Hey, sorry for the late reply, i was on vacation

@p4rancesc0:
Only the first version is public, the following updates are private.

@grizzlydj:
Unfortunately that is the expected behavior. In the current implementation it takes about 2.5ms before the next midi note is played. For example, triggering 4 notes before the hihat at a step results in about 10ms delay. As i wrote earlier i'm working on it in the upcoming expander mode after i have finished flam. But the technical limitation is still 1ms, the time it takes to transmit one midi command. Here's a screenshot how it looks like (noteOn ch10: 153, note 35, velocity 48):




@tb303:
Have you tried v1.027? It has a small fix that could solve your problem.
Besides i do not recommend to stay below v1.026 because of a buggy mux board handling!

@gihaume:
I suspect the same issue as above, spread out notes over time. Is there a variation in sound when you listen to a single voice only?

@selliercdrc:
I see no reason for that. Nava has only one physical trigger out.

Interesting.... these delays can not be removed in any way?

by the way I tried yocto2 in slave mode and an empty pattern when the sequencer is running and now the midi is working fine, but when the sequencer in master mode there are the same delays as nava ... maybe this will help ... nava in slave mode with an empty pattern does not work correctly at all - open hihats reproduced instead of closed any time... strange

Thanks:)

I can get rid of (almost) all the internal delay, just wait for the next update. All expander suffer from the 1ms delay due to the slow midi baud rate.

Like i wrote earlier, i do not recommend to use slave mode for midi play. Future firmwares will have that "functionality" removed, just like the flam test version. Instead there'll be the dedicated expader mode.


Btw, is there a midi vst wizard in the forum? I'm thinking of a special function to "collect" midi notes, maybe with midi ccs as start and end markers. Something like this:

-> send cc xy value 0 //start collecting notes
-> send midi note a //nava doesn't trigger, accumulates the notes
-> send midi note b //
-> send midi note c //
-> ...
-> send cc xy value 127 // trigger all notes at once.

If someone is able to code this (cc before and after a "chord" and the packet should be sent earlier by the number of notes in millisec), i'll include it in the firmware. Maybe that's even possible mith max4live...

I still really hope that everything will work out and it will be possible to reduce all delays to a minimum I look forward to the next update! Thank you!

Zabox I edited my last message, it seems there is no issue as I described finally.

Bump! Any updates on development?

I didn‘t have much time the last month and i‘m on vacation again this week.
Planned release is in mid/late august.

Another bump!

Maybe i find the time to finish the firmware next week.

but the expander itself is already working great.

v1.027 Ch1: BD trigger, Ch2: SD trigger, Ch 3 LT trigger. Ch4: Midi in (sending 3 notes at once, midi messages are between the vertical cursors)


v1.028:


v1.028: (magnified) latency is < 0.1ms