Autosub vtt file contains duplicated subtilte in each frame

[ali-masoudi]

Hey guys every autosub i download has duplicate string in multiple frame!

youtube-dl -v --sub-lang en --write-auto-sub --skip-download https://www.youtube.com/watch?v=v5cngxo4mIg
[debug] System config: []
[debug] User config: []
[debug] Custom config: []
[debug] Command-line args: ['-v', '--sub-lang', 'en', '--write-auto-sub', '--skip-download', 'https://www.youtube.com/watch?v=v5cngxo4mIg']
[debug] Encodings: locale UTF-8, fs utf-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2019.06.21
[debug] Python version 3.7.3 (CPython) - Linux-5.1.14-arch1-1-ARCH-x86_64-with-arch
[debug] exe versions: ffmpeg 4.1.3, ffprobe 4.1.3, rtmpdump 2.4
[debug] Proxy map: {}
[youtube] v5cngxo4mIg: Downloading webpage
[youtube] v5cngxo4mIg: Downloading video info webpage
[youtube] v5cngxo4mIg: Looking for automatic captions
[debug] Default format spec: bestvideo+bestaudio/best
[info] Writing video subtitles to: PyTorch Prerequisites - Syllabus for Neural Network Programming Course-v5cngxo4mIg.en.vtt

resulting subtitle:

WEBVTT
Kind: captions
Language: en

00:00:00.469 --> 00:00:03.169 align:start position:0%
 
welcome<00:00:01.469><c> to</c><00:00:01.620><c> the</c><00:00:01.740><c> series</c><00:00:02.159><c> on</c><00:00:02.460><c> neural</c><00:00:02.909><c> network</c>

00:00:03.169 --> 00:00:03.179 align:start position:0%
welcome to the series on neural network
 

00:00:03.179 --> 00:00:05.510 align:start position:0%
welcome to the series on neural network
programming<00:00:03.689><c> with</c><00:00:04.110><c> PI</c><00:00:04.440><c> torch</c><00:00:04.740><c> in</c><00:00:05.040><c> this</c><00:00:05.250><c> video</c>

00:00:05.510 --> 00:00:05.520 align:start position:0%
programming with PI torch in this video
 

00:00:05.520 --> 00:00:07.369 align:start position:0%
programming with PI torch in this video
we<00:00:05.970><c> will</c><00:00:06.089><c> look</c><00:00:06.270><c> at</c><00:00:06.359><c> the</c><00:00:06.480><c> prerequisites</c><00:00:07.319><c> needed</c>

00:00:07.369 --> 00:00:07.379 align:start position:0%
we will look at the prerequisites needed
 

00:00:07.379 --> 00:00:09.680 align:start position:0%
we will look at the prerequisites needed
to<00:00:07.859><c> be</c><00:00:07.980><c> best</c><00:00:08.340><c> prepared</c><00:00:08.429><c> for</c><00:00:08.910><c> the</c><00:00:09.000><c> series</c><00:00:09.360><c> we'll</c>

00:00:09.680 --> 00:00:09.690 align:start position:0%
to be best prepared for the series we'll
 

00:00:09.690 --> 00:00:11.330 align:start position:0%
to be best prepared for the series we'll
get<00:00:09.809><c> an</c><00:00:09.900><c> overview</c><00:00:10.380><c> of</c><00:00:10.559><c> the</c><00:00:10.679><c> series</c><00:00:11.010><c> and</c><00:00:11.219><c> a</c>

00:00:11.330 --> 00:00:11.340 align:start position:0%
get an overview of the series and a
 

00:00:11.340 --> 00:00:12.980 align:start position:0%
get an overview of the series and a
sneak<00:00:11.610><c> peek</c><00:00:11.849><c> at</c><00:00:11.969><c> a</c><00:00:12.000><c> project</c><00:00:12.450><c> we'll</c><00:00:12.599><c> be</c><00:00:12.690><c> working</c>

00:00:12.980 --> 00:00:12.990 align:start position:0%
sneak peek at a project we'll be working
 

00:00:12.990 --> 00:00:14.900 align:start position:0%
sneak peek at a project we'll be working
on<00:00:13.080><c> this</c><00:00:13.440><c> will</c><00:00:13.620><c> give</c><00:00:13.650><c> us</c><00:00:13.830><c> a</c><00:00:13.950><c> good</c><00:00:14.219><c> idea</c><00:00:14.280><c> about</c>

00:00:14.900 --> 00:00:14.910 align:start position:0%
on this will give us a good idea about
 

00:00:14.910 --> 00:00:16.700 align:start position:0%
on this will give us a good idea about
what<00:00:15.150><c> we'll</c><00:00:15.299><c> be</c><00:00:15.420><c> learning</c><00:00:15.599><c> and</c><00:00:15.990><c> what</c><00:00:16.260><c> skills</c>

00:00:16.700 --> 00:00:16.710 align:start position:0%
what we'll be learning and what skills

notice duplicated sentence in multiple frame!

[ali-masoudi]

#16032