It's been almost a year since I've touched this project. Coming back, there's quite a few things about it I'm changing:
I have no plans to develop WavMod anymore. I plan to ship Fluid Vocal with the W4U or EFB-GT resampler, and the open source wavtool-yawu if I can get it to work. I hoped that it would work with Moresampler, but unfortunately there's some compatibility issue that would probably require a ton of modification to fix. I'm more focused on getting an Utau-like program, rather than beating it (as I wanted to before), and I'm putting more emphasis on compatibility. For one, I'll most likely use .ust files as the native format. I need to look at some Japanese VCV multipitch banks, since this is pretty much all of Fluid vocal's weak spots (it's been so long since I've looked at voicebanks, and I never used Japanese ones anyway).
There is a good chance that most current Japanese voicebanks won't work in the long run, due to the locale issues I'm sure all Utau users are all too familiar with. With Fluid vocal it's this whole issue but in reverse. Fluid vocal is a US locale program, and won't play nicely with the Shift-JIS that virtually all Utau banks use (but should work fine with Japanese file names made on a US locale machine). I could support these voice banks by making Fluid vocal in Japanese locale, however that just perpetuates the problem - it would only be compatible with JP locale just like Utau, and everyone would still be making Shift-JIS voice-banks. I think it's better to stick to supporting US locale, even if it means some compatibility issues. Hopefully, voicebank devs would move away from Shift-JIS, and thus fix the problem in the long run. I'm just talking off the top of my head here however, and all this needs much more consideration. First priority is to get this program to work at all.
Which brings me to "working". As of now, I think I've got the biggest issues with the rendering system fixed. Before, the overlap and envelopes were all messed up terribly, not to mention a stupid mistake with file names that allowed no more then 10 notes to properly play back. That's fixed now, and for the most part (with English CVVC/VCCV) it sounds like completely un-tuned utau, since pretty much what it is as there's no pitch control of any sort yet (this is a big issue I'll talk about in a minute). It also does so semi-reliably, there's no constant crashing and glitches quite like their was. I've also added the ability to insert notes into the middle of the row (before you could only add them to the very end), however the notes don't line themselves up horizontally like they should yet. Zooming is also implemented and working as far as I've tested.
I'm abandoning the idea of the 2-view workspace I originally had in mind. My original idea involved 2 "modes": an editing mode for adding notes and lyrics, and a tuning mode for adjusting pitch and properties. As of now the GUI plan is just a single view, with a Vocaloid-style pitch clip (and other properties, I'll probably put flag-editing here) right below the toolbar. Right-clicking a note reveals a menu with all the properties as well as the envelope controls, which mostly work now. Exactly how everything else will work is still in the air for now.
Anyways, I really do hope to finally get this project up and going! It looks more likely now than it ever has.
I have no plans to develop WavMod anymore. I plan to ship Fluid Vocal with the W4U or EFB-GT resampler, and the open source wavtool-yawu if I can get it to work. I hoped that it would work with Moresampler, but unfortunately there's some compatibility issue that would probably require a ton of modification to fix. I'm more focused on getting an Utau-like program, rather than beating it (as I wanted to before), and I'm putting more emphasis on compatibility. For one, I'll most likely use .ust files as the native format. I need to look at some Japanese VCV multipitch banks, since this is pretty much all of Fluid vocal's weak spots (it's been so long since I've looked at voicebanks, and I never used Japanese ones anyway).
There is a good chance that most current Japanese voicebanks won't work in the long run, due to the locale issues I'm sure all Utau users are all too familiar with. With Fluid vocal it's this whole issue but in reverse. Fluid vocal is a US locale program, and won't play nicely with the Shift-JIS that virtually all Utau banks use (but should work fine with Japanese file names made on a US locale machine). I could support these voice banks by making Fluid vocal in Japanese locale, however that just perpetuates the problem - it would only be compatible with JP locale just like Utau, and everyone would still be making Shift-JIS voice-banks. I think it's better to stick to supporting US locale, even if it means some compatibility issues. Hopefully, voicebank devs would move away from Shift-JIS, and thus fix the problem in the long run. I'm just talking off the top of my head here however, and all this needs much more consideration. First priority is to get this program to work at all.
Which brings me to "working". As of now, I think I've got the biggest issues with the rendering system fixed. Before, the overlap and envelopes were all messed up terribly, not to mention a stupid mistake with file names that allowed no more then 10 notes to properly play back. That's fixed now, and for the most part (with English CVVC/VCCV) it sounds like completely un-tuned utau, since pretty much what it is as there's no pitch control of any sort yet (this is a big issue I'll talk about in a minute). It also does so semi-reliably, there's no constant crashing and glitches quite like their was. I've also added the ability to insert notes into the middle of the row (before you could only add them to the very end), however the notes don't line themselves up horizontally like they should yet. Zooming is also implemented and working as far as I've tested.
I'm abandoning the idea of the 2-view workspace I originally had in mind. My original idea involved 2 "modes": an editing mode for adding notes and lyrics, and a tuning mode for adjusting pitch and properties. As of now the GUI plan is just a single view, with a Vocaloid-style pitch clip (and other properties, I'll probably put flag-editing here) right below the toolbar. Right-clicking a note reveals a menu with all the properties as well as the envelope controls, which mostly work now. Exactly how everything else will work is still in the air for now.
Anyways, I really do hope to finally get this project up and going! It looks more likely now than it ever has.