the following command:

ffmpeg.exe -i "movie.mp4" -f wav -vn -acodec pcm_s16le -ar 16000 -ac 1 "movie.wav"

(please change movie.mp4 and movie.wav with the correct names for your movie file).

To understand the command:

-i "movie.mp4": the input file
-f wav: the output format (a .wav file)
-vn: no video
-acodec pcm_s16le: the audio codec (the correct one for .wav file)
-ar 16000: the sample rate (16khz)
-ac 1: the number of channels (1 for mono, 2 for stereo)
"movie.wav": the output file

The process should be very fast and will give you a .wav file with the same length as the movie.

Transcribing the audio

The last step is to do the actual audio transcribing using whisper.cpp. The great thing about whisper.cpp is that it can directly create .srt (subtitle) files. The command to use is:

main.exe -osrt -m ggml-base.en.bin -f movie.wav

Notice that

The whisper.cpp bundle you downloaded should have a main.exe file
-osrt: Generate a .srt file as output
-m ggml-base.en.bin: The model to use for the transcription
-f movie.wav: The input file

The above command will start outputing the text it detects and when it finishes it will generate a movie.wav.srt. You can then use that .srt file to add subtitles for your movie!

Results

To test the process I used the first 14 minutes of the movie Kill Bill 2 as input. To generate the .wav file I used the following command:

ffmpeg.exe -i "kb2.mp4" -f wav -vn -acodec pcm_s16le -ar 16000 -ac 1  -ss 00:00:00 -to 00:14:00 "kb2.wav"

(please notice the -ss and -to to select the first 14 minutes of the movie).

This generated a 26 MB wav file. Then I tried the results using three different models:

ggml-tiny.en.bin with a size of 77 MB
ggml-base.en.bin with a size of 147 MB
ggml-small.en.bin with a size of 488 MB
ggml-medium.bin with a size of 1533 MB

ggml-tiny.en.bin

Some stats:

whisper_model_load: mem required  =  201.00 MB (+    3.00 MB per decoder)
whisper_model_load: model size    =   73.54 MB

whisper_print_timings:    total time = 166330.89 ms

So time needed was ~ 160 seconds

And the actual transcription:

[00:00:00.000 --> 00:00:03.000]   [MUSIC PLAYING]
[00:00:03.000 --> 00:00:16.480]   Do you finally sit just now?
[00:00:16.480 --> 00:00:19.120]   [MUSIC PLAYING]
[00:00:19.120 --> 00:00:21.480]   No, I can't do.
[00:00:21.480 --> 00:00:27.600]   I'd like to believe you're aware enough even now.
[00:00:27.600 --> 00:00:35.080]   No, that there is nothing suggesting in my actions.
[00:00:35.080 --> 00:00:44.000]   This moment, this is me and my most nice against them.
[00:00:44.000 --> 00:00:47.440]   Well, it's your name.
[00:00:47.440 --> 00:00:52.920]   [MUSIC PLAYING]
[00:00:52.920 --> 00:00:55.120]   But Dad didn't I?
[00:00:55.120 --> 00:00:56.640]   Well, I wasn't.
[00:00:56.640 --> 00:00:58.040]   But it wasn't from lack of trying.
[00:00:58.040 --> 00:01:00.360]   I can tell you that.
[00:01:00.360 --> 00:01:03.600]   Actually, Bill's last bullet put me in a coma.
[00:01:03.600 --> 00:01:07.600]   A coma, I was to lie in for four years.
[00:01:07.600 --> 00:01:09.360]   And I woke up.
[00:01:09.360 --> 00:01:11.280]   I went on with a movie advertisements
[00:01:11.280 --> 00:01:15.720]   for two as a roaring rampage of revenge.
[00:01:15.720 --> 00:01:17.120]   I roared.
[00:01:17.120 --> 00:01:18.800]   And I relanged.
[00:01:18.800 --> 00:01:22.480]   And I got bloody satisfaction.
[00:01:22.480 --> 00:01:26.360]   I've killed a hell of a lot of people to get to this point.
[00:01:26.360 --> 00:01:29.720]   But I have only one more.
[00:01:29.720 --> 00:01:31.760]   The last one.
[00:01:31.760 --> 00:01:35.360]   The one I'm driving to right now.
[00:01:35.360 --> 00:01:38.280]   The only one left.
[00:01:38.280 --> 00:01:42.040]   And when I arrive at my destination,
[00:01:42.040 --> 00:01:44.200]   I have going to kill Bill.
[00:01:44.200 --> 00:01:47.200]   [MUSIC PLAYING]
[00:01:47.200 --> 00:02:10.880]   [MUSIC PLAYING]
[00:02:10.880 --> 00:02:13.480]   Now the incident that happened at the two pines wedding
[00:02:13.480 --> 00:02:17.360]   chapel that put this whole gory story into motion
[00:02:17.360 --> 00:02:20.120]   has since become legend.
[00:02:20.120 --> 00:02:22.800]   Massacre at two pines.
[00:02:22.800 --> 00:02:24.800]   That's what the newspapers called it.
[00:02:24.800 --> 00:02:28.760]   The local TV news called it the El Paso, Texas wedding
[00:02:28.760 --> 00:02:30.920]   chapel massacre.
[00:02:30.920 --> 00:02:32.280]   How it happened?
[00:02:32.280 --> 00:02:33.680]   Who was there?
[00:02:33.680 --> 00:02:36.600]   How many got killed and who killed them?
[00:02:36.600 --> 00:02:40.320]   Changes depending on who's telling the story.
[00:02:40.320 --> 00:02:43.080]   In actual fact, the massacre didn't happen
[00:02:43.080 --> 00:02:45.720]   during a wedding at all.
[00:02:45.720 --> 00:02:48.360]   It was a wedding rehearsal.
[00:02:48.360 --> 00:02:51.800]   Now when we come to the park where I say you make kiss,
[00:02:51.800 --> 00:02:55.120]   the bride, you make kiss, the bride.
[00:02:55.120 --> 00:02:56.960]   But don't stick your tongue in her mouth.
[00:02:56.960 --> 00:03:01.840]   This might be funny to your friends,
[00:03:01.840 --> 00:03:06.440]   but it would be embarrassing to your parents.
[00:03:06.440 --> 00:03:08.200]   We'll try to do strange things.
[00:03:08.200 --> 00:03:11.080]   [LAUGHTER]
[00:03:11.080 --> 00:03:12.040]   Y'all got a song?
[00:03:12.040 --> 00:03:18.000]   How about love me, tender?
[00:03:18.000 --> 00:03:18.840]   I'd play that.
[00:03:18.840 --> 00:03:24.360]   Let me tender be great.
[00:03:24.360 --> 00:03:27.520]   Rufus, he's the man.
[00:03:27.520 --> 00:03:30.640]   Rufus, who was that he used to play for?
[00:03:30.640 --> 00:03:32.640]   Rufus Thomas.
[00:03:32.640 --> 00:03:34.760]   Rufus Thomas.
[00:03:34.760 --> 00:03:35.680]   Rufus Thomas.
[00:03:35.680 --> 00:03:37.080]   I was a dreel.
[00:03:37.080 --> 00:03:38.600]   I was a drifter.
[00:03:38.600 --> 00:03:40.080]   I was a colister.
[00:03:40.080 --> 00:03:41.920]   I was a part of the game.
[00:03:41.920 --> 00:03:43.760]   I was a bar kid.
[00:03:43.760 --> 00:03:48.200]   If they come through Texas, I can play with him.
[00:03:48.200 --> 00:03:50.720]   Rufus, he's the man.
[00:03:50.720 --> 00:03:56.840]   I never forgot anything.
[00:03:56.840 --> 00:04:00.440]   Oh, yes, you forgot the seating arrangements.
[00:04:00.440 --> 00:04:03.200]   Thank you, Mother.
[00:04:03.200 --> 00:04:07.840]   Now the way we normally do this, we have the bride's side,
[00:04:07.840 --> 00:04:09.560]   and then we have the groom's side.
[00:04:09.560 --> 00:04:12.880]   But since the bride ain't got nobody coming,
[00:04:12.880 --> 00:04:16.960]   and the groom's got far too many people coming--
[00:04:16.960 --> 00:04:18.920]   Well, yeah, they're coming all the way from Oklahoma.
[00:04:18.920 --> 00:04:21.600]   [LAUGHTER]
[00:04:21.600 --> 00:04:23.160]   Right.
[00:04:23.160 --> 00:04:27.880]   Well, I don't see no problem with the groom's side
[00:04:27.880 --> 00:04:29.840]   sharing the bride's side.
[00:04:29.840 --> 00:04:30.760]   Do you, Mother?
[00:04:30.760 --> 00:04:32.720]   Not a problem with that.
[00:04:32.720 --> 00:04:38.880]   But honey, you know, it would be good if you had somebody come.
[00:04:38.880 --> 00:04:42.840]   You know, it was a sign of good faith.
[00:04:42.840 --> 00:04:49.400]   Well, I don't have anybody except for Tommy and my friends.
[00:04:49.400 --> 00:04:52.000]   You have no family?
[00:04:52.000 --> 00:04:53.600]   Well, I'm working on changing that.
[00:04:53.600 --> 00:04:55.360]   Mrs. Harmony, we're all the family.
[00:04:55.360 --> 00:04:56.760]   This is Langel's ever going to need.
[00:04:56.760 --> 00:05:01.760]   I don't feel very well in this bitch.
[00:05:01.760 --> 00:05:03.680]   This started to piss me off.
[00:05:03.680 --> 00:05:08.280]   So while you all blather on, I'm going to go outside and get some air.
[00:05:08.280 --> 00:05:10.320]   I'm a Reverend, sorry.
[00:05:10.320 --> 00:05:11.840]   She's going to go out and get some air.
[00:05:11.840 --> 00:05:12.440]   Yeah.
[00:05:12.440 --> 00:05:15.320]   Given her delicate condition, she just needs a few minutes
[00:05:15.320 --> 00:05:15.920]   to get it together.
[00:05:15.920 --> 00:05:16.920]   She'll be OK.
[00:05:16.920 --> 00:05:19.880]   [MUSIC PLAYING]
[00:05:19.880 --> 00:05:23.880]   [MUSIC PLAYING]
[00:05:23.880 --> 00:05:27.880]   [MUSIC PLAYING]
[00:05:27.880 --> 00:05:31.880]   [MUSIC PLAYING]
[00:05:31.880 --> 00:05:35.880]   [MUSIC PLAYING]
[00:05:35.880 --> 00:05:39.880]   [MUSIC PLAYING]
[00:05:39.880 --> 00:05:43.880]   [MUSIC PLAYING]
[00:05:43.880 --> 00:05:47.880]   [MUSIC PLAYING]
[00:05:47.880 --> 00:05:51.880]   [MUSIC PLAYING]
[00:05:51.880 --> 00:05:55.880]   [MUSIC PLAYING]
[00:05:55.880 --> 00:05:59.880]   [MUSIC PLAYING]
[00:05:59.880 --> 00:06:01.880]   [MUSIC PLAYING]
[00:06:01.880 --> 00:06:03.880]   [MUSIC PLAYING]
[00:06:03.880 --> 00:06:05.880]   [MUSIC PLAYING]
[00:06:05.880 --> 00:06:07.880]   [MUSIC PLAYING]
[00:06:07.880 --> 00:06:09.880]   [MUSIC PLAYING]
[00:06:35.880 --> 00:06:37.880]   Hello, kiddo.
[00:06:37.880 --> 00:06:45.880]   How did you find me?
[00:06:45.880 --> 00:06:46.880]   I'm the man.
[00:06:46.880 --> 00:06:54.880]   What are you doing here?
[00:06:54.880 --> 00:06:57.880]   Am I doing?
[00:06:57.880 --> 00:07:02.880]   Well, I'm only going to play my flute.
[00:07:02.880 --> 00:07:05.880]   [MUSIC PLAYING]
[00:07:05.880 --> 00:07:14.880]   This moment, I'm looking at the most beautiful
[00:07:14.880 --> 00:07:18.880]   bright and these old eyes of every scene.
[00:07:18.880 --> 00:07:20.880]   Where are you here?
[00:07:20.880 --> 00:07:21.880]   Nice look.
[00:07:21.880 --> 00:07:26.880]   Are you going to be nice?
[00:07:26.880 --> 00:07:30.880]   I've never been nice my whole life.
[00:07:30.880 --> 00:07:33.880]   But I'll do my best to be sweet.
[00:07:33.880 --> 00:07:42.880]   I was told you, your sweet side is your best side.
[00:07:42.880 --> 00:07:45.880]   I guess that's why you're the only one who's ever seen it.
[00:07:45.880 --> 00:07:51.880]   See, you got a bun in the oven.
[00:07:51.880 --> 00:07:54.880]   Hmm.
[00:07:54.880 --> 00:07:57.880]   I'm knocked out.
[00:07:57.880 --> 00:07:59.880]   I'm not sure what you're doing.
[00:07:59.880 --> 00:08:01.880]   I'm not sure what you're doing.
[00:08:01.880 --> 00:08:03.880]   I'm not sure what you're doing.
[00:08:03.880 --> 00:08:05.880]   I'm not sure what you're doing.
[00:08:05.880 --> 00:08:07.880]   I'm not sure what you're doing.
[00:08:07.880 --> 00:08:09.880]   I'm not sure what you're doing.
[00:08:09.880 --> 00:08:11.880]   I'm not sure what you're doing.
[00:08:11.880 --> 00:08:13.880]   I'm not sure what you're doing.
[00:08:13.880 --> 00:08:15.880]   I'm not sure what you're doing.
[00:08:15.880 --> 00:08:17.880]   I'm not sure what you're doing.
[00:08:17.880 --> 00:08:19.880]   I'm not sure what you're doing.
[00:08:19.880 --> 00:08:21.880]   I'm not sure what you're doing.
[00:08:21.880 --> 00:08:23.880]   I'm not sure what you're doing.
[00:08:23.880 --> 00:08:25.880]   I'm not sure what you're doing.
[00:08:25.880 --> 00:08:28.880]   That's hardly a prompt.
[00:08:28.880 --> 00:08:30.880]   But you're right.
[00:08:30.880 --> 00:08:35.880]   What is your young man do for a living?
[00:08:35.880 --> 00:08:39.880]   He owns a used record store here in El Paso.
[00:08:39.880 --> 00:08:41.880]   Music lover, right?
[00:08:41.880 --> 00:08:44.880]   He's fond of music.
[00:08:44.880 --> 00:08:46.880]   Not me all.
[00:08:46.880 --> 00:08:54.880]   And what are you doing for a J.O.B. these days?
[00:08:55.880 --> 00:08:58.880]   I work in the record store.
[00:08:58.880 --> 00:09:06.880]   Oh, so it all suddenly seems so clear.
[00:09:06.880 --> 00:09:09.880]   Do you like it?
[00:09:09.880 --> 00:09:13.880]   Yeah, I like it a lot, smartass.
[00:09:13.880 --> 00:09:15.880]   I get to listen to music all day.
[00:09:15.880 --> 00:09:17.880]   Talk about music all day.
[00:09:17.880 --> 00:09:20.880]   It's really cool.
[00:09:20.880 --> 00:09:22.880]   It's going to be a great environment
[00:09:22.880 --> 00:09:27.880]   for a little girl to grow up in.
[00:09:27.880 --> 00:09:30.880]   As opposed to getting around the world,
[00:09:30.880 --> 00:09:33.880]   killing human beings, and being paid best
[00:09:33.880 --> 00:09:36.880]   sums of money.
[00:09:36.880 --> 00:09:38.880]   Precisely.
[00:09:38.880 --> 00:09:41.880]   Well, my own friend.
[00:09:41.880 --> 00:09:43.880]   Do we choose someone?
[00:09:43.880 --> 00:09:48.880]   However, all cocklockery aside,
[00:09:48.880 --> 00:09:51.880]   I am looking forward to meeting your young man.
[00:09:51.880 --> 00:09:54.880]   I happen to be more or less particular.
[00:09:54.880 --> 00:09:59.880]   Who might get married?
[00:09:59.880 --> 00:10:01.880]   You want to come to the wedding?
[00:10:01.880 --> 00:10:04.880]   Only if I can sit on the bright side.
[00:10:04.880 --> 00:10:07.880]   You'll find it a bit lonely on my side.
[00:10:07.880 --> 00:10:11.880]   Your side always was a bit lonely.
[00:10:11.880 --> 00:10:15.880]   But I wouldn't sit anywhere else.
[00:10:15.880 --> 00:10:20.880]   You know, I had a lovely stream of money.
[00:10:20.880 --> 00:10:22.880]   Lovely stream about you.
[00:10:22.880 --> 00:10:23.880]   Oh, here's Tommy.
[00:10:23.880 --> 00:10:25.880]   Call me, Arlene.
[00:10:25.880 --> 00:10:26.880]   You must be Tommy.
[00:10:26.880 --> 00:10:27.880]   Uh-huh.
[00:10:27.880 --> 00:10:29.880]   Arlene's told me so much about you.
[00:10:29.880 --> 00:10:30.880]   Aren't you okay?
[00:10:30.880 --> 00:10:31.880]   Oh, I'm fine.
[00:10:31.880 --> 00:10:34.880]   Tommy, I'd like you to meet my father.
[00:10:34.880 --> 00:10:37.880]   Oh, my God.
[00:10:37.880 --> 00:10:39.880]   Oh, my God, this is great.
[00:10:39.880 --> 00:10:41.880]   I'm so glad to meet you, sir.
[00:10:41.880 --> 00:10:42.880]   Oh, Dad.
[00:10:42.880 --> 00:10:44.880]   The name is Bill.
[00:10:44.880 --> 00:10:46.880]   Well, it's great to meet you.
[00:10:46.880 --> 00:10:47.880]   Bill.
[00:10:47.880 --> 00:10:49.880]   Arlene told me you could make it.
[00:10:49.880 --> 00:10:50.880]   Surprise.
[00:10:50.880 --> 00:10:52.880]   That's my pop for you.
[00:10:52.880 --> 00:10:54.880]   Always full of surprises.
[00:10:54.880 --> 00:11:00.880]   Well, in the surprise department, the apple doesn't fall far from the tree.
[00:11:00.880 --> 00:11:02.880]   When did you get in?
[00:11:02.880 --> 00:11:03.880]   Just now.
[00:11:03.880 --> 00:11:05.880]   Did you come straight from Australia?
[00:11:05.880 --> 00:11:06.880]   Of course.
[00:11:06.880 --> 00:11:08.880]   Daddy, I told Tommy that you were in Perth,
[00:11:08.880 --> 00:11:11.880]   lining for silver and no one could meet you.
[00:11:11.880 --> 00:11:15.880]   Lucky for us all, that's not the case.
[00:11:15.880 --> 00:11:19.880]   So, what's this all about?
[00:11:19.880 --> 00:11:26.880]   I've heard of wedding rehearsals, but I don't believe I've ever heard of a wedding dress rehearsal before.
[00:11:26.880 --> 00:11:33.880]   We thought, well, I paid so much money for a dress you only going to wear once, especially when Arlene looks so goddamn beautiful in it.
[00:11:33.880 --> 00:11:38.880]   So, uh, we're going to try to get all the mileage we can out of it.
[00:11:38.880 --> 00:11:43.880]   Isn't it supposed to be bad luck for the groom to see the bride and her wedding dress?
[00:11:43.880 --> 00:11:45.880]   People with a ceremony?
[00:11:45.880 --> 00:11:50.880]   Well, I guess I just believe I'm having dangerous.
[00:11:50.880 --> 00:11:53.880]   I know just what you mean.
[00:11:53.880 --> 00:11:54.880]   Some.
[00:11:54.880 --> 00:11:56.880]   Some of us are places to be.
[00:11:56.880 --> 00:11:58.880]   Show them to do.
[00:11:58.880 --> 00:12:01.880]   Look, we got to go through this one more time.
[00:12:01.880 --> 00:12:03.880]   So, uh, why don't you have a--
[00:12:03.880 --> 00:12:05.880]   Oh, my God.
[00:12:05.880 --> 00:12:06.880]   What am I thinking?
[00:12:06.880 --> 00:12:07.880]   You should give her away.
[00:12:07.880 --> 00:12:10.880]   Tommy, that's not exactly Daddy's cup of tea.
[00:12:10.880 --> 00:12:15.880]   I'm not even sure if you're much more comfortable sitting with the rest of the cats.
[00:12:15.880 --> 00:12:17.880]   Really?
[00:12:17.880 --> 00:12:20.880]   That's asking a lot.
[00:12:20.880 --> 00:12:21.880]   No.
[00:12:21.880 --> 00:12:22.880]   Okay.
[00:12:22.880 --> 00:12:23.880]   Forget it.
[00:12:23.880 --> 00:12:25.880]   But how about we go out to dinner tonight, celebrate?
[00:12:25.880 --> 00:12:28.880]   Only on the condition that I pay for everything.
[00:12:28.880 --> 00:12:29.880]   Deal.
[00:12:29.880 --> 00:12:31.880]   We have to do this now.
[00:12:31.880 --> 00:12:32.880]   Can I watch?
[00:12:32.880 --> 00:12:33.880]   Absolutely.
[00:12:33.880 --> 00:12:35.880]   Have a seat.
[00:12:35.880 --> 00:12:37.880]   Which is the bride's side.
[00:12:37.880 --> 00:12:39.880]   Right over here.
[00:12:39.880 --> 00:12:44.880]   Mother, here we go.
[00:12:44.880 --> 00:12:45.880]   Yes.
[00:12:45.880 --> 00:12:48.880]   Now, son, pop them, bowels.
[00:12:48.880 --> 00:12:49.880]   Yeah.
[00:12:49.880 --> 00:12:51.880]   You can't hear me.
[00:12:51.880 --> 00:12:53.880]   You should be.
[00:12:53.880 --> 00:12:54.880]   No.
[00:12:54.880 --> 00:12:58.880]   I just don't want to.
[00:12:58.880 --> 00:13:00.880]   You don't want me to put them there.
[00:13:00.880 --> 00:13:07.880]   If he's the man you want, then go stand by.
[00:13:07.880 --> 00:13:31.880]   [MUSIC]
[00:13:31.880 --> 00:13:33.880]   Does it look pretty?
[00:13:33.880 --> 00:13:35.880]   Oh, yeah.
[00:13:35.880 --> 00:13:45.880]   [MUSIC]
[00:13:45.880 --> 00:13:47.880]   Thank you.
[00:13:47.880 --> 00:13:52.880]   [MUSIC]
[00:13:52.880 --> 00:13:59.880]   [MUSIC]

ggml-base.en.bin

Some stats:

whisper_model_load: mem required  =  310.00 MB (+    6.00 MB per decoder)
whisper_model_load: model size    =  140.54 MB

whisper_print_timings:    total time = 433205.62 ms

So time needed was ~ 433 seconds

Actual transcription:

[00:00:00.000 --> 00:00:15.000]   [Music]
[00:00:15.000 --> 00:00:17.000]   Do you find me sadistic?
[00:00:17.000 --> 00:00:20.000]   No, Kato.
[00:00:20.000 --> 00:00:23.000]   I'd like to believe
[00:00:23.000 --> 00:00:27.000]   you're aware enough, even now.
[00:00:27.000 --> 00:00:33.000]   I know that there is nothing sadistic in my actions.
[00:00:33.000 --> 00:00:36.000]   This moment,
[00:00:36.000 --> 00:00:40.000]   this is me,
[00:00:40.000 --> 00:00:44.000]   and I most miss the case.
[00:00:44.000 --> 00:00:45.000]   Well,
[00:00:45.000 --> 00:00:48.000]   it's your baby.
[00:00:48.000 --> 00:00:54.000]   The dead didn't I?
[00:00:55.000 --> 00:00:57.000]   Well, I wasn't.
[00:00:57.000 --> 00:01:00.000]   But it wasn't from lack of try and I can tell you that.
[00:01:00.000 --> 00:01:03.000]   Actually, Bill's last bullet put me in a coma.
[00:01:03.000 --> 00:01:07.000]   A coma I was to lie in for four years.
[00:01:07.000 --> 00:01:09.000]   When I woke up,
[00:01:09.000 --> 00:01:15.000]   I went on with the movie advertisements referred to as a roaring rampage of revenge.
[00:01:15.000 --> 00:01:17.000]   I roared,
[00:01:17.000 --> 00:01:19.000]   and I rampaged,
[00:01:19.000 --> 00:01:22.000]   and I got bloody satisfaction.
[00:01:22.000 --> 00:01:26.000]   I've killed a hell of a lot of people who get to this point.
[00:01:26.000 --> 00:01:29.000]   But I have only one more.
[00:01:29.000 --> 00:01:31.000]   The last one.
[00:01:31.000 --> 00:01:35.000]   The one I'm driving to right now.
[00:01:35.000 --> 00:01:38.000]   The only one left.
[00:01:38.000 --> 00:01:41.000]   And when I arrive at my destination,
[00:01:41.000 --> 00:01:45.000]   I am gonna kill Bill.
[00:01:45.000 --> 00:02:02.000]   [Music]
[00:02:02.000 --> 00:02:14.000]   Now the incident that happened at the two pines wedding chapel
[00:02:14.000 --> 00:02:20.000]   that put this whole gory story into motion has since become legend.
[00:02:20.000 --> 00:02:23.000]   Massacre at two pines.
[00:02:23.000 --> 00:02:25.000]   That's what the newspapers called it.
[00:02:25.000 --> 00:02:31.000]   The local TV news called it the El Paso Texas Wedding Chapel Massacre.
[00:02:31.000 --> 00:02:32.000]   How it happened?
[00:02:32.000 --> 00:02:33.000]   Who was there?
[00:02:33.000 --> 00:02:36.000]   How many got killed and who killed them?
[00:02:36.000 --> 00:02:40.000]   Changes depending on who's telling the story.
[00:02:40.000 --> 00:02:42.000]   In actual fact,
[00:02:42.000 --> 00:02:46.000]   the massacre didn't happen during a wedding at all.
[00:02:46.000 --> 00:02:48.000]   It was a wedding rehearsal.
[00:02:48.000 --> 00:02:53.000]   Now when we come to the park where I say you may kiss the bride,
[00:02:53.000 --> 00:02:55.000]   you may kiss the bride.
[00:02:55.000 --> 00:02:59.000]   But don't stick your tongue in her mouth.
[00:02:59.000 --> 00:03:02.000]   This might be funny to your friends,
[00:03:02.000 --> 00:03:06.000]   but it would be embarrassing to your parents.
[00:03:06.000 --> 00:03:11.000]   We'll try to be strange.
[00:03:11.000 --> 00:03:16.000]   Y'all got a song?
[00:03:16.000 --> 00:03:18.000]   How about "Love Me Tender"?
[00:03:18.000 --> 00:03:22.000]   I can play that.
[00:03:22.000 --> 00:03:24.000]   Let me tend to be great.
[00:03:24.000 --> 00:03:27.000]   Rufus, he's the man.
[00:03:27.000 --> 00:03:28.000]   Rufus?
[00:03:28.000 --> 00:03:30.000]   Who was that to use to play for?
[00:03:30.000 --> 00:03:32.000]   Rufus Thomas.
[00:03:32.000 --> 00:03:34.000]   Rufus Thomas.
[00:03:34.000 --> 00:03:36.000]   Rufus Thomas.
[00:03:36.000 --> 00:03:37.000]   I was a drill.
[00:03:37.000 --> 00:03:38.000]   I was a drifted.
[00:03:38.000 --> 00:03:40.000]   I was a coaster.
[00:03:40.000 --> 00:03:42.000]   I was part of the gang.
[00:03:42.000 --> 00:03:44.000]   I was a bar-k.
[00:03:44.000 --> 00:03:48.000]   If they come through Texas, I'd play with him.
[00:03:48.000 --> 00:03:50.000]   Rufus?
[00:03:50.000 --> 00:03:54.000]   He's the man.
[00:03:54.000 --> 00:03:57.000]   Have I forgot anything?
[00:03:57.000 --> 00:04:00.000]   Oh yes, you forgot the seating arrangements.
[00:04:00.000 --> 00:04:03.000]   Thank you, mother.
[00:04:03.000 --> 00:04:05.000]   Now the way we normally do this,
[00:04:05.000 --> 00:04:08.000]   we have the bride's side,
[00:04:08.000 --> 00:04:12.000]   but since the bride ain't got nobody coming,
[00:04:12.000 --> 00:04:16.000]   and the groom's got far too many people coming,
[00:04:16.000 --> 00:04:20.000]   well yeah, they're coming all the way from Oklahoma.
[00:04:20.000 --> 00:04:22.000]   Right.
[00:04:22.000 --> 00:04:27.000]   Well, I don't see no problem with the groom's side
[00:04:27.000 --> 00:04:29.000]   sharing the bride's side.
[00:04:29.000 --> 00:04:30.000]   Do you, mother?
[00:04:30.000 --> 00:04:32.000]   Not a problem with that,
[00:04:32.000 --> 00:04:37.000]   but honey, you know, it would be good if you had somebody come
[00:04:37.000 --> 00:04:41.000]   You know, is it sign of good faith?
[00:04:41.000 --> 00:04:44.000]   Well, I don't have anybody,
[00:04:44.000 --> 00:04:48.000]   except for Tommy and my friends.
[00:04:48.000 --> 00:04:51.000]   You have no family?
[00:04:51.000 --> 00:04:53.000]   Well, I'm working on changing then.
[00:04:53.000 --> 00:04:55.000]   Mrs. Harmony, we're all the family
[00:04:55.000 --> 00:04:58.000]   this Alangel's ever going to need.
[00:04:58.000 --> 00:05:00.000]   I'm not feeling very well,
[00:05:00.000 --> 00:05:03.000]   and this bitch is starting to piss me off.
[00:05:03.000 --> 00:05:05.000]   So while you all bother on,
[00:05:05.000 --> 00:05:08.000]   I'm going to go outside and get some air.
[00:05:08.000 --> 00:05:10.000]   I'm sorry.
[00:05:10.000 --> 00:05:12.000]   She's going to go out and get some air.
[00:05:12.000 --> 00:05:14.000]   Yeah, given her delicate condition,
[00:05:14.000 --> 00:05:17.000]   she just needs a few minutes to give it to get us to be okay.
[00:05:18.000 --> 00:05:21.000]   [music]
[00:05:22.000 --> 00:05:25.000]   [music]
[00:05:26.000 --> 00:05:29.000]   [music]
[00:05:30.000 --> 00:05:33.000]   [music]
[00:05:34.000 --> 00:05:37.000]   [music]
[00:05:38.000 --> 00:05:41.000]   [music]
[00:05:42.000 --> 00:05:45.000]   [music]
[00:05:46.000 --> 00:05:49.000]   [music]
[00:05:50.000 --> 00:05:53.000]   [music]
[00:05:53.000 --> 00:05:57.000]   [music]
[00:05:57.000 --> 00:06:01.000]   [music]
[00:06:01.000 --> 00:06:05.000]   [music]
[00:06:05.000 --> 00:06:09.000]   [music]
[00:06:09.000 --> 00:06:13.000]   [music]
[00:06:36.000 --> 00:06:40.000]   Hello, kiddo.
[00:06:40.000 --> 00:06:46.000]   How did you find me?
[00:06:46.000 --> 00:06:50.000]   I'm the man.
[00:06:50.000 --> 00:06:54.000]   What are you doing here?
[00:06:54.000 --> 00:06:58.000]   What am I doing?
[00:06:58.000 --> 00:07:00.000]   Well,
[00:07:00.000 --> 00:07:06.000]   I was playing my flute.
[00:07:06.000 --> 00:07:12.000]   At this moment,
[00:07:12.000 --> 00:07:15.000]   I'm looking at the most beautiful bride
[00:07:15.000 --> 00:07:19.000]   in these old eyes of every scene.
[00:07:19.000 --> 00:07:21.000]   Why are you here?
[00:07:21.000 --> 00:07:25.000]   Last look.
[00:07:25.000 --> 00:07:27.000]   Are you going to be nice?
[00:07:27.000 --> 00:07:31.000]   I've never been nice my whole life.
[00:07:31.000 --> 00:07:36.000]   But I'll do my best to be sweet.
[00:07:36.000 --> 00:07:39.000]   I always told you,
[00:07:39.000 --> 00:07:42.000]   your sweet side is your best side.
[00:07:42.000 --> 00:07:49.000]   I guess that's why you're the only one who's ever seen it.
[00:07:49.000 --> 00:07:54.000]   See, you got a bun in the oven?
[00:07:54.000 --> 00:07:58.000]   I'm knocked up.
[00:07:58.000 --> 00:08:00.000]   Chase Louise.
[00:08:00.000 --> 00:08:02.000]   That young man here is sure
[00:08:02.000 --> 00:08:06.000]   it doesn't believe in wasting time, does he?
[00:08:06.000 --> 00:08:10.000]   Have you seen Tommy?
[00:08:10.000 --> 00:08:12.000]   A guy on the tux?
[00:08:12.000 --> 00:08:13.000]   Yes.
[00:08:13.000 --> 00:08:16.000]   And I saw him.
[00:08:16.000 --> 00:08:20.000]   I like his hair.
[00:08:20.000 --> 00:08:24.000]   You promised you'd be nice.
[00:08:24.000 --> 00:08:26.000]   I said I'd do my best.
[00:08:26.000 --> 00:08:29.000]   That's hardly a promise.
[00:08:29.000 --> 00:08:31.000]   But you're right.
[00:08:31.000 --> 00:08:36.000]   What does your young man do for a living?
[00:08:36.000 --> 00:08:40.000]   He owns a used record store here in El Paso.
[00:08:40.000 --> 00:08:42.000]   He's a lover, right?
[00:08:42.000 --> 00:08:45.000]   He's fond of music.
[00:08:45.000 --> 00:08:51.000]   Aren't we all?
[00:08:51.000 --> 00:08:56.000]   And what are you doing for a J.O.B. these days?
[00:08:56.000 --> 00:08:59.000]   I work in the record store.
[00:08:59.000 --> 00:09:03.000]   Ah, so...
[00:09:03.000 --> 00:09:08.000]   it all suddenly seems so clear.
[00:09:08.000 --> 00:09:10.000]   Do you like it?
[00:09:10.000 --> 00:09:14.000]   Yeah, I like it a lot, smart ass.
[00:09:14.000 --> 00:09:17.000]   I get to listen to music all day.
[00:09:17.000 --> 00:09:21.000]   Talk about music all day. It's really cool.
[00:09:21.000 --> 00:09:28.000]   It's going to be a great environment for my little girl to grow up in.
[00:09:28.000 --> 00:09:33.000]   As opposed to jetting around the world, killing human beams,
[00:09:33.000 --> 00:09:38.000]   and being paid best sums of money.
[00:09:38.000 --> 00:09:40.000]   Precisely.
[00:09:40.000 --> 00:09:44.000]   I have a friend.
[00:09:44.000 --> 00:09:47.000]   Do each you own.
[00:09:47.000 --> 00:09:51.000]   However, all clockwork re-assigned.
[00:09:51.000 --> 00:09:55.000]   I am looking forward to meeting your young man.
[00:09:55.000 --> 00:09:59.000]   I happen to be more or less particular.
[00:09:59.000 --> 00:10:03.000]   Oh, my God, Mary.
[00:10:03.000 --> 00:10:05.000]   You want to come to the wedding?
[00:10:05.000 --> 00:10:08.000]   Only if I can sit on the bride's side.
[00:10:08.000 --> 00:10:12.000]   Your side always was a bit lonely.
[00:10:12.000 --> 00:10:17.000]   But I wouldn't sit anywhere else.
[00:10:17.000 --> 00:10:22.000]   You know, I had a lovely stream about you.
[00:10:22.000 --> 00:10:25.000]   Oh, here's Tommy. Call me Arlene.
[00:10:25.000 --> 00:10:29.000]   You must be Tommy. Arlene's told me so much about you.
[00:10:29.000 --> 00:10:31.000]   Are you okay?
[00:10:31.000 --> 00:10:35.000]   Oh, I'm fine. Tommy, I'd like you to meet my father.
[00:10:35.000 --> 00:10:38.000]   Oh, my God.
[00:10:38.000 --> 00:10:40.000]   Oh, my God, this is great.
[00:10:40.000 --> 00:10:42.000]   I'm so glad to meet you, sir.
[00:10:42.000 --> 00:10:45.000]   Oh, Dad, the name is Bill.
[00:10:45.000 --> 00:10:48.000]   Well, it's great to meet you. Bill.
[00:10:48.000 --> 00:10:50.000]   Arlene told me you could make it.
[00:10:50.000 --> 00:10:51.000]   Surprise.
[00:10:51.000 --> 00:10:54.000]   That's my pot for you. Always full of surprises.
[00:10:54.000 --> 00:10:58.000]   Well, in a surprise department.
[00:10:58.000 --> 00:11:01.000]   The Apple doesn't fall far from the tree.
[00:11:01.000 --> 00:11:04.000]   When did you get in? Just now.
[00:11:04.000 --> 00:11:06.000]   Did you come straight from Australia?
[00:11:06.000 --> 00:11:07.000]   Of course.
[00:11:07.000 --> 00:11:09.000]   Daddy, I told Tommy that you were in Perth,
[00:11:09.000 --> 00:11:13.000]   finding for silver, and no one could reach you.
[00:11:13.000 --> 00:11:16.000]   Lucky for us all, that's not the case.
[00:11:16.000 --> 00:11:20.000]   So, what's this all about?
[00:11:20.000 --> 00:11:27.000]   I've heard of wedding rehearsals, but I don't believe I've ever heard of a wedding dress rehearsal before.
[00:11:27.000 --> 00:11:31.000]   We thought, "Why pay so much money for a dress you only gonna wear once?"
[00:11:31.000 --> 00:11:34.000]   Especially when Arlene looks so goddamn beautiful in it.
[00:11:34.000 --> 00:11:39.000]   So, we're gonna try to get all the mileage we can out of it.
[00:11:39.000 --> 00:11:44.000]   Isn't it supposed to be bad luck for the groom to see the bride in her wedding dress?
[00:11:44.000 --> 00:11:46.000]   People with a ceremony?
[00:11:46.000 --> 00:11:51.000]   Wow. I guess I just believe in them dangerously.
[00:11:51.000 --> 00:11:54.000]   I know just what you mean.
[00:11:54.000 --> 00:11:59.000]   "San, Sama Lacha, places to be. It's your duty."
[00:11:59.000 --> 00:12:02.000]   But we gotta go through this one more time.
[00:12:02.000 --> 00:12:05.000]   So, uh, why don't you have a... oh my god.
[00:12:05.000 --> 00:12:08.000]   What am I thinking? You should give her away!
[00:12:08.000 --> 00:12:11.000]   Tommy, that's not exactly Daddy's cup of tea.
[00:12:11.000 --> 00:12:16.000]   I think Father seemed much more comfortable sitting with the rest of the guests.
[00:12:16.000 --> 00:12:18.000]   Really?
[00:12:18.000 --> 00:12:21.000]   That's asking a lot.
[00:12:21.000 --> 00:12:24.000]   Oh. Okay. We'll forget it.
[00:12:24.000 --> 00:12:26.000]   But how about we go out to dinner tonight and celebrate?
[00:12:26.000 --> 00:12:29.000]   Only on the condition that I pay for everything.
[00:12:29.000 --> 00:12:32.000]   Deal. We have to do this now.
[00:12:32.000 --> 00:12:33.000]   Can I watch?
[00:12:33.000 --> 00:12:36.000]   Absolutely. You have a seat.
[00:12:36.000 --> 00:12:38.000]   Which is the bride's side?
[00:12:38.000 --> 00:12:40.000]   Right over here.
[00:12:40.000 --> 00:12:45.000]   Father, here we go.
[00:12:45.000 --> 00:12:50.000]   Yeah. Now, Sean, drop them bows.
[00:12:50.000 --> 00:12:53.000]   Yeah.
[00:12:53.000 --> 00:12:56.000]   Oh, Sean.
[00:12:56.000 --> 00:12:59.000]   Oh. I just want...
[00:12:59.000 --> 00:13:03.000]   You don't want me a damn thing.
[00:13:03.000 --> 00:13:08.000]   If he's the man you want, then go stand by.
[00:13:08.000 --> 00:13:11.000]   [Sighs]
[00:13:11.000 --> 00:13:34.000]   Does it look pretty?
[00:13:34.000 --> 00:13:37.000]   Oh, yes.
[00:13:37.000 --> 00:13:40.000]   [Sighs]
[00:13:40.000 --> 00:13:49.000]   Thank you.
[00:13:49.000 --> 00:13:56.000]   [Sighs]
[00:13:56.000 --> 00:14:00.000]   [Music]

ggml-small.en.bin

Some stats:

whisper_model_load: mem required  =  743.00 MB (+   16.00 MB per decoder)
whisper_model_load: model size    = 464.44 MB

whisper_print_timings:    total time = 1713762.12 ms

So time needed was ~ 1713 seconds (almost half an hour)

And the actual transcription:

[00:00:00.000 --> 00:00:10.000]   [MUSIC]
[00:00:10.000 --> 00:00:17.000]   Do you find me sadistic?
[00:00:17.000 --> 00:00:21.000]   No, kiddo.
[00:00:21.000 --> 00:00:27.000]   I'd like to believe you're aware enough, even now,
[00:00:27.000 --> 00:00:33.000]   to know that there is nothing sadistic in my actions.
[00:00:33.000 --> 00:00:44.000]   This moment, this is me and my most nicer kiss to be.
[00:00:44.000 --> 00:00:48.000]   Well, it's your baby.
[00:00:48.000 --> 00:00:55.000]   Look, Dad, didn't I?
[00:00:55.000 --> 00:01:00.000]   Well, I wasn't. But it wasn't from lack of trying, I can tell you that.
[00:01:00.000 --> 00:01:03.000]   Actually, Bill's last bullet put me in a coma.
[00:01:03.000 --> 00:01:07.000]   A coma I was to lie in for four years.
[00:01:07.000 --> 00:01:11.000]   When I woke up, I went on with the movie advertisements
[00:01:11.000 --> 00:01:15.000]   referred to as a roaring rampage of revenge.
[00:01:15.000 --> 00:01:22.000]   I roared, and I rampaged, and I got bloody satisfaction.
[00:01:22.000 --> 00:01:26.000]   I've killed a hell of a lot of people to get to this point.
[00:01:26.000 --> 00:01:29.000]   But I have only one more.
[00:01:29.000 --> 00:01:31.000]   The last one.
[00:01:31.000 --> 00:01:35.000]   The one I'm driving to right now.
[00:01:35.000 --> 00:01:38.000]   The only one left.
[00:01:38.000 --> 00:01:45.000]   And when I arrive at my destination, I am gonna kill Bill.
[00:01:45.000 --> 00:02:10.000]   [Music]
[00:02:10.000 --> 00:02:14.000]   Now, the incident that happened at the Two Pines Wedding Chapel
[00:02:14.000 --> 00:02:20.000]   that put this whole gory story into motion has since become legend.
[00:02:20.000 --> 00:02:22.000]   Massacre at Two Pines.
[00:02:22.000 --> 00:02:24.000]   That's what the newspapers called it.
[00:02:24.000 --> 00:02:30.000]   The local TV news called it the El Paso, Texas Wedding Chapel Massacre.
[00:02:30.000 --> 00:02:36.000]   How it happened, who was there, how many got killed, and who killed them.
[00:02:36.000 --> 00:02:40.000]   Changes depending on who's telling the story.
[00:02:40.000 --> 00:02:45.000]   In actual fact, the massacre didn't happen during a wedding at all.
[00:02:45.000 --> 00:02:48.000]   It was a wedding rehearsal.
[00:02:48.000 --> 00:02:53.000]   Now, when we come to the park where I say you may kiss the bride,
[00:02:53.000 --> 00:02:55.000]   you may kiss the bride.
[00:02:55.000 --> 00:02:59.000]   But don't stick your tongue in her mouth.
[00:02:59.000 --> 00:03:06.000]   This might be funny to your friends, but it would be embarrassing to your parents.
[00:03:06.000 --> 00:03:10.000]   We'll try and restrain ourselves from that.
[00:03:10.000 --> 00:03:16.000]   Y'all got a song?
[00:03:16.000 --> 00:03:19.000]   How about "Love Me, Tender"? I can play that.
[00:03:19.000 --> 00:03:21.000]   Sure.
[00:03:21.000 --> 00:03:24.000]   "Love Me, Tender" would be great.
[00:03:24.000 --> 00:03:27.000]   Rufus, he's the man.
[00:03:27.000 --> 00:03:30.000]   Rufus, who was that you used to play for?
[00:03:30.000 --> 00:03:32.000]   Rufus Thomas.
[00:03:32.000 --> 00:03:34.000]   Rufus Thomas.
[00:03:34.000 --> 00:03:35.000]   Rufus Thomas.
[00:03:35.000 --> 00:03:39.000]   I was a drill, I was a drifter, I was a coaster,
[00:03:39.000 --> 00:03:43.000]   I was part of the gang, I was a bar-k.
[00:03:43.000 --> 00:03:47.000]   If they come through Texas, I haven't played with them.
[00:03:47.000 --> 00:03:53.000]   Rufus, he's the man.
[00:03:53.000 --> 00:03:56.000]   Have you forgotten anything?
[00:03:56.000 --> 00:04:00.000]   Oh, yes, you forgot the seating arrangements.
[00:04:00.000 --> 00:04:02.000]   Thank you, Mother.
[00:04:02.000 --> 00:04:07.000]   Now, the way we normally do this, we have the bride's side,
[00:04:07.000 --> 00:04:09.000]   and then we have the groom's side.
[00:04:09.000 --> 00:04:12.000]   But since the bride ain't got nobody coming,
[00:04:12.000 --> 00:04:16.000]   and the groom's got far too many people coming...
[00:04:16.000 --> 00:04:19.000]   Well, yeah, they're coming all the way from Oklahoma.
[00:04:19.000 --> 00:04:22.000]   Right.
[00:04:22.000 --> 00:04:29.000]   Well, I don't see no problem with the groom's side sharing the bride's side.
[00:04:29.000 --> 00:04:30.000]   Do you, Mother?
[00:04:30.000 --> 00:04:32.000]   I don't have a problem with that.
[00:04:32.000 --> 00:04:38.000]   But, honey, you know, it would be good if you had somebody come.
[00:04:38.000 --> 00:04:41.000]   You know, is that a sign of good faith?
[00:04:41.000 --> 00:04:48.000]   Well, I don't have anybody, except for Tommy and my friends.
[00:04:48.000 --> 00:04:51.000]   You have no family?
[00:04:51.000 --> 00:04:53.000]   Well, I'm working on changing that.
[00:04:53.000 --> 00:04:57.000]   Mrs. Harmony, we're all the family this little angel's ever gonna need.
[00:04:57.000 --> 00:05:03.000]   I'm not feeling very well, and this bitch is starting to piss me off.
[00:05:03.000 --> 00:05:07.000]   So while you all blather on, I'm gonna go outside and get some air.
[00:05:07.000 --> 00:05:09.000]   Um, uh, Reverend, sorry.
[00:05:09.000 --> 00:05:11.000]   She's gonna go out and get some air.
[00:05:11.000 --> 00:05:13.000]   Yeah, given her delicate condition.
[00:05:13.000 --> 00:05:16.000]   She just needs a few minutes to get it together. She'll be okay.
[00:05:16.000 --> 00:05:18.000]   Okay.
[00:05:19.000 --> 00:05:22.000]   [♪♪♪]
[00:05:23.000 --> 00:05:25.000]   [♪♪♪]
[00:05:26.000 --> 00:05:28.000]   [♪♪♪]
[00:05:28.000 --> 00:05:30.000]   [♪♪♪]
[00:05:30.000 --> 00:05:32.000]   [♪♪♪]
[00:05:33.000 --> 00:05:35.000]   [♪♪♪]
[00:05:36.000 --> 00:05:38.000]   [♪♪♪]
[00:05:39.000 --> 00:05:41.000]   [♪♪♪]
[00:05:41.000 --> 00:05:43.000]   [♪♪♪]
[00:05:43.000 --> 00:05:45.000]   [♪♪♪]
[00:05:45.000 --> 00:05:47.000]   [♪♪♪]
[00:05:47.000 --> 00:05:49.000]   [♪♪♪]
[00:05:49.000 --> 00:05:51.000]   [♪♪♪]
[00:05:51.000 --> 00:05:53.000]   [♪♪♪]
[00:05:53.000 --> 00:05:55.000]   [♪♪♪]
[00:05:55.000 --> 00:05:57.000]   [♪♪♪]
[00:05:57.000 --> 00:05:59.000]   [♪♪♪]
[00:05:59.000 --> 00:06:01.000]   [♪♪♪]
[00:06:01.000 --> 00:06:03.000]   [♪♪♪]
[00:06:03.000 --> 00:06:05.000]   [♪♪♪]
[00:06:05.000 --> 00:06:07.000]   [♪♪♪]
[00:06:07.000 --> 00:06:09.000]   [♪♪♪]
[00:06:36.000 --> 00:06:38.000]   Hello, kiddo.
[00:06:38.000 --> 00:06:45.000]   How did you find me?
[00:06:45.000 --> 00:06:48.000]   I'm the man.
[00:06:48.000 --> 00:06:53.000]   What are you doing here?
[00:06:53.000 --> 00:06:57.000]   What am I doing?
[00:06:57.000 --> 00:07:03.000]   Well, a moment ago I was playing my flute.
[00:07:04.000 --> 00:07:06.000]   [♪♪♪]
[00:07:06.000 --> 00:07:15.000]   At this moment, I'm looking at the most beautiful bride
[00:07:15.000 --> 00:07:18.000]   these old eyes have ever seen.
[00:07:18.000 --> 00:07:22.000]   Why are you here?
[00:07:22.000 --> 00:07:25.000]   Last look.
[00:07:25.000 --> 00:07:28.000]   Are you gonna be nice?
[00:07:28.000 --> 00:07:30.000]   I've never been nice my whole life.
[00:07:32.000 --> 00:07:34.000]   I'm just a guest to be sweet.
[00:07:34.000 --> 00:07:42.000]   I always told you, your sweet side is your best side.
[00:07:42.000 --> 00:07:47.000]   I guess that's why you're the only one who's ever seen it.
[00:07:47.000 --> 00:07:52.000]   See, you got a bun in the oven.
[00:07:52.000 --> 00:07:57.000]   I'm knocked up.
[00:07:58.000 --> 00:08:01.000]   I'm not a man of the way.
[00:08:01.000 --> 00:08:04.000]   I'm not a man of the way.
[00:08:04.000 --> 00:08:07.000]   I'm not a man of the way.
[00:08:07.000 --> 00:08:10.000]   I'm not a man of the way.
[00:08:10.000 --> 00:08:13.000]   I'm not a man of the way.
[00:08:13.000 --> 00:08:16.000]   I'm not a man of the way.
[00:08:16.000 --> 00:08:19.000]   I'm not a man of the way.
[00:08:19.000 --> 00:08:22.000]   I'm not a man of the way.
[00:08:22.000 --> 00:08:25.000]   I'm not a man of the way.
[00:08:26.000 --> 00:08:28.000]   It's hardly a promise.
[00:08:28.000 --> 00:08:31.000]   But you're right.
[00:08:31.000 --> 00:08:34.000]   What does your young man do for a living?
[00:08:34.000 --> 00:08:39.000]   He owns a used record store here in El Paso.
[00:08:39.000 --> 00:08:42.000]   A music lover, right?
[00:08:42.000 --> 00:08:44.000]   He's fond of music.
[00:08:44.000 --> 00:08:47.000]   Aren't we all?
[00:08:47.000 --> 00:08:55.000]   And what are you doing for a J-O-B these days?
[00:08:56.000 --> 00:08:58.000]   I work in the record store.
[00:08:58.000 --> 00:09:01.000]   Ah, so...
[00:09:01.000 --> 00:09:06.000]   it all suddenly seems so clear.
[00:09:06.000 --> 00:09:10.000]   Do you like it?
[00:09:10.000 --> 00:09:13.000]   Yeah, I like it a lot, smartass.
[00:09:13.000 --> 00:09:16.000]   I get to listen to music all day,
[00:09:16.000 --> 00:09:19.000]   talk about music all day. It's really cool.
[00:09:19.000 --> 00:09:24.000]   It's gonna be a great environment for my little girl to grow up in.
[00:09:24.000 --> 00:09:30.000]   As opposed to jetting around the world,
[00:09:30.000 --> 00:09:32.000]   killing human beings,
[00:09:32.000 --> 00:09:35.000]   and being paid vast sums of money.
[00:09:35.000 --> 00:09:39.000]   Precisely.
[00:09:39.000 --> 00:09:41.000]   Well, my old friend,
[00:09:41.000 --> 00:09:43.000]   to each his own.
[00:09:43.000 --> 00:09:45.000]   However,
[00:09:45.000 --> 00:09:49.000]   all cock-blockery aside,
[00:09:49.000 --> 00:09:52.000]   I am looking forward to meeting your young man.
[00:09:52.000 --> 00:09:55.000]   I happen to be more or less particular
[00:09:55.000 --> 00:09:58.000]   who my gal marries.
[00:09:58.000 --> 00:10:02.000]   You wanna come to the wedding?
[00:10:02.000 --> 00:10:04.000]   Only if I can sit on the bride's side.
[00:10:04.000 --> 00:10:08.000]   You'll find it a bit lonely on my side.
[00:10:08.000 --> 00:10:12.000]   Your side always was a bit lonely,
[00:10:12.000 --> 00:10:15.000]   but I wouldn't sit anywhere else.
[00:10:15.000 --> 00:10:19.000]   You know,
[00:10:19.000 --> 00:10:22.000]   I had the loveliest dream about you.
[00:10:22.000 --> 00:10:24.000]   Oh, here's Tommy. Call me Arlene.
[00:10:24.000 --> 00:10:27.000]   You must be Tommy.
[00:10:27.000 --> 00:10:29.000]   Arlene's told me so much about you.
[00:10:29.000 --> 00:10:31.000]   Honey, you okay?
[00:10:31.000 --> 00:10:32.000]   Oh, I'm fine.
[00:10:32.000 --> 00:10:35.000]   Tommy, I'd like you to meet my father.
[00:10:35.000 --> 00:10:38.000]   Oh, my God.
[00:10:38.000 --> 00:10:40.000]   Oh, my God, this is great.
[00:10:40.000 --> 00:10:42.000]   I'm so glad to meet you, sir.
[00:10:42.000 --> 00:10:43.000]   Oh, Dad.
[00:10:43.000 --> 00:10:45.000]   The name's Bill.
[00:10:45.000 --> 00:10:47.000]   Well, it's great to meet you.
[00:10:47.000 --> 00:10:50.000]   So, Arlene told me you couldn't make it.
[00:10:50.000 --> 00:10:51.000]   Surprise.
[00:10:51.000 --> 00:10:53.000]   That's my pot for you.
[00:10:53.000 --> 00:10:55.000]   Always full of surprises.
[00:10:55.000 --> 00:10:58.000]   Well, in the surprise department,
[00:10:58.000 --> 00:11:01.000]   the apple doesn't fall far from the tree.
[00:11:01.000 --> 00:11:03.000]   When did you get in?
[00:11:03.000 --> 00:11:04.000]   Just now.
[00:11:04.000 --> 00:11:06.000]   Did you come straight from Australia?
[00:11:06.000 --> 00:11:07.000]   Of course.
[00:11:07.000 --> 00:11:09.000]   Daddy, I told Tommy that you were in Perth
[00:11:09.000 --> 00:11:12.000]   lining for Silver and no one could reach you.
[00:11:12.000 --> 00:11:16.000]   Lucky for us all, that's not the case.
[00:11:16.000 --> 00:11:20.000]   So, what's this all about?
[00:11:20.000 --> 00:11:22.000]   I've heard of wedding rehearsals,
[00:11:22.000 --> 00:11:24.000]   but I don't believe I've ever heard
[00:11:24.000 --> 00:11:26.000]   of a wedding dress rehearsal before.
[00:11:26.000 --> 00:11:29.000]   We thought, why pay so much money for a dress
[00:11:29.000 --> 00:11:31.000]   you're only gonna wear once?
[00:11:31.000 --> 00:11:34.000]   Especially when Arlene looks so goddamn beautiful in it.
[00:11:34.000 --> 00:11:38.000]   So, I think we're gonna try to get all the mileage we can out of it.
[00:11:38.000 --> 00:11:41.000]   Isn't it supposed to be bad luck
[00:11:41.000 --> 00:11:44.000]   for the groom to see the bride in her wedding dress?
[00:11:44.000 --> 00:11:46.000]   People of the ceremony?
[00:11:46.000 --> 00:11:50.000]   I guess I just believe in living dangerously.
[00:11:50.000 --> 00:11:54.000]   I know just what you mean.
[00:11:54.000 --> 00:11:57.000]   Some...some of us are places to be.
[00:11:57.000 --> 00:11:59.000]   It's your duty.
[00:11:59.000 --> 00:12:01.000]   Look, we gotta go through this one more time.
[00:12:01.000 --> 00:12:03.000]   So, why don't you have a s...
[00:12:03.000 --> 00:12:05.000]   Oh, my God.
[00:12:05.000 --> 00:12:07.000]   What am I thinking? You should give her away.
[00:12:07.000 --> 00:12:11.000]   Tommy, that's not exactly Daddy's cup of tea.
[00:12:11.000 --> 00:12:14.000]   I think Father would be much more comfortable
[00:12:14.000 --> 00:12:16.000]   sitting with the rest of the guests.
[00:12:16.000 --> 00:12:18.000]   Really?
[00:12:18.000 --> 00:12:20.000]   That's asking a lot.
[00:12:20.000 --> 00:12:24.000]   Oh. Okay. We'll forget it.
[00:12:24.000 --> 00:12:26.000]   But how about we go out to dinner tonight and celebrate?
[00:12:26.000 --> 00:12:29.000]   Only on the condition that I pay for everything.
[00:12:29.000 --> 00:12:32.000]   Deal. We gotta do this now.
[00:12:32.000 --> 00:12:33.000]   Can I watch?
[00:12:33.000 --> 00:12:35.000]   Absolutely. Have a seat.
[00:12:35.000 --> 00:12:38.000]   Which is the bride's side?
[00:12:38.000 --> 00:12:40.000]   Right over here.
[00:12:41.000 --> 00:12:44.000]   Father, here we go.
[00:12:44.000 --> 00:12:45.000]   Yes.
[00:12:45.000 --> 00:12:48.000]   Now, son, about them vows.
[00:12:48.000 --> 00:12:57.000]   Belle.
[00:12:57.000 --> 00:12:58.000]   I just don't want...
[00:12:58.000 --> 00:13:01.000]   You know only a damn thing.
[00:13:01.000 --> 00:13:04.000]   If he's the man you want,
[00:13:04.000 --> 00:13:07.000]   then go stand by.
[00:13:07.000 --> 00:13:09.000]   Stand by.
[00:13:09.000 --> 00:13:33.000]   Do I look pretty?
[00:13:33.000 --> 00:13:35.000]   Oh, yes.
[00:13:35.000 --> 00:13:37.000]   Thank you.
[00:13:38.000 --> 00:13:40.000]   Thank you.
[00:13:41.000 --> 00:13:43.000]   Thank you.
[00:13:44.000 --> 00:13:46.000]   Thank you.
[00:13:47.000 --> 00:13:49.000]   Thank you.
[00:13:49.000 --> 00:13:52.000]   [♪♪♪]
[00:13:53.000 --> 00:13:55.000]   [♪♪♪]
[00:13:55.000 --> 00:13:57.420]   (soft music)
[00:13:57.420 --> 00:13:59.320]   (slow, dramatic music)

ggml-medium.en.bin

Some stats:

whisper_model_load: mem required  = 1899.00 MB (+   43.00 MB per decoder)
whisper_model_load: model size    = 1462.12 MB

whisper_print_timings:    total time = 3563774.75 ms

So time needed was ~ 3563 seconds (almost 1 hour)

And the actual transcription:

[00:00:00.000 --> 00:00:15.000]   [Music]
[00:00:15.000 --> 00:00:19.000]   Do you find me sadistic?
[00:00:19.000 --> 00:00:21.000]   No, kiddo.
[00:00:21.000 --> 00:00:24.000]   I'd like to believe
[00:00:24.000 --> 00:00:27.000]   you're aware enough, even now,
[00:00:27.000 --> 00:00:31.000]   to know that there's nothing sadistic
[00:00:31.000 --> 00:00:34.000]   in my actions.
[00:00:34.000 --> 00:00:38.000]   This moment,
[00:00:38.000 --> 00:00:41.000]   this is me
[00:00:41.000 --> 00:00:44.000]   and my most masochistic.
[00:00:44.000 --> 00:00:46.000]   Well,
[00:00:46.000 --> 00:00:48.000]   it's your baby.
[00:00:48.000 --> 00:00:50.000]   [Gunshot]
[00:00:50.000 --> 00:00:53.000]   [Music]
[00:00:53.000 --> 00:00:55.000]   You looked dead, didn't I?
[00:00:55.000 --> 00:00:57.000]   Well, I wasn't.
[00:00:57.000 --> 00:01:00.000]   But it wasn't from lack of trying, I can tell you that.
[00:01:00.000 --> 00:01:03.000]   Actually, Bill's last bullet put me in a coma.
[00:01:03.000 --> 00:01:07.000]   A coma I was to lie in for four years.
[00:01:07.000 --> 00:01:09.000]   When I woke up,
[00:01:09.000 --> 00:01:15.000]   I went on what the movie advertisements refer to as a "roaring rampage of revenge."
[00:01:15.000 --> 00:01:18.000]   I roared, and I rampaged,
[00:01:18.000 --> 00:01:22.000]   and I got bloody satisfaction.
[00:01:22.000 --> 00:01:26.000]   I've killed a hell of a lot of people to get to this point.
[00:01:26.000 --> 00:01:29.000]   But I have only one more.
[00:01:29.000 --> 00:01:31.000]   The last one.
[00:01:31.000 --> 00:01:35.000]   The one I'm driving to right now.
[00:01:35.000 --> 00:01:38.000]   The only one left.
[00:01:38.000 --> 00:01:42.000]   And when I arrive at my destination,
[00:01:42.000 --> 00:01:45.000]   I am gonna kill Bill.
[00:01:45.000 --> 00:02:10.000]   [Music]
[00:02:10.000 --> 00:02:17.000]   Now, the incident that happened at the Two Pines wedding chapel that put this whole gory story into motion
[00:02:17.000 --> 00:02:20.000]   has since become legend.
[00:02:20.000 --> 00:02:22.000]   "Massacre at Two Pines."
[00:02:22.000 --> 00:02:24.000]   That's what the newspapers called it.
[00:02:24.000 --> 00:02:30.000]   The local TV news called it the "El Paso, Texas Wedding Chapel Massacre."
[00:02:30.000 --> 00:02:32.000]   How it happened.
[00:02:32.000 --> 00:02:33.000]   Who was there.
[00:02:33.000 --> 00:02:36.000]   How many got killed and who killed them.
[00:02:36.000 --> 00:02:40.000]   Changes depending on who's telling the story.
[00:02:40.000 --> 00:02:45.000]   In actual fact, the massacre didn't happen during a wedding at all.
[00:02:45.000 --> 00:02:48.000]   It was a wedding rehearsal.
[00:02:48.000 --> 00:02:55.000]   Now, when we come to the part where I say, "You may kiss the bride, you may kiss the bride,
[00:02:55.000 --> 00:02:59.000]   but don't stick your tongue in her mouth."
[00:02:59.000 --> 00:03:06.000]   This might be funny to your friends, but it would be embarrassing to your parents.
[00:03:06.000 --> 00:03:11.000]   We'll try to restrain ourselves from it.
[00:03:11.000 --> 00:03:16.000]   Y'all got a song?
[00:03:16.000 --> 00:03:20.000]   How about "Love Me Tender." I can play that.
[00:03:20.000 --> 00:03:22.000]   Sure.
[00:03:22.000 --> 00:03:24.000]   "Love Me Tender" would be great.
[00:03:24.000 --> 00:03:27.000]   Rufus, he's the man.
[00:03:27.000 --> 00:03:30.000]   Rufus, who was that you used to play for?
[00:03:30.000 --> 00:03:32.000]   Rufus Thomas.
[00:03:32.000 --> 00:03:35.000]   Rufus Thomas. Rufus Thomas.
[00:03:35.000 --> 00:03:43.000]   I was a drill, I was a drifter, I was a coaster, I was part of the gang, I was a bar-quet.
[00:03:43.000 --> 00:03:47.000]   If they come through Texas, I done played with them.
[00:03:47.000 --> 00:03:51.000]   Rufus, he's the man.
[00:03:54.000 --> 00:03:57.000]   Have you ever forgotten anything?
[00:03:57.000 --> 00:04:00.000]   Oh yes, you forgot the seating arrangements.
[00:04:00.000 --> 00:04:03.000]   Thank you, Mother.
[00:04:03.000 --> 00:04:10.000]   Now, the way we normally do this, we have the bride's side and then we have the groom's side.
[00:04:10.000 --> 00:04:17.000]   But since the bride ain't got nobody coming, and the groom's got far too many people coming.
[00:04:17.000 --> 00:04:21.000]   Well yeah, they're coming all the way from Oklahoma.
[00:04:21.000 --> 00:04:30.000]   Right. Well I don't see no problem with the groom's side sharing the bride's side. Do you, Mother?
[00:04:30.000 --> 00:04:32.000]   No, I don't have a problem with that.
[00:04:32.000 --> 00:04:38.000]   But, honey, you know it would be good if you had somebody come.
[00:04:38.000 --> 00:04:42.000]   You know, is it a sign of good faith?
[00:04:42.000 --> 00:04:49.000]   Well, I don't have anybody. Except for Tommy and my friends.
[00:04:49.000 --> 00:04:52.000]   You have no family?
[00:04:52.000 --> 00:04:54.000]   Well, I'm working on changing that.
[00:04:54.000 --> 00:04:58.000]   Mrs. Harmony, we're all the family this little angel's ever gonna need.
[00:04:58.000 --> 00:05:04.000]   I'm not feeling very well, and this bitch is starting to piss me off.
[00:05:04.000 --> 00:05:08.000]   So while you all blather on, I'm gonna go outside and get some air.
[00:05:08.000 --> 00:05:10.000]   Um, uh, Reverend, sorry.
[00:05:10.000 --> 00:05:12.000]   She's gonna go out and get some air?
[00:05:12.000 --> 00:05:14.000]   Yeah, given her delicate condition.
[00:05:14.000 --> 00:05:19.000]   She just needs a few minutes to get it together. She'll be okay.
[00:05:20.000 --> 00:05:25.000]   [Music]
[00:05:26.000 --> 00:05:31.000]   [Music]
[00:05:31.000 --> 00:05:36.000]   [Music]
[00:05:36.000 --> 00:05:41.000]   [Music]
[00:05:41.000 --> 00:05:46.000]   [Music]
[00:05:46.000 --> 00:05:51.000]   [Music]
[00:05:51.000 --> 00:05:56.000]   [Music]
[00:05:56.000 --> 00:06:01.000]   [Music]
[00:06:01.000 --> 00:06:06.000]   [Music]
[00:06:06.000 --> 00:06:11.000]   [Music]
[00:06:11.000 --> 00:06:16.000]   [Music]
[00:06:16.000 --> 00:06:21.000]   [Music]
[00:06:21.000 --> 00:06:26.000]   [Music]
[00:06:26.000 --> 00:06:31.000]   [Music]
[00:06:31.000 --> 00:06:36.000]   [Music]
[00:06:36.000 --> 00:06:38.000]   Hello, kiddo.
[00:06:38.000 --> 00:06:45.000]   How did you find me?
[00:06:45.000 --> 00:06:48.000]   I'm the man.
[00:06:48.000 --> 00:06:53.000]   What are you doing here?
[00:06:55.000 --> 00:06:57.000]   What am I doing?
[00:06:57.000 --> 00:07:03.000]   Well, a moment ago I was playing my flute.
[00:07:03.000 --> 00:07:17.000]   This moment, I'm looking at the most beautiful bride these whole eyes have ever seen.
[00:07:17.000 --> 00:07:21.000]   Why are you here?
[00:07:21.000 --> 00:07:23.000]   Last look.
[00:07:24.000 --> 00:07:26.000]   Are you gonna be nice?
[00:07:26.000 --> 00:07:29.000]   I've never been nice my whole life.
[00:07:29.000 --> 00:07:34.000]   But I'll do my best to be sweet.
[00:07:34.000 --> 00:07:41.000]   I always told you, your sweet side is your best side.
[00:07:41.000 --> 00:07:46.000]   I guess that's why you're the only one who's ever seen it.
[00:07:46.000 --> 00:07:51.000]   See, you got a bun in the oven.
[00:07:52.000 --> 00:07:53.000]   Hmm.
[00:07:53.000 --> 00:07:56.000]   I'm knocked up.
[00:07:56.000 --> 00:07:59.000]   Jeez, Louise.
[00:07:59.000 --> 00:08:04.000]   That young man of yours sure doesn't believe in wasting time, does he?
[00:08:04.000 --> 00:08:07.000]   Have you seen Tommy?
[00:08:07.000 --> 00:08:11.000]   Big guy in the tux?
[00:08:11.000 --> 00:08:12.000]   Yes.
[00:08:12.000 --> 00:08:14.000]   Then I saw him.
[00:08:14.000 --> 00:08:18.000]   I like his hair.
[00:08:20.000 --> 00:08:22.000]   You promised you'd be nice.
[00:08:22.000 --> 00:08:28.000]   I said I'd do my best. That's hardly a promise.
[00:08:28.000 --> 00:08:30.000]   But you're right.
[00:08:30.000 --> 00:08:34.000]   What does your young man do for a living?
[00:08:34.000 --> 00:08:39.000]   He owns a used record store here in El Paso.
[00:08:39.000 --> 00:08:41.000]   Ah. Music lover, eh?
[00:08:41.000 --> 00:08:44.000]   He's fond of music.
[00:08:44.000 --> 00:08:47.000]   Aren't we all?
[00:08:48.000 --> 00:08:52.000]   And what are you doing for a J.O.B. these days?
[00:08:52.000 --> 00:08:55.000]   I work in the record store.
[00:08:55.000 --> 00:08:59.000]   Ah, so...
[00:08:59.000 --> 00:09:03.000]   it all suddenly seems so clear.
[00:09:03.000 --> 00:09:07.000]   Do you like it?
[00:09:07.000 --> 00:09:10.000]   Yeah, I like it a lot, smartass.
[00:09:10.000 --> 00:09:13.000]   I get to listen to music all day.
[00:09:14.000 --> 00:09:17.000]   Talk about music all day. It's really cool.
[00:09:17.000 --> 00:09:22.000]   It's gonna be a great environment for my little girl to grow up in.
[00:09:22.000 --> 00:09:30.000]   As opposed to jetting around the world, killing human beings,
[00:09:30.000 --> 00:09:33.000]   and being paid vast sums of money?
[00:09:33.000 --> 00:09:36.000]   Precisely.
[00:09:36.000 --> 00:09:39.000]   Well, my old friend,
[00:09:39.000 --> 00:09:41.000]   to each his own,
[00:09:42.000 --> 00:09:44.000]   to each his own.
[00:09:44.000 --> 00:09:49.000]   However, all cock-luckery aside,
[00:09:49.000 --> 00:09:52.000]   I am looking forward to meeting your young man.
[00:09:52.000 --> 00:09:56.000]   I happen to be more or less particular,
[00:09:56.000 --> 00:09:58.000]   whom my gout marries.
[00:09:58.000 --> 00:10:02.000]   You wanna come to the wedding?
[00:10:02.000 --> 00:10:05.000]   Only if I can sit on the bride's side.
[00:10:05.000 --> 00:10:09.000]   You'll find it a bit lonely on my side.
[00:10:09.000 --> 00:10:12.000]   Your side always was a bit lonely.
[00:10:12.000 --> 00:10:15.000]   But I wouldn't sit anywhere else.
[00:10:15.000 --> 00:10:19.000]   You know,
[00:10:19.000 --> 00:10:22.000]   I had the loveliest dream about you.
[00:10:22.000 --> 00:10:25.000]   Oh, here's Tommy. Call me Arlene.
[00:10:25.000 --> 00:10:27.000]   You must be Tommy.
[00:10:27.000 --> 00:10:29.000]   Arlene's told me so much about you.
[00:10:29.000 --> 00:10:31.000]   Arlene, you okay?
[00:10:31.000 --> 00:10:32.000]   Oh, I'm fine.
[00:10:32.000 --> 00:10:35.000]   Tommy, I'd like you to meet my father.
[00:10:35.000 --> 00:10:38.000]   Oh, my God!
[00:10:38.000 --> 00:10:40.000]   Oh, my God! This is great!
[00:10:40.000 --> 00:10:42.000]   I'm so glad to meet you, sir.
[00:10:42.000 --> 00:10:43.000]   Oh, Dad.
[00:10:43.000 --> 00:10:45.000]   The name's Bill.
[00:10:45.000 --> 00:10:47.000]   Well, it's great to meet you, Bill.
[00:10:47.000 --> 00:10:49.000]   Arlene told me you couldn't make it.
[00:10:49.000 --> 00:10:51.000]   Surprise.
[00:10:51.000 --> 00:10:52.000]   That's my pop for you.
[00:10:52.000 --> 00:10:54.000]   Always full of surprises.
[00:10:54.000 --> 00:10:57.000]   Well, in the surprise department,
[00:10:57.000 --> 00:11:00.000]   the apple doesn't fall far from the tree.
[00:11:00.000 --> 00:11:02.000]   When did you get in?
[00:11:02.000 --> 00:11:03.000]   Just now.
[00:11:03.000 --> 00:11:05.000]   Did you come straight from Australia?
[00:11:05.000 --> 00:11:07.000]   Of course.
[00:11:07.000 --> 00:11:09.000]   Daddy, I told Tommy that you were in Perth
[00:11:09.000 --> 00:11:12.000]   mining for silver, and no one could reach you.
[00:11:12.000 --> 00:11:16.000]   Lucky for us all, that's not the case.
[00:11:16.000 --> 00:11:20.000]   So, what's this all about?
[00:11:20.000 --> 00:11:22.000]   I've heard of wedding rehearsals,
[00:11:22.000 --> 00:11:24.000]   but I don't believe I've ever heard
[00:11:24.000 --> 00:11:27.000]   of a wedding dress rehearsal before.
[00:11:27.000 --> 00:11:28.000]   We thought,
[00:11:28.000 --> 00:11:29.000]   "Why pay so much money for a dress
[00:11:29.000 --> 00:11:31.000]   you're only gonna wear once?"
[00:11:31.000 --> 00:11:34.000]   Especially when Arlene looks so goddamn beautiful in it.
[00:11:34.000 --> 00:11:36.000]   So, uh, I think we're gonna try to get all the mileage
[00:11:36.000 --> 00:11:37.000]   we can out of it.
[00:11:37.000 --> 00:11:41.000]   Isn't it supposed to be bad luck
[00:11:41.000 --> 00:11:44.000]   for the groom to see the bride in her wedding dress
[00:11:44.000 --> 00:11:46.000]   before the ceremony?
[00:11:46.000 --> 00:11:50.000]   Well, I guess I just believe I live in danger, so...
[00:11:50.000 --> 00:11:54.000]   I know just what you mean.
[00:11:54.000 --> 00:11:57.000]   Son, some of us have places to be.
[00:11:57.000 --> 00:11:59.000]   It's your old dude.
[00:11:59.000 --> 00:12:01.000]   Look, we gotta go through this one more time.
[00:12:01.000 --> 00:12:03.000]   So, uh, why don't you have a s--
[00:12:03.000 --> 00:12:05.000]   Oh, my God.
[00:12:05.000 --> 00:12:07.000]   What am I thinking? You should give her away.
[00:12:07.000 --> 00:12:11.000]   Tommy, that's not exactly Daddy's cup of tea.
[00:12:11.000 --> 00:12:14.000]   I think Father would be much more comfortable
[00:12:14.000 --> 00:12:16.000]   sitting with the rest of the guests.
[00:12:16.000 --> 00:12:17.000]   Really?
[00:12:17.000 --> 00:12:19.000]   That's asking a lot.
[00:12:19.000 --> 00:12:21.000]   Oh.
[00:12:21.000 --> 00:12:23.000]   Okay. Well, forget it.
[00:12:23.000 --> 00:12:26.000]   But how about we go out to dinner tonight and celebrate?
[00:12:26.000 --> 00:12:29.000]   Only on the condition that I pay for everything.
[00:12:29.000 --> 00:12:31.000]   Deal. We gotta do this now.
[00:12:31.000 --> 00:12:33.000]   Can I watch?
[00:12:33.000 --> 00:12:35.000]   Absolutely. Have a seat.
[00:12:35.000 --> 00:12:37.000]   Which is the bride's side?
[00:12:37.000 --> 00:12:39.000]   Right over here.
[00:12:39.000 --> 00:12:44.000]   Mother, here we go.
[00:12:44.000 --> 00:12:49.000]   Now, son, about them vows.
[00:12:49.000 --> 00:12:57.000]   No.
[00:12:57.000 --> 00:12:59.000]   I just don't want...
[00:12:59.000 --> 00:13:02.000]   You don't owe me a damn thing.
[00:13:03.000 --> 00:13:05.000]   If he's the man you want,
[00:13:05.000 --> 00:13:08.000]   then go stand by.
[00:13:08.000 --> 00:13:10.000]   [chuckles]
[00:13:10.000 --> 00:13:34.000]   Do I look pretty?
[00:13:34.000 --> 00:13:36.000]   Oh, yeah.
[00:13:36.000 --> 00:13:39.000]   [♪♪♪]
[00:13:39.000 --> 00:13:48.000]   Thank you.
[00:13:48.000 --> 00:13:51.000]   [♪♪♪]
[00:13:51.000 --> 00:13:54.000]   [♪♪♪]
[00:13:54.000 --> 00:13:57.000]   [♪♪♪]
[00:13:57.000 --> 00:13:59.920]   [MUSIC]

Comparison of results

tiny 160 s	base 433 s	small 1713 s	medium 3563 s
Do you finally sit just now?	Do you find me sadistic?	Do you find me sadistic?	Do you find me sadistic?
[MUSIC PLAYING]	No, Kato.	No, kiddo.	No, kiddo.
No, I can’t do.	I’d like to believe	I’d like to believe you’re aware enough, even now,	I’d like to believe
I’d like to believe you’re aware enough even now.	you’re aware enough, even now.	to know that there is nothing sadistic in my actions.	you’re aware enough, even now,
No, that there is nothing suggesting in my actions.	I know that there is nothing sadistic in my actions.	This moment, this is me and my most nicer kiss to be.	to know that there’s nothing sadistic
This moment, this is me and my most nice against them.	This moment,	Well, it’s your baby.	in my actions.
Well, it’s your name.	this is me,	Look, Dad, didn’t I?	This moment,
[MUSIC PLAYING]	and I most miss the case.	Well, I wasn’t. But it wasn’t from lack of trying, I can tell you that.	this is me
But Dad didn’t I?	Well,	Actually, Bill’s last bullet put me in a coma.	and my most masochistic.
Well, I wasn’t.	it’s your baby.	A coma I was to lie in for four years.	Well,
But it wasn’t from lack of trying.	The dead didn’t I?	When I woke up, I went on with the movie advertisements	it’s your baby.
I can tell you that.	Well, I wasn’t.	referred to as a roaring rampage of revenge.	[Gunshot]
Actually, Bill’s last bullet put me in a coma.	But it wasn’t from lack of try and I can tell you that.	I roared, and I rampaged, and I got bloody satisfaction.	[Music]
A coma, I was to lie in for four years.	Actually, Bill’s last bullet put me in a coma.	I’ve killed a hell of a lot of people to get to this point.	You looked dead, didn’t I?
And I woke up.	A coma I was to lie in for four years.	But I have only one more.	Well, I wasn’t.
I went on with a movie advertisements	When I woke up,	The last one.	But it wasn’t from lack of trying, I can tell you that.
for two as a roaring rampage of revenge.	I went on with the movie advertisements referred to as a roaring rampage of	The one I’m driving to right now.	Actually, Bill’s last bullet put me in a coma.
I roared.	I roared,	The only one left.	A coma I was to lie in for four years.
And I relanged.	and I rampaged,	And when I arrive at my destination, I am gonna kill Bill.	When I woke up,
And I got bloody satisfaction.	and I got bloody satisfaction.		I went on what the movie advertisements refer to as a “roaring rampage of revenege.”
I’ve killed a hell of a lot of people to get to this point.	I’ve killed a hell of a lot of people who get to this point.		I roared, and I rampaged,
But I have only one more.	But I have only one more.		and I got bloody satisfaction.
The last one.	The last one.		I’ve killed a hell of a lot of people to get to this point.
The one I’m driving to right now.	The one I’m driving to right now.		But I have only one more.
The only one left.	The only one left.		The last one.
And when I arrive at my destination,	And when I arrive at my destination,		The one I’m driving to right now.
I have going to kill Bill.	I am gonna kill Bill.		The only one left.
			And when I arrive at my destination,
			I am gonna kill Bill.

Conclusion

If you take a peek at the above results (and you remember the movie or download an .srt of the actual subtitles) you’ll see that the results for the small and medium model were almost perfect! The base model was also good enough considering that it took much less time than these two.

The tiny model wasn’t so good however even that model is good enough to understand everything that is being said in the movie from the subtitles and context. Finally, consider that both the tiny and base models were faster than real-time even on my (very slow) computer.

HTML form disable after submit

Fri 12 May 2023

One of the most common problems I get in my apps is double submissions of forms. A lot of users can’t understand the difference between single and double click and end up double clicking the form submit button. Also, if the form takes too long to submit they might thing that they didn’t press the button correctly and click it again.

This, depending on how your app is built could result in either working perfectly, or showing errors to users or (which is the worst) duplicate entries in your database.

There is a very simple fix for that: Disable the submit button after the first click. Here’s how to do it with jQuery:

$(document).ready(function () {
    $('form').submit(function () {
        let submit = $(this).find(':input[type=submit]')
        submit.prop('disabled', true);
        if(submit.val()) {
            submit.val(submit.val() + ' ⌛' )
        } else {
            submit.html(submit.html() + ' ⌛')
        }
    })
});

and with vanilla.js if you don’t use jquery

document.addEventListener('DOMContentLoaded', function () {
    document.querySelectorAll('form').forEach(function (form) {
        form.addEventListener('submit', function (event) {
            let submit = this.querySelector('input[type="submit"], button[type="submit"]');
            submit.disabled = true;
            if(submit.value) {
                submit.value = submit.value + ' ⌛';
            } else {
                submit.innerHTML = submit.innerHTML + ' ⌛';
            }
        });
    });
});

The above will find all forms and add an event listener to the submit event. When the form is submitted it will find the submit button and disable it. Finally, it adds a unicode hourglass character (⌛) to the displayed button text so the user gets a quick feedback that the form is being submitted.

The above snippets should work correctly no matter if you use an <input type="submit"> or a <button type="submit"> element (that’s why we use :input in jquery to capture both types of elements or we do the double check on the querySelector, also notice that it checks if the element has a val()/value and sets sets val()/value or html()/innerHTML accordingly).

I use the above snippet on every project I work on and it has saved me a lot of headaches. Please be advised that if you do funny JS things with your form this snippet may not work and break its functionality, but in this case you are probably handling the form disabling yourself.

Multiple storages for the same FileField in Django

Tue 11 April 2023

When you need to support user-uploaded files from Django (usually called media) you will probably use a FileField in your models. This is translated to a simple varchar (text) field in the database that contains a unique identifier for the file. This usually would be the path to the file, however this is not always the case!

What really happens is that Django has an underlying concept called a File Storage which is a class that has information on how to talk to the actual storage backend, and particularly how to translate the unique identifier stored on the db to an actual file object. By default Django stores files in the file system using the FileSystemStorage however it is possible to use different backends through an add-on (for example Amazon S3) or even write your own.

Each FileField can be configured to use a different storage backend by passing the storage parameter; if you don’t use this parameter then the default storage backend is used. So you can easily configure a FileField that would upload files to your filesystem and another one that would upload files to S3.

However, one thing that is not supported though is to use multiple storages for the same FileField depending on some parameter of the model instance. Unfortunately, in a recent project I had to do exactly that: We had a FileField on a model that contained hundreds of GBs of files stored on the filesystem; we wanted to be able to upload the files of new instances of that model on S3 but also wanted to keep the old files on the filesystem to avoid moving all these to S3 (which would result to a lot of downtime). I also wanted a way to be “flexible” on this i.e to be able to change again the storage backend for some instances if needed and definitely not move/copy all these files!

If you take a peek at the FileField options you’ll see that there’s a storage parameter that can be a callable. However this callable is initialized with the models and is not evaluated again until the app is restarted so it can’t be used to decide on the storage for each model instance.

The only thing that is evaluated each time a file is uploaded through the FileField is when upload_to is a function. This function receives the model instance and returns the path that the file will be uploaded to.

The idea is to use this upload_to function to return a different path depending on the model instance and then use a custom storage backend that will use the path to decide on the actual storage backend to use.

This is the code I ended up with for the upload_to function:

def file_upload_path(instance, filename):
    dt_str = app.created_on.strftime("%Y/%m/%d")
    file_storage = ""

    if instance.id >= settings.STORAGE_CHANGE_ID: 
        file_storage = settings.STORAGE_SELECTION_STR + "/"

    return "protected/{0}{1}/{2}/{3}".format(file_storage, dt_str, instance.id, filename)

class Model(models.Model):
    file = models.FileField(upload_to=file_upload_path)

What happens here is that I have a setting STORAGE_CHANGE_ID that is the id of the instance after which all instances will use the different storage backend. You can use whatever method you want here to decide on the storage that would be used; the only thing to keep in mind is to put the storage somewhere on the returned path.

I also have a setting STORAGE_SELECTION_STR that is the string that will be used in the path to differentiate the storage backend. The STORAGE_SELECTION_STR has the value of minios3 for this project.

Using this function the paths of the instances that are >= STORAGE_CHANGE_ID will be of the form protected/minios3/2021/04/11/1234/filename.ext while for the old files these will be of the form protected/2021/04/11/1234/filename.ext. Notice the minios3 string in between.

Of course this is not enough. We also need to tell Django to use the different storage backend for the new files. In order to do this we have to implement a custom storage class like this:

from django.core.files.storage import FileSystemStorage, Storage
from storages.backends.s3boto3 import S3Boto3Storage
from django.conf import settings


class FilenameBasedStorage(Storage):
    minio_choice = settings.STORAGE_SELECTION_STR

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)

    def _open(self, name, mode="rb"):
        if self.minio_choice in name:
            return S3Boto3Storage().open(name, mode)
        else:
            return FileSystemStorage().open(name, mode)

    def _save(self, name, content):
        if self.minio_choice in name:
            return S3Boto3Storage().save(name, content)
        else:
            return FileSystemStorage().save(name, content)

    def delete(self, name):
        if self.minio_choice in name:
            return S3Boto3Storage().delete(name)
        else:
            return FileSystemStorage().delete(name)

    def exists(self, name):
        if self.minio_choice in name:
            return S3Boto3Storage().exists(name)
        else:
            return FileSystemStorage().exists(name)

    def size(self, name):
        if self.minio_choice in name:
            return S3Boto3Storage().size(name)
        else:
            return FileSystemStorage().size(name)

    def url(self, name):
        if self.minio_choice in name:
            return S3Boto3Storage().url(name)
        else:
            return FileSystemStorage().url(name)

    def path(self, name):
        if self.minio_choice in name:
            return S3Boto3Storage().path(name)
        else:
            return FileSystemStorage().path(name)

This class should be self explainable: It uses the settings.STORAGE_SELECTION_STR we mentioned above to decide which storage backend to use and then it forwards each method to the corresponding backend (either the filesystem storage or the S3 storage).

One thing to notice is that the django.core.files.storage.Storage class this class inherits from has more methods that can be implemented (and would raise if called without implementing them) however this implementation works fine for my needs.

One question some readers may have is what happens if the user uploads a file named test-minios3.pdf (i.e a file containing the STORAGE_SELECTION_STR). Well you may just as well ignore it; it will just be saved always on the minio-s3 storage backend. Or you can make sure to remove that string from the filename before saving it on the file_upload_path. I chose to ignore it since it doesn’t matter for my use case.

Finally, we need to tell Django to use this storage class for the file field. We can do this by adding it to the FileField like:

file = models.FileField(upload_to=file_upload_path, storage=FilenameBasedStorage)

or we can configure it on the DEFAULT_FILE_STORAGE setting (for Django < 4.2) or on the STORAGES dict (for Django >= 4.2).

I hope this helps someone else that needs to do something similar!

← Older

/var/

Various programming stuff

A simple OpenID connect tutorial

Introduction

Authentication flow

Authentication server

Automatic Discovery

The Authentication url

User Authentication and redirect

Getting the token

Decoding a JWT (JSON Web Token)

Retrieving the public key

Verifying the JWT token

Calling the userinfo endpoint

What about the authorization server?

Conclusion

Simple Django - DataTables integration

AI auto-subtitling

Introduction

Whisper.cpp

Extract audio from the movie

Transcribing the audio

Results

ggml-tiny.en.bin

ggml-base.en.bin

ggml-small.en.bin

ggml-medium.en.bin

Comparison of results

Conclusion

HTML form disable after submit

Multiple storages for the same FileField in Django