It’s been about 10 months since any update or new features on the Mac, I haven’t check the ios app, don’t really use that very much. Any way just really wondering what’s coming and possibly when, it’s my main transcription tool.
I’m hoping that you won’t have to wait too much longer for an update. I’ve been trying to wrap up the all-new MIDI and keyboard commands support. It is proving to be a total beast of a feature to build, mainly because I had to build all the (cross-platform!!) GUI to support the configuration.[1]
I shared a little sneak peak at the feature on Mastodon a few months ago: https://mastodon.social/@liscio/113556932417132567
You know the old saying about “the last 10%” of a project that takes 90% of the time? Unfortunately I’m buried pretty deep right now in that final 10%.
If I had stuck with the configuration text files, the release would have been done a while ago. However, it’s unlikely that anybody would want to use it. ↩︎
3 Likes
Hi Chris,
Delighted to see something finally happening in Capo!
Can I suggest you post these updates here in the Capo forum please? I don’t use Mastodon and don’t want yet another social media account. As a result none of your work has been visible to me and frankly I debated whether or not to renew at the end of December.
Seeing as you’re working on the app again my key requests are:
- Ability to overlap regions. Even if this was only possible with 2 or 3 regions it would be a massive help.
- Allow me to set the start of the song and only then calculate the bars, beats etc.
- Have more controls visible in the UI. For example volume AND tempo controls both visible together
I also only use Capo on my Macbook btw.
Thanks and regards,
James
James,
Sorry I missed this when you posted. I’m not sure why I didn’t receive the notification about your message here.
Anyway, I probably should have shared that progress report on here rather than Mastodon. In the next update, I’ve added some in-app links to this community site which—I hope—will drive more people to use it and make things less quiet around here. That’ll certainly force me to be better about that in the future.
Cheers,
Chris
No worries thanks for the response.
I started following you on Mastodon several weeks back when I
saw your post about Mastodon. In fact checked Mastodon this morning and saw the post on your ACM article. Congrats, took a quick look at the article before work will read more this weekend.
Just a quick follow up on the ACM, when I first started using music slow down tools, Jennifer Batten has some youtube videos on using the music slow down tools and transcribing, that helped me, I don’t think the videos were so much about ear training as they were on how to use a slow down tool to transcribe, IMHO ear training is an important aspect of the using the tools. Looking forward to reading more of your paper this weekend
1 Like
Hi Chris,
No problem at all. I’m glad to see your excellent product being developed further!
Any plans for AI?
All the best,
James
I have nothing planned for Capo that involves AI. If anything, I dabbled with LLMs a little bit to assist with repetitive coding tasks for the new update, but sadly it’s not as much of a productivity booster as I would have hoped. If I wasn’t paying close attention, it would have introduced a lot of hard-to-find bugs.
I’m curious, though—what sort of AI features would you have hoped to find in Capo?
Hi Chris,
I took a look at Moises. The thing they do very well is the separation between “segments” ie: vocal/guitar/bass/drums, etc. I’m assuming that’s done using AI.
You can select which segments you want, what volume they should be at, and you can loop segments. There is very little crossfade between instruments etc so the result is impressively clean.
I really don’t want to imply that it’s a better tool - it isn’t, and Capo remains my number one tool which I use daily for learning songs, practising and so on.
If that is indeed AI in action then there must be other ways it’s useful, perhaps for beat recognition for example.
If it helps you I can send you a screen capture of it?
Regards,
James
I’m well aware of the app, and also the technology behind it. Unfortunately, I don’t know if or when you’ll ever see that get into Capo. A major part of what’s holding me back is my own ability to build it.[1]
Beat recognition is another area that I could certainly improve in Capo with deep learning, but the last time I checked there were still a lot of major limitations in the cutting-edge techniques. For example, changing time signatures, triplet feel, and wild fluctuations in tempo are all going to throw things off significantly. This is why I added the beat editing features in the first place, and why I think that making them easier to use might be a better use of my time.
I could go on for days about why this is challenging. It’s not just a highly complicated project technically, but there are significant barriers to (legally) obtaining a good collection of training data. ↩︎
All good and I agree with you about improving the beat recognition. While I like the separation in Moises it isn’t a core feature for me, just useful to practise over.
1 Like
In case you folks aren’t notified about new topics here, yesterday I created a thread to invite everyone to try the Capo 4.4 Public Beta.
If you are able to help, please join in and try out the new release. Thanks!
1 Like