Feature request: Exchange voices is the web app

[CarlosUnchained]

It would be nice to have an option to exchange voices in the web app. This would help to re organize voices without having to delete and re upload the content for each voice. I say exchange, because there's no data loss in that process, so for example exchange voice 5 samples and settings for voice 9.

[marchingband]

That would be great but, the WEB UI doesn't have all those files. It would have to fetch them from the WVR, and then re-upload them to their new voice. There are a number of convenience features like this that arnt possible because of this.

[marchingband]

What I have considered in the past is a big matrix to "patch" channels to default voices, which could fulfil the same function. Would that work?

[CarlosUnchained]

Ah I see. Makes sense. The matrix would help visually to do what you can with PrCh. I'm def not a fan of having different channels assigned a voice other than the same channel, with that many voices is easy to loose track. The matrix could help in that regard by having visual feedback when needed.

On another topic, don't want to start an issue for a question: How is the mixing approached?
I think it's pretty good for the amount of sources and levels to handle, I'm not sure if there's room for improvement.
I run into clipping with some multi sampled instruments. Gain staging is a bit hard and maybe some kind of limiter would be nice to have, also simple compression. Again, no idea what's in there already and if there's any room left.

I have been recommending the project on different sites and people seem to be interested :)
Thanks again for all the feedback so far!

[marchingband]

So this matrix would just be for the defaults. When WVR first boots, it would use this table to assign channels to voices. This would avoid the user having to send a bunch of prch messages to get the channel/voice map into the desired state. And it would make it easier to fit a new device on a new channel into the mix.

I would say that it sounds to me like you may not need to use all these channels. They typically are reserved for discrete devices. If I have multiple sets of sounds (drums, synths, voices, etc) but all are controlled by one device, I would usually keep them all in one voice, and use ex. octaves to organize them rather then voices. There are 127 slots in each voice, which is very often enough to work with for a single device. Is it the case that you are using all 127 notes and require more voices? Just trying to get clearer on the use case that has you struggling with the channel/voice map approach.

So there are no real-time effects on WVR like compression or limiting, there is not enough processor power for that. The mixing consists of a single >> operator, which effectively divides every mixed sample by (IIRC) 16. This is not an exact science, it's a balance of on one hand estimating the likely max volume, and on the other, limiting maximum potential fidelity, as there are only so many bits available.

I havn't experienced clipping very often, and I consider the final gain stage to be fairly conservative. I noticed that you asked about normalizing. If you have saturated samples that are all normalized, it may make it possible to clip. I would suggest to lower the volume of your samples at some point in the process, that could be before you upload them, in the UI as you do the upload, by lowering the velocities you are sending, or by lowering the volume or expression on that midi channel.

There is also always the option of increasing that final >> operator, but that would be a case of a custom firmware, as it has been working well the way it is, so far, for most users.

[marchingband]

Also FYI I created a discord server, using the new community mode, to replace the Google group. It's a handier place to ask questions, if you like. I am still getting the hang of it, but this invite should work. https://discord.gg/WugqSsUXTf

[marchingband]

Oh and thank you so much for spreading the word on WVR! I appreciate that so much, I would love to see the community grow more this year :)

[CarlosUnchained]

Is it the case that you are using all 127 notes and require more voices?

In my case each voice is an instrument, and I may want to control them from different sources, specially with button/gates of the Eurorack shield. Think of a voice as drum racks, pools of textures or multi sampled chromatic instruments. Mentally is way easier to have instrument per voice, and not several split by note ranges.

I can't transpose what the gates and buttons target without pin reconfiguration so the obvious is to change bank (channel1).

Thanks for the tip regarding levels. I'll try a different approach for the samples I've had troubles with. Also saw there's a way to adjust overall volume with the pins, may try that too.

Great to see there's a Discord! Just joined. I'd be great if there could be a chat-like board to talk or share related content, I have a little video using WVR that may help you to see how I use it with midi. It doesn't really deserve its own post.

[marchingband]

That makes sense. I havn't spent a lot of energy considering the combination of pin triggered notes and midi triggered notes. The eurorack module does present some new use cases. Let me think about it.

[marchingband]

Added a general-chat text thread to the discord. Thanks for the suggestion!