Tool Bigger thread on audio works and their translation

[Ezekyle Abaddon]

If anyone wants to transcript a ASMR that doesn't provide .vtt or .srt files with it and wants a more efficient/inexpensive method than using whisper try instead using an LMM ai to do all the work for you, this might seem obvious but it doesn't hurt to tell the people who didn't think to try before

currently the one that I'm using is gemeni 2.5 pro via Google's ai studio to transcript and translate the audio for me, but there might come more advanced ones that does it better in the future, also the ai must have audio input as a feature otherwise it can't process audio files


this method personally saved me tons of time as I had to find free alternatives to whisper (as most paying methods are yet to be available with my country)

Interesting. I just try Gemeni 2.5 a few days ago but because I'm not good with AI stuff, I just goofing around with mathematics function on the app. I did not know there is a transcript and translate function for the AI.

 

[obzoruch]

If anyone wants to transcript a ASMR that doesn't provide .vtt or .srt files with it and wants a more efficient/inexpensive method than using whisper try instead using an LMM ai to do all the work for you, this might seem obvious but it doesn't hurt to tell the people who didn't think to try before

currently the one that I'm using is gemeni 2.5 pro via Google's ai studio to transcript and translate the audio for me, but there might come more advanced ones that does it better in the future, also the ai must have audio input as a feature otherwise it can't process audio files


this method personally saved me tons of time as I had to find free alternatives to whisper (as most paying methods are yet to be available with my country)

Whisper can be used for free by downloading to your device, well or using google colab, which is what I do

 

[bilboswaggings]

How trustworthy is asmr.one as far as file safety is concerned? Downloaded a pdf file by accident with the work's transcription, analyzed it then deleted. It's probably clean but I'm a Schizo.

 

[Jimmyodi841]

If anyone wants to transcript a ASMR that doesn't provide .vtt or .srt files with it and wants a more efficient/inexpensive method than using whisper try instead using an LMM ai to do all the work for you, this might seem obvious but it doesn't hurt to tell the people who didn't think to try before

currently the one that I'm using is gemeni 2.5 pro via Google's ai studio to transcript and translate the audio for me, but there might come more advanced ones that does it better in the future, also the ai must have audio input as a feature otherwise it can't process audio files


this method personally saved me tons of time as I had to find free alternatives to whisper (as most paying methods are yet to be available with my country)

I've already tried it and it works great!
It's really amazing!

 

[bilboswaggings]

I've already tried it and it works great!
It's really amazing!

Sorry to be a bother but is there a specific way to set up this "Google Ai studio"? Or is it just a plug and play kind of thing.

 

[Jimmyodi841]

Sorry to be a bother but is there a specific way to set up this "Google Ai studio"? Or is it just a plug and play kind of thing.

I don't know much about this.
I just use the regular chat, just make sure it's with Gemini 2.5 Pro.

 

[dontseemeeee]

So how do I use these though? The gumroad has the translated scripts in the forms of pdfs and docx but no vtt or subtitle file, am i supposed to just find those myself?

 

[ccgssdtttew]

Wow! It took me 2 days of searching to finally find a post on converting ASMR to english. I aint no genius but i think i shall try if the steps are simple enough.

Anyway, great that you compiled the websites used for ASMR. I found one more to add that I feel seems good.

You must be registered to see the links

You must be registered to see the links

Re-pasting these as mentioned in the original post:

You must be registered to see the links

You must be registered to see the links

 

[Ezekyle Abaddon]

I don't know much about this.
I just use the regular chat, just make sure it's with Gemini 2.5 Pro.
View attachment 4932321

Google Gemini is damn good but half of the times, it stop working after detect 'harmful' content and whatnot.

 

[Jimmyodi841]

Google Gemini is damn good but half of the times, it stop working after detect 'harmful' content and whatnot.

What I do is try different prompts like in ChatGPT, making it clear that they are completely fictional, hypothetical situations or things like that.
I haven’t had any problems with it.
I’ve been able to translate all the lines from several voice works.

 

[Ezekyle Abaddon]

What I do is try different prompts like in ChatGPT, making it clear that they are completely fictional, hypothetical situations or things like that.
I haven’t had any problems with it.
I’ve been able to translate all the lines from several voice works.

Make sense. I'm an idiot so I usually just said it translated this or that and forget to input other command/ conditions. Thanks for your help.

 

[Jimmyodi841]

Make sense. I'm an idiot so I usually just said it translated this or that and forget to input other command/ conditions. Thanks for your help.

I usually use something like this.
This is just an example, it might not always work.

I need the most accurate possible transcription of this audio.
Transcribe all the dialogue without omitting anything that's said.
A transcription as exact as possible, keeping the original tone.
First, transcribe the Japanese audio, and then also translate it into English.

 

[mtl poet]

I've been waiting for this:

You must be registered to see the links

Unfortunately, I haven't had much luck. The audio filter and low volume makes the good parts almost unintelligible. My only hope is that the chinese take it and write a script for it

 

[Magicshot]

Ok, so i moved away from whisper to gemini 2.5, when it works its straight up perfect, srt with both japanese and translation with perfect timing and no erros. BUT in most files, even when attempting jailbreak, its going to hit the harmful content detected.

 

[Ezekyle Abaddon]

Ok, so i moved away from whisper to gemini 2.5, when it works its straight up perfect, srt with both japanese and translation with perfect timing and no erros. BUT in most files, even when attempting jailbreak, its going to hit the harmful content detected.

I got the same issue. In my experience, translate text is much easier than audio so I try to look for the Chinese version of the script and just go it from there. Usually, Chinese script have timestamp too and even more proper than the original version most of the times.