My #1 wish for SunVox: Record into existing pattern

[jbw]

I realize SunVox and trackers are sort of set up for entering data with a keyboard, and "programming" songs, however I think being able to record live playing and record live automation into a preexisting pattern of a fixed length would be a HUGE step forward for SunVox.

Example:

-create a new empty pattern of 64 beats
-enter some notes and other data (current way to use SunVox)
-hit record (maybe have a specific "Record into selected pattern" button)
-listen to looping pattern and add notes (maybe even have an option to delete)
-keep listening and add parameter changes (filter sweeps, or other automation movements)
-and so on and so forth

This would be a great way to work with patterns, and create variations of patterns, without having to trim anything.

Maybe you don't want to take SunVox this direction, but I for one think it would be a really nice addition without compromising the spirit of SunVox, because after all you could keep the old way of creating patterns. This would just be an addition.

I hope to see it someday soon!

[jbw]

For notes you could also toggle the record to either replace notes on a line where previous existing notes were overwritten, or add to notes on a line so that previously existing notes remained.

[Twindle Jumper]

I'm also searching around for how to do this, so that I can keep overdubbing the same loop without going outside of its boundaries. I even feel like I remember doing it and just changed a setting or something... But I can't find how to do it now.

[Keres]

You can do this easily with Linux/Jack and 2 copies of sunvox. Sunvox even has a built in Keyboard1 and Keyboard2 MIDI device that works between copies under normal ALSA. Timing is an issue with MIDI, but Jack is a real-time audio server so latency is really good.

so you can say... get a collection of metamodules with each multisynth sequence in them, and fade between them using a slider and select different patterns in the timeline, and record pattern data to another copy of Sunvox using MIDI.

several things to note:
1. latency causes MIDI glitches. sometimes if your latency is too low then what you hear when you record will not quite match what you play back. you can quantize this or increase your latency.
2. Helps very much to have a 3rd app to send MIDI to start both copies at the same time.
3. diving into a metamodule while using 2 copies of Sunvox causes havoc with Jack sound server MIDI connections and disconnects them, but you can just put them back.
4. Sunvox .ini and setting files can be a real headache in linux. You can easily cause huge logical problems as to what copy you are actually editing because the ini files in sunvox store your metamodule edits. you can dive into a metamodule from one copy of sunvox and when you pop out it either crashes or you show up in the other copy of sunvox you are running. that's because they are sharing the same .ini and settings files!!! needless to say you loose your edits of that metamodule, if it does not outright die and crash. the solution is to run sunvox as 2 different users. this is hard because of permissions, so right now i am exploring how to give user /home/sunvox2/ the same permissions as my session here. since the user sunvox2 doesn't have console or hardware permissions, it won't even run in my session. the workaround is just to open sunvox as the root user with a separate config in /root/. it works perfectly, but i suspect that it is dangerous.

for all the work it takes to set up, it's worth it. The fastest way to set up jack is the UbuntuStudio distro. IM pretty sure that you can just do it from the livedisk. install libsdl2 and sunvox and be jammin without even an OS install.

[Keres]

PS: my no.1 wish for sunvox was the 29xxyy command, i just didn't know it till i played with it ;p

[transient]

i would like to +1 this request. ability to record into existing pattern (and loop for overdub) would be a fantastic addition. the current method of recording works but really limits any kind of live playing in terms of fitting that playing into workable chunks of pattern.

[Keres]

wait, cant you already do this with XXYY levels drawing???

[zumbasa]

@Keres: Can you elaborate that?

+1 on this one, too.

I think a simple option would be to introduce a loop marker to the timeline. So additionally to the track automatically looping after the last pattern, the track could loop from (or within) the loop markers and so would the recording while the loop markers are aimed (Midi control!). Internally maybe the recording could just be re-triggered, overdubbing the existing record. On re-aiming it would create a new pattern. I just do not see an easy option the re-trigger an overdubbing recording, but thats cherry on top I think.

I think this would be a great addition for grooving, studio or live setup alike.


Side idea:
I just thought about multiple numbered loop marker pairs that are are stepped through in sequence, giving the ability to quickly arrange a very compact track.

[Keres]

have you tried recording from sequencer1 (or another open copy of sunvox) to the main sunvox where you are recording, and use the pattern jump commands?

31 - Jump to line XXYY (address) right after the end of the current line.

32 - Set jump address mode YY:

0: jump to absolute address, relative to the start of the timeline (default);
1: jump to ( pattern beginning + address );
2: jump to ( pattern beginning - address );
3: jump to ( next line + address );
4: jump to ( next line - address ).

im not sure how that would interact with recording patterns. my thought is that the pattern jump commands would be more useful in the performance sequencer.

[zumbasa]

Thank you for your help. I kinda got what I needed by patching together two instances of sunvox, using the jump command, Input and sampler modules.

This would be perfect if only for the sampler. It lacks the possibility for midi triggering for recording. Also the sampler would need to starting recording right away when the threshold is set to zero. I can come around that part by adding a very silent noise module though Still, the MIDI triggering of the recording for the sampler module is the major roadblock for me, now.