Transcription Project

[markhavemann]

added:

Light The Night Season 2 (华灯初上)

 

[markhavemann]

added:

Light The Night Season 3 (华灯初上)

[markhavemann]

Added 预支未来 S1 (Futmalls)

 

This show was kind of unique with its Black Mirror sci-fi style, hopefully there will be a season 2.

[JinWenSen]

On 12/25/2018 at 11:57 PM, markhavemann said:

Financial Contributions

If anybody wants to contribute financially to this and have stuff transcribed professionally at around 20rmb/3USD for a 30 minute episode, you can add me on WeChat or send money via PayPal with the info below: 

 

Hey thanks for doing this, it's awesome ! I might want to contribute, I tried looking for you on WeChat but couldn't find you... Anyway, I was looking for the 男人帮 transcript but all I could find was the show with hard subtitles..

[markhavemann]

On 9/21/2022 at 3:19 AM, JinWenSen said:

Hey thanks for doing this, it's awesome ! I might want to contribute, I tried looking for you on WeChat but couldn't find you... Anyway, I was looking for the 男人帮 transcript but all I could find was the show with hard subtitles..

I changed my wechat ID and forgot to update it here. It's now: justanothermark

[JinWenSen]

On 9/20/2022 at 7:29 PM, markhavemann said:

changed my wechat ID and forgot to update it here. It's now:

Sent you a friend invite

[markhavemann]

For something a little different, I've added all subtitles for the Chinese dub of Netflix version of My Hero Academia S1-4. The subtitles match about 95% of the time and it's a nice way to get some Chinese while watching something that isn't a Chinese show. 

 

I'm thinking about uploading the actual mp3 audio tracks of each episode too, since I think the Chinese dubs are only available if you watch on Netflix in Taiwan or Hong Kong. Not sure about the legality of this though?

[MTH123]

On 10/9/2022 at 6:01 AM, markhavemann said:

Not sure about the legality of this though?

 

I'm not sure either. There’s at least one long thread about this kind of thing on this website. It’s a great, interesting read with a lot of different, informative opinions. For me, the sort of bottom line is that it isn’t for profit. It’s for learning. There aren’t that many people trying to learn this way either. So, there isn’t much for any company to get too excited about in a legal way. Having said that, I still prefer to point people to websites where they can download stuff themselves.

[markhavemann]

On 10/12/2022 at 5:55 AM, MTH123 said:

For me, the sort of bottom line is that it isn’t for profit. It’s for learning.

That's what I tend to think too.

 

On 10/12/2022 at 5:55 AM, MTH123 said:

I still prefer to point people to websites where they can download stuff themselves.

Unfortunately, lots of the websites that were around before have disappeared. If you are lucky enough to find a website to get stuff like this, it usually doesn't last for very long, and all that's left is broken links and useful content that's lost forever. 

[MTH123]

On 10/11/2022 at 8:15 PM, markhavemann said:

Unfortunately, lots of the websites that were around before have disappeared. If you are lucky enough to find a website to get stuff like this, it usually doesn't last for very long, and all that's left is broken links and useful content that's lost forever. 

 

Yes, great point. Even websites that are still around have broken links.

[markhavemann]

Added 

 

Chinese Name	English Name	Year	Episodes	Complete	Subs	English Subs	Text Transcripts	Links
延禧攻略	Story of Yanxi Palace		70	✔️	✔️		✔️

[大块头]

I wonder how much it would cost per hour to transcribe podcasts? Presumably a lot more than $6/hour... I know I'd get a lot more out of Gushi FM if I could skim through a transcript and then listen to each episode again.

 

In my experience hiring Chinese freelancers, $15/hour is typical for unspecialized labor. That works out to maybe $30 for transcribing an hour of audio; too much for me to support personally but feasible with a Patreon if there's enough interest.

[pon00050]

On 2/1/2023 at 5:17 PM, 大块头 said:

I wonder how much it would cost per hour to transcribe podcasts? Presumably a lot more than $6/hour... I know I'd get a lot more out of Gushi FM if I could skim through a transcript and then listen to each episode again.

 

In my experience hiring Chinese freelancers, $15/hour is typical for unspecialized labor. That works out to maybe $30 for transcribing an hour of audio; too much for me to support personally but feasible with a Patreon if there's enough interest.

 

In my experience, transcribing an hour of audio usually takes more than an hour of real time.

 

I'd like to suggest an alternative.

Here is the method that I am currently using to get transcripts.

This method does not produce perfectly accurate transcripts.

If you could find another willing learner to allocate the amount of work to make one correct transcript,

I think that should also be feasible and get you results you want faster.

 

[大块头]

On 2/1/2023 at 6:19 PM, pon00050 said:

Here is the method that I am currently using to get transcripts.

This method does not produce perfectly accurate transcripts.

 

I think that people producing transcriptions probably use AI tools like that as a first step and then correct errors manually? A quick search didn't yield any companies offering human transcription (人工转录), but I found several freelancers on Upwork charging ~$15 (per hour of labor, not per hour of audio) for this service.

 

Edit: found a transcription service, but it's expensive

[pon00050]

On 2/1/2023 at 7:32 PM, 大块头 said:

I think that people producing transcriptions probably use AI tools like that as a first step and then correct errors manually?

Did you read the writing that I linked?

While it's not using specifically "AI", it is incorporating speech to text technology to get some of the transcripts automatically generated and human(myself) is doing the rest to make the transcription complete and accurate. So, I think it's still the idea of using technology as a first step and correcting errors manually.

[markhavemann]

https://www.iflyrec.com/ has got human transcription for 1.34rmb/min of audio, their AI is even cheaper and pretty good too.

 

If you get it done with AI you could post on the forum and get people to help check as a joint learning activity, I'd probably be interested in helping if the content was interesting.

[大块头]

On 2/3/2023 at 2:38 AM, markhavemann said:

https://www.iflyrec.com/ has got human transcription for 1.34rmb/min of audio

 

Wow, that's pretty inexpensive. I assume that at that cost the names of people and things will probably not be 100% accurate, but I don't think that's an issue for our purposes. 

 

From experience, I know I'll be more likely to fit this listening practice in among my other projects if it doesn't require a lot of time sitting in front of the computer correcting transcripts, running scripts, etc. I'll look into getting a small Patreon started in a few months when I have more time.

[markhavemann]

On 2/4/2023 at 1:55 AM, 大块头 said:

Wow, that's pretty inexpensive. I assume that at that cost the names of people and things will probably not be 100% accurate, but I don't think that's an issue for our purposes. 

This is a big company, I don't see why there would be mistakes. It's very likely more or less on par with any other transcription services.  

 

As for names, I'm not sure anyone could guarantee accuracy at any price, unless someone actually mentions the exact characters in their name. This is the same for English, if you are transcribing and all you have to go on is sound, would you write "Ashleigh", "Ashley", "Ashlee"? No matter what your fee was, there's no way to know unless it's a famous person that you could Google, or the people in the audio say which one it is. 

 

I guess you might get placeholders with pinyin for some names that can't easily be guessed. I might try get something transcribed at some stage soon and post the results.  

[markhavemann]

Added:

 

Chinese Name	English Name	Year	Episodes	Complete	Subs	English Subs	Text Transcripts	Links
狂飙	The Knockout	2023	39	✔️	✔️	✔️	✔️	iqiyi

 

Really enjoying this one so far, would definitely recommend. 

[markhavemann]

I've created a thread for Anki vocabulary decks to help watch 家有儿女, as well as general vocabulary stats for the series. 

 

The thread can be found here. I've also added a place for links in the main post of this thread as I add for more shows.