Currently you are offered the ability to insert an audio effect pre or post fader. You can then move the effect up and down the chain to where you want it. That may be pre or post fader and in series or parallel with other audio processors. (FYI the same is true for MIDI effects, but without pre/post fader.)
In a future version of zynthian you will not be offered where to insert an audio effect. It will always be at the same position within the chain and you can then move it to where you want it. This simplifie the workflow, making it more consistent. I would like your opinion on where the default location for new processor insertion should be. Options are:
At the start of the chain, before any other effects and pre-fader.
Immediately before the fader, at the end of the pre-fader effects.
Immediately after the fader, at the start of the post-fader effects.
because if I add something, I think it’s logical that it’s added behind what i already have
also Post-fader, as that is how I work with hardware mixers and FX.
I don’t see why people prefer pre-fade above, you’d never want that on delays and reverbs (which is usually my first go-to fx).
This is probably the most common use of effects in hardware mixers, especially in a recording workflow. I’m not sure if it is as common in this instrument workflow. We will see…
I was working on that today! It should be available in testing soon but unlikely to ever be in Oram.
Keep on voting and do include your reasoning / logic if you can as it helps us all to understand what workflows benefit from such a decision.
It may influence your decision to know that there will also be proper send & return loops with the send pre/post fader switchable on each chain.
IMHO, It should be at the end of prefader segment (option 2).
It should be explained that postfader fxs adds 1 period of latency, so you have to use It with care for live playing.
For avoiding the extra period of latency we would need to split the mixer in 2 separated jack clients. This could be planned for the future, but it’s not in our current list of tasks.
This extra latency is removed in the dev version which does indeed split the the mixer in two. So it is on our current list of tasks. I am working on that now. This won’t be in Oram.
Maybe I am not educated enough in this very field. But reading up on the reasons of pre- versus post-fader, I only see explanations involving (aux) sends, used commonly for shared effects or sidechaining. But as far as I know, in Zynthian, we only have the signal being processed within a single chain, before it leaves that chain. Please correct me if my understanding is faulty.
So does it really matter in the current situation?
Either way, I’d vote for an end position, I.e. 2 or 4.
From the presented options I’d prefer No. 2. Two other options would be:
After the last effect already present, be it pre- oder post fader
After the currently focused controller page in the chain, be it pre- or post fader.
Personally I’d say that for time based and filter fx it doesn’t even matter if it’s pre- or post fader. We are still talking “insert” effects, not “send” effects inside the chain position.
Talking about non-linear or dynamic effects I rather would place them prefader always, because otherwise moving the fader in the mixer would not only change volume like intended, but also sonic quality. Imagine you put a compressor (or overdrive) post fader, then fader movement would affect the compression behaviour or the gain before the “overdrive cirquit”. I’d narrow down the purpose of a channel fader just affecting volume. This still doesn’t mean that this kind of behavour might be intended by you, but then you can move these effects manually.
I think (correct me if I’m wrong), what @rv5 is referring to where you want your delays and reverbs is that you might want them on a “send bus”. Personally if I would put these effect on a send bus (which I would not in a Zynthian workflow most of the time) even there I would put them pre fader, because some Reverbs or delays might inherit some nonlinearities (“tape delay”, “tube delay”, “plate reverb input gain”) which you might not want to affect by the send busses fader.
You can keep voting and giving opinion but with the current results it looks like an overwhealming preference for 2: insert immediately before fader, i.e. end of pre-fader section of chain. That is handy because that is what is currently coded in the development branch.
Thanks for the extra suggestions. (I knew there would be…)
I don’t like the first of these options because it feels inconsistent. I see the logic that someone who puts an effect post-fader may want the next processor to also be post-fader but then the next chain may have different behaviour. I think this variable behaviour is likely to be more confusing.
The second option has some logical sense and I considered this (and maybe when we get a more graphical interface to this we can revisit (but we don’t really normally see the chain when we are adding processors so again, it may feel a bit random or variable to many users.
I don’t see what makes the workflow simpler, if two or more steps are to be done after the “improvement” when before that it was possible to simply do it in one step and have the choice of which one.
It really makes me angry to be forced to do many steps (with possible software failures and shortcomings) when it has been much simpler before that disimprovement.
If the fx appear where they naturally should then you are not forced to do anything else.
If you are asked where you want them then you are required to have knowledge you might not have.
