PreenFM > preenfm2 and preenfm3

Volume mixer

<< < (2/3) > >>

lokki:
well yes and no... you see, when you use a sound with less harmonics, the headroom becomes a problem pretty fast. you can hear a distorted sound on a sine VERY quickly. that is why i think it is smart to have it that way. also where would you draw the line? have it always at 16 voices (the dac setting that is) so that when you play monophonic you only use 1/16th of the dynamic range? hmm.

to circumvent your volume problem in live settings i would create at least one instrument which fills up all the voices. once you need more on the bass for example, you simply reduce on the other instrument first.

that being said, it would be relatively trivial to change this in firmware. (xavier has done this for the 8timbres version and it is just one calculation that has to be changed)

mzero:
yeah, I guess I would set as if all 16 voices were playing... So playing a single voice is giving up almost 4 bits of the DAC's range, so from 18 bits down to 14. But that's still a tremendous dynamic range for a single voice. I'd be surprised if it were a problem....

But, one approach would be - assuming you have a "master instrument" mixer like O.P. in this thread proposes - and assuming that is scaled so that you can achieve +/-12db (say). Now, if the output of each instrument is fixed to the maximal polyphony possible, then.... if you are only playing one instrument or just one voice, you can use this mixer page to regain those bits in the conversion.

joradsilver:
+1

Just got the box. Love the sounds and the editing but for performance 100% need this functionality to use well. Please advise when this get updated or someone finds a way.

mzero:
Reading through the source code, I discovered that there is already a CC mapped to volume, it just isn't obvious:

CC 73, Filter Gain, is actually gain1 no matter what filter or final effect mode is selected, even none.

If Xavier's intention is that parameter always be simple channel gain as part of any future effects, then this parameter could be used for a in-box mixer mode, as well as externally.

I suppose if I were to attempt to code this (which I might), I'd be tempted to put an additional, per-instrument, pan and gain stage after the filter. I don't think it would cost too much in processing, and it would leave the row of filter parameters alone for future filter use. I would also be likely to implement a "constant voice gain" mode so that number of voices doesn't change the relative volume of an instrument.2

In the original post, there is a fx level. I suspect the intention was wet/dry. This doesn't exist in the synth. calculations today, but could be easily added (just another two multiplies and an add per instrument.)

Thoughts?

1: This filter stage gain calculation is linear. This gives a rather unnatural feel to the control. I'd think a quadratic response (squaring the gain) would yield a wider usable range, with a more natural feel, where the user can feel the control as scaled dbs.

2: It is understood that in a "constant voice gain", an instrument with but one voice would be 1/8 the volume it could be - thus losing some available range on the DAC if it were the only instrument. Personally, I'd be willing to give up these 3 bits in exchange for constant leveling of voices: one voice in a mono instrument is as loud as a single voice in a 3 voice instrument.

joradsilver:

--- Quote from: mzero on March 01, 2018, 04:41:11 PM ---Reading through the source code, I discovered that there is already a CC mapped to volume, it just isn't obvious:

CC 73, Filter Gain, is actually gain1 no matter what filter or final effect mode is selected, even none.

If Xavier's intention is that parameter always be simple channel gain as part of any future effects, then this parameter could be used for a in-box mixer mode, as well as externally.


--- End quote ---

After digging a little I found this option too. From a user experience though, it's still quite clunky to jump between instruments.

At the most basic level, all I'd like to have my hands on would be a simple volume mixer of the four instruments on on panel using the four bottom encoders. Seems like using the filter gain parameter could achieve this nicely if coded to do so. But your idea on per instrument pan/gain would be even better !  I hope you give it a try :)

I support any efforts to make this happen but unfortunately my own technical ability is not quite there yet  :-\


Will keep an ear here and hope to assist in any way I can from testing/usability perspective.

Navigation

[0] Message Index

[#] Next page

[*] Previous page