CV amount readout inconsistencies - Trigger / Env Follower

[Audiotic]

Hi there,

I'm running into a problem when setting up Drum Triggering with CV or Envelope Followers. I'm afraid this is an issue with Reason handling CV and there won't be a solution, but maybe some of you solved the issue.
There are no problems when you don't need velocity changes (except a slight delay that can be easily fixed), but for triggering Drums from a recorded drum-track, the cv amount / velocity readout is essential.

My setup:
- Original drumtrack (lets say Snare) has a CV generating device connected (via parallel or FX out or Direct Out, doesnt matter). This can be M-Class Compressor, Pulverizer, Sweeper, DEG-2 generator etc... Doesn't matter which device is used, I tried all of them
- Triggering device can be any Reason Device allowing CV/Gate Trigger... lets say KONG (Tried most of them)

When I record the triggered drums (Audio or MIDI via CV to MIDI player), the velocity is not consistent. Some of the triggers are much quieter than they should be.
You can go in and manually edit the faulty velocity hits, but that's obviously a suboptimal workaround.

The screenshots show the problem:
The original drum-track is (for the purpose of demonstration) the same sample, duplicated (same level, sample start etc.)
The recorded triggered drums-track shows different levels on some of the hits (screenshot 1)
The recorded midi shows different velocity on some of the hits (screenshot 2)

What's weird is that the readout looks more or less consistent when using a CV readout device (screenshot 3)


Any thoughts?

Cheers

[deeplink]

I think some more info is required.. like what devices you are actually using..

First thing that comes to mind though, is that there a big difference between a trigger and gate - one of which had Velocity info "contained within it".

Secondly, could be that the CV out device is essentially mono Legato, so the envelope continues even after its been retriggered

[Audiotic]

I think some more info is required.. like what devices you are actually using..

First thing that comes to mind though, is that there a big difference between a trigger and gate - one of which had Velocity info "contained within it".

Secondly, could be that the CV out device is essentially mono Legato, so the envelope continues even after its been retriggered

As mentioned I have used many different devices for testing. The screenshots are recorded with the MClass Compressor, but same behavior with Pulverizer, sweeper, etc.
Velocity info is contained, as seen in the CV readout screenshot.
Legato is not the problem, since the samples are re-triggered correctly as seen in the cv readout.

[jam-s]

A minimal example project which exhibits this behaviour would help in debugging it. You can zip it and attach it here if it is not too big.

[Loque]

I speculate it is related to MIDI value range. AFAIU the following can have an impact to the result:
* Envelope Follower settings at all
* Value conversion
* value range
* recording speed (MIDI recording)
* timing of sampling (recording of value)
* general delays in recording line

And I speculate if you lower the song speed, the result meets better the expectations.

Are you recording gate signals? They behave different then automation values and require a specific change to be recognized as a Gate. Try to scale and shift (offset) the values to pass the 0 more recognizable. Maybe faster song speed with increased space between trigger samples has better results with recording Gate signals... Worth trying...

[Audiotic]

I speculate it is related to MIDI value range. AFAIU the following can have an impact to the result:
* Envelope Follower settings at all
* Value conversion
* value range
* recording speed (MIDI recording)
* timing of sampling (recording of value)
* general delays in recording line

And I speculate if you lower the song speed, the result meets better the expectations.

Are you recording gate signals? They behave different then automation values and require a specific change to be recognized as a Gate. Try to scale and shift (offset) the values to pass the 0 more recognizable. Maybe faster song speed with increased space between trigger samples has better results with recording Gate signals... Worth trying...

Thanks for the input.

I tried lots of different things, using cv utilities to change cv offset, scale and envelope. Gate or cv outputs makes no difference.
Recording the triggered audio gives the same results as recording (or sending to track) the midi data via a cv to midi player.

I think it comes down to some internal timing issue when translating the cv values to midi (velocity) because the inconsistencies sometimes have a weird consistency. (i.e. Getting quiet at the same instances after starting playback)
Changing playback speed does change the timing behavior, but doesn't fix it.