Whisper Mate 4+

A Clear Labor of Love!

Whisper Mate is an incredibly advanced app for transcription automation, correction and translation. It has one of the strongest feature sets I’ve ever seen in any app. There’s a bit of a learning curve, but the developer has designed Whisper Mate to be sufficiently intuitive to figure things out, with just a little trial and error. Especially for those of us who need to correct transcriptions, the app is indispensable, with integrated access for automated segmentation by speaker(AKA diarization), regex find-replace, and plenty of other bells and whistles to make a painstaking task significantly less painful.

The best SRT app. Better than Mac Whisper!

The developer is listening to feedback, thank you!

This developer has made improvements and now has a balanced setting that improves subtitles although a bit slow.

Also the addition of Deelgram is wonderful, their Whisper translations are really the most accurate if you choose Whisper Large. Also it does not break senteces in half etc. I wish Offiline Whisper was more like Online DeepGram using Whisper.

Developer Response ,

Thanks you for you detail message. Can you send me the wrong media file. Whisper has difference model and difference tuned model, and some call arguments. Difference app
may call diff arguments. Not one argument will work fit all the media file. It need to self tune the params. Can't you send the test media file. So i can test which params will give you the wanted subtitles as you see in other apps.

I understand that you want Whisper Mate to improve and that is why you wrote such a detailed comment. Thank you for your suggestion. I will see how to make the configuration model parameters easier to understand by providing a single option

---
Can you send me the test media file to test? Whisper Mate provide more arguments to config to transcribe the media. I can test the media why is slow and use what arguments to make it faster. You can change the AI Model arguments in the Model Config Panel right corner's Extra button.

iniitializaing model took forever

I didn't use the app for a while and came back to open and run it.

But "initializing model" took forever and doesn't function to transcribe mp3 file.

I already downloaded model onto my local machine so this initializing part is unnecessary since I've used the program multiple times previously.

what caused this one? I deleted and downloaded from the app store but still same

Developer Response ,

in V5.4.2 use deepgram transcribe English will has an diarization option.
It can auto-tag the speaker. then you can use @speaker name keyword to filter the speaker's rows then batch change the speaker name or use replace feature "@speaker-->new name"

whisper transcribe engine diarization feature will continuous research for the best implementation approach.