We all pay a lot of money for the Adobe Suite subscription, but for me, one of the most useful applications is one that was "promised" but still doesn't exist. A few years ago, Adobe teased us with an astonishing "Photoshop" for voice prototype. And that was it! Years have past and despite the interest, not another word about this product from Adobe, not even a lesser version.
Now, I learn that Techsmith just released an application that does almost everything I've wanted, that is, Audiate. Go check it out. It allows for direct editing of voice waveforms with a Voice to Text interface. I haven't tested how good it is, but even if it's not perfect and it doesn't synthesize sentences, at least it's a start, especially for editing voice tracks.
How did Techsmith beat Adobe at their own game?
I'm looking forward to someone telling me that I am wrong and Adobe does have some hidden gem in the vast array of tools that are part of the Cloud package (75% that don't need).
I'm mostly asking this question because I need to know if I should purchase the Techsmith application or look at some application that already exist in the Adobe Cloud.
Copy link to clipboard
I'm assuming that you are talking about VoCo. And the reason it's not available has nothing to do with being 'beaten at their own game' or anything like that - technically at least at a basic level it was quite feasible. Adobe's lawyers, when they heard about it had other ideas, mainly revolving around potentially massive court cases, possibly resulting in the company being cleaned out. And it was dropped.
So you're not wrong about it not being there, but that's not for any technical reason, as far as I'm aware. Just a pretty good legal one.
There are already free AI libraries that with a bit of work, can effectively synthesize voices with minimal voice sampling. See the YouTube channel, "Two Minute Papers" for some astounding examples of AI projects.
But sorry, my post was a bit misleading... all I really want at this time is a way to edit audio using the text representation of the content, as opposed to looking at a less meaningful waveform and needing to waste time reviewing the audio. I spend a lot of time editing audio, and I'm thinking that Techsmith's newest application may save me countless editing hours.
I'm just a little bit annoyed that Adobe hasn't released that ability in all these years, they basically prototyped it four years ago. Again, without the voice sampling and synthesis part of that prototype.
Oh, and I had a look at Audiate. At present it does not do what VoCo was claimed to do. And I think that there's a very good chance that Audiate won't be able to do that in the future either. Lyrebird made similar promises and that hasn't materialised, very likely for exactly the same reasons. In terms of what Audiate does now, then yes some people might find that useful. That said, I'm not aware of any Adobe announcement that said they were going to develop an app that let you edit speech by turning it into written words, so if it happens at all, it won't be for a while.
Like I said, I started this thread on the wrong track. LOL I don't really care that much for the cool VoCo or Lyrebird voice synthesis... cool but not really what I need now.
But the ability to edit audio speech in its text equivalent is something many people who regularly edit audio files need.
> some people might find that useful.
Ha ha... I'm going to guess audio speech editing isn't something you do often.
I just downloaded the application trial. Let's see if it really works as well as advertised.
Copy link to clipboard
I'm rather disappointed regarding the transcription speed of 1:1, meaning that it takes about 1 minute to transcribe 1 minute of audio. Since it's speaker independent (unlike Dragon Dictate), I have nothing to compare it to. But at this speed it's not at all useful because I can complete a less accurate but acceptable editing in about half the time it takes Audiate to complete its transcription.
Accuracy? I'm not sure. It looks pretty good with the different speakers involved but I need to wait another 35 minutes before I can really get into the editing.