Add post-mixer Sends

[Din]

I’ve done some searching to make sure this isn’t a duplicate, so excuse me if I’ve missed another thread to the same effect.

I’d love to see Renoise Send devices that use the channel mixer output (after the mixer fader and pan pot settings) instead of the current pre-mixer-only setup. Pre-mixer Sends are nice, but Sends that use the channel mixer output itself (after the fader and pan adjustments are applied) need to be available.

Suppose I’ve got a nice drum track with a Send going to a Room reverb plugin used by other track Sends. I get the Send level just right so the drums and their reverb sound perfect together. After more work I decide to raise the level of the drums in the mix. Because the Send is using the channel signal before the post-fx mixer, raising the level of the post-fx mixer will have no effect on the signal used by the Send; i.e. it won’t raise the level of the reverb effect. Translate this to several tracks needing minor level adjustments, and I’m losing all kinds of time tweaking Send levels to match.

Adjusting the pre-fx mixer level isn’t really a good solution, either, since it ignores dynamics-processing (compressor, gate, expander, etc.) in the effects chain between the pre-fx mixer and the Send device. That really just moves the time spent tweaking from the Send levels over to adjusting dynamics processor thresholds.

It gets a little bit uglier when I want to use track level automation for crescendos/decrescendos, because I also have to automate the Send levels to keep them proportional.

This is all made worse for me due to my preference for mixing using K-14 metering (RMS level for forte passages is around -14dBFS instead of being normalized to 0dBFS; it’s fantastic for wide dynamics). As such I do a lot of mixer level adjustments while I work.

TLDR:
I love the current Send device, but adding the option for it to use the post-mixer signal instead of just the post-fx signal, or creating a new post-mix Send device would be a big boon for us mixer-fiddlers.

[kazakore]

@Din

I’d love to see Renoise Send devices that use the channel mixer output (after the mixer fader and pan pot settings) instead of the current pre-mixer-only setup. Pre-mixer Sends are nice, but Sends that use the channel mixer output itself (after the fader and pan adjustments are applied) need to be available.

The Pan pots in Renoise are at the begging of the channel, not the end. Add a sample, hard pan to left or right, add a stereo effect and listen to check for yourself.

Send Device has Pan, so you can adjust position into the Send Track here, actually better than with the one built into the track.

Adjusting the pre-fx mixer level isn’t really a good solution, either, since it ignores dynamics-processing (compressor, gate, expander, etc.) in the effects chain between the pre-fx mixer and the Send device. That really just moves the time spent tweaking from the Send levels over to adjusting dynamics processor thresholds.

It gets a little bit uglier when I want to use track level automation for crescendos/decrescendos, because I also have to automate the Send levels to keep them proportional.

And you can’t automate the Post Faders at all! So you are already one step ahead of using them here… Something I have repeatedly said is very strange behaviour, as it goes against the design of pretty much every hardware (digital) mixer, where you will be able to automate the Post Fader for certain, but the Pre Fader/Gain/Trim is less likely.

Suppose I’ve got a nice drum track with a Send going to a Room reverb plugin used by other track Sends. I get the Send level just right so the drums and their reverb sound perfect together. After more work I decide to raise the level of the drums in the mix. Because the Send is using the channel signal before the post-fx mixer, raising the level of the post-fx mixer will have no effect on the signal used by the Send; i.e. it won’t raise the level of the reverb effect. Translate this to several tracks needing minor level adjustments, and I’m losing all kinds of time tweaking Send levels to match.

Whether the fader is on the Send Device or the Post Fader you are still going to have to make as many adjustments. The Post Faders are larger and easier to use when looking at the Mixer view though (rather than the small Send Level you might display on each Send Device in Mixer view to make them all accessible at once.

In short: I don’t think having the Send after the Post Fader will actually help you much. Not with how it is currently implemented anyway. To get around these problems a lot of problems a lot of people always an a Gainer before the Send Device, so they can change levels above 0dB if required and also effect the level through the Send if using Keep Source.

[Din]

Then let me add “post-fx mixer automation” to my list of requested features. I’m more than a little surprised by its absence, although that may have something to do with the hard limit on the number of DSP devices in a chain. One way around that would be to tack the post-fx mixer and any post-mix sends on the end of the chain (device-numbering-wise). It would reduce the number of devices you could use before the post-fx mixer, but the maximum number of devices would be the same. That way, you could automate everything.

[Cas]

I agree with the problems posed in this topic. The routing options in Renoise are pretty damn good, but very counterintuitive for a beginner and frustrating for more experienced users. Making a Hydra on every track with a send, is also too much work and takes too much space to form a good solution. Something really cool would be native or a tool, to link postfx volume to the last send device in the chain for designated tracks, so that you can tweak the send level by itself, but then if you change the post vol in the mixer the send level would come along proportionally.
Anyway everything about the sends, and group-sends, would benefit from a better visual representation.

[Din]

I don’t think my explanation is as good as it can be, so here are some signal-flow-charts to show what it is I’m after. This is in response to a few replies to this post in a different thread. Behold my mighty MS Paint skillz!!

All levels for arrows represent the output level of the device relative to the original pre-DSP audio signal. All levels in devices represent any gain applied to the signal by that device. Thus a device that applies -6dB gain will output -6dB relative to its input signal. For simplicity all DSP chains are assumed not to affect the signal level.

First, we have what Renoise currently does:

Note that the Send is passing on an unaffected signal to the Mixer, but attenuates the signal passed to the Aux Return track by -6dB. This is how the Sends currently work when set to keep the signal.

Apply -6dB gain to the mixer output and you get this:

Now the levels of both the Mixer and the Aux Return are -6dB relative to the original signal, which is generally not desirable.

Now here’s what I would like to be able to do:

Being able to move a Send to the right of the Mixer (or add a new Post-mixer Send device) you can do the same -6dB gain on the Mixer and you get this:

Now the Aux Return Send level remains -6dB relative to its input, for a total gain of -12dB relative to the original audio signal.

Without this capability, Renoise will never be very strong in the mixing arena, especially for auto-mapping to control surfaces.

[wahrk]

@Din

I don’t think my explanation is as good as it can be, so here are some signal-flow-charts to show what it is I’m after. This is in response to a few replies to this post in a different thread. Behold my mighty MS Paint skillz!!

All levels for arrows represent the output level of the device relative to the original pre-DSP audio signal. All levels in devices represent any gain applied to the signal by that device. Thus a device that applies -6dB gain will output -6dB relative to its input signal. For simplicity all DSP chains are assumed not to affect the signal level.

First, we have what Renoise currently does:

Note that the Send is passing on an unaffected signal to the Mixer, but attenuates the signal passed to the Aux Return track by -6dB. This is how the Sends currently work when set to keep the signal.

Apply -6dB gain to the mixer output and you get this:

Now the levels of both the Mixer and the Aux Return are -6dB relative to the original signal, which is generally not desirable.

Now here’s what I would like to be able to do:

Being able to move a Send to the right of the Mixer (or add a new Post-mixer Send device) you can do the same -6dB gain on the Mixer and you get this:

Now the Aux Return level remains -6dB relative to its input, for a total gain of -12dB relative to the original audio signal.

Without this capability, Renoise will never be very strong in the mixing arena, especially for auto-mapping to control surfaces.

Use the PreFX fader if you’re okay with the DSP input levels going down. Otherwise, just toss a Gainer in there. I still see this as unnecessary.

[2_daze_j (Jonathan)]

The truth is: “its not a bad idea. and it would add a lot.” The other truth is: It may be very far from the original Renoise design, and consequently very difficult to implement.

What Renoise does have is: Access to industry standard Rewire.

Sometimes: Things are the way they are supposed to be, and when we change them, we destroy their original purpose.

[Din]

Sorry, but this thing has ballooned a bit in the writing. If anyone hasn’t already figured it out, I’m wordy.

Also, I’d like to thank everyone who has participated in this thread so far. It has helped me really iron things out

@wahrk

Use the PreFX fader if you’re okay with the DSP input levels going down.

The only reason to adjust the PreFX fader (input gain, really) is for reining in incoming audio that’s too hot for the DSP chain, like VSTi’s with built-in feedback, flange, chorus, distortion/gain, and the like. Also, as I said in my first post, using the pre-fx level screws up dynamics processing DSPs.

Otherwise, just toss a Gainer in there.

This is actually the best kludge available, but there are some cons. If I were to use a pre-send gainer for all tracks with sends, I’d usually have gainers in all tracks but the bassline. I’d be using the gainers to control levels in all tracks but the bassline. The post-fx mixer faders would be entirely unnecessary in all tracks but the bassline.

I still see this as unnecessary.

I disagree. Given you’re suggestion to use a gainer to achieve the same result, the functionality itself is not unneeded. The problem is in mapping controls to a hardware controller. Ableton Live and REAPER both support the Mackie control protocol, which allows basic mixer controls to be automatically (this is the important bit) mapped to the hardware when a new track is added. I assume Duplex is a way of doing the same thing within Renoise without the developers having to break their hardware-agnosticism. But if Duplex works the way I think it does, it’s going after default channel controls, and a Gainer tacked into the DSP chain is not a channel default. Keep in mind that templates are something else entirely: template != default. Right now the post-fx mixer is really just a dressed-up closing brace for a DSP chain container. For me, the Mixer view itself is only really useful for managing DSP chains across multiple tracks; it’s no use to me for mixing.

Even leaving out hardware control, the mixer itself suffers for the lack of this basic functionality. If you use a gainer for any track-level control, why not use it for all track-level control? And then who really needs the post-fx mixer if everyone is using gainers instead?

@wahrk

The truth is: “its not a bad idea. and it would add a lot.” The other truth is: It may be very far from the original Renoise design, and consequently very difficult to implement.

I’ll leave it for the developers to decide on whether or not it’s too difficult to implement. I’m offering a suggestion. I love Renoise as it is, but without better mixing capabilities it can never be a one-app solution for me. Part of the reason I ultimately gave up on Buzz was that I was sick of having to use multiple programs for composition and mixing (the main being its instability).

I think the problem is that current Renoise sends only work as DSP inserts. They’re a part of the DSP chain. They can only be placed within that chain. Hardware mixers and DAWs integrate the sends into the mixer itself. The trade off is that they’re there whether you use them or not. It also imposes some limits. For example, Live can only handle up to 8 auxiliary tracks, and all are automatically available to all tracks; you just have to adjust the send levels. Renoise lets you use as many send tracks as your computer can handle, and you can have tracks using none, some, or all of them.

What Renoise does have is: Access to industry standard Rewire.

Renoise already has the mixer infrastructure in place. Why ignore the possibility to improve what’s there? Rewire is not a good solution because I would need to use it for every single Renoise composition where I want full mixing control. It would be far easier on me (and my CPU) to just do a multitrack mixdown in Renoise and mix with another program. Again, not a one-app solution.

Sometimes: Things are the way they are supposed to be, and when we change them, we destroy their original purpose.

Software is supposed to be whatever the developers decide it should be. It has no purpose other than that. Metaphysics doesn’t belong in computing.

[Cas]

I’ve suggested once that instead of having Mute Source/Keep Source buttons the #Send device could perhaps have a ‘Keep’ slider… if these could be then linked and unlinked in a smart way… That would probably be a relatively simple solution to a lot of annoyances in routing department.

[vvoois (vV)]

@Din

Now here’s what I would like to be able to do:

You can do this by adding a gainer for the level control to the master and tick the “Keep source” button in the send.

@Din

This is actually the best kludge available, but there are some cons. If I were to use a pre-send gainer for all tracks with sends, I’d usually have gainers in all tracks but the bassline. I’d be using the gainers to control levels in all tracks but the bassline. The post-fx mixer faders would be entirely unnecessary in all tracks but the bassline.

You can group the tracks that you want to process this way and add the gainer and send to the group rather than the individual track.

[Din]

@vV

You can do this by adding a gainer for the level control to the master and tick the “Keep source” button in the send.

Sorry, I don’t understand this.

You can group the tracks that you want to process this way and add the gainer and send to the group rather than the individual track.

This only works if all tracks are already volume-balanced before being routed to the group (using their respective mixers or gainers), and when all tracks use the same send amount to the same return tracks. That’s fine when everything uses the same level of same effects. I’ve done this with reverb on drums where I want all the pieces to sound like they’re playing in the same acoustic space. But suppose the rest of my tracks are using different send levels to different return tracks with different effects. Even within a group I would sometimes want to use sends for parts without affecting the whole.

And even where it applies, this solution is exactly the same kludge as before. I use the gainer instead of the post-fx mixer. The post-fx mixer becomes unnecessary.

I’m suddenly curious: What do people use the post-fx mixer for? I don’t really bother doing anything with it myself, but I’d like to know what uses other people find for it. It won’t change my mind about this, but it might give me some insight.

[vvoois (vV)]

@Din

I’m suddenly curious: What do people use the post-fx mixer for?

It is an end mixing fader, one that you cannot automate (by design). You set the ultimate maximum level for that track and can automate the pre-mixer to variate the level between that and the bottom (-inf) level.
There are various people that also want to automate the post mixer but currently are also tied to using a gainer to resolve that part.
The post mixer also turns useless if you put a send in between and do not keep the source audio streaming.
If you want to have your track 0 DB, you could consider use the keep button and use the post mixer for setting the track level to 0db at max instead of using the gainer. The gainer is usually added to raise the volume above achievable levels by the post or pre mixer.

To answer your first question "Keep source"means the original audio from the track not only streams to the send track, but also streams directly to the master (in parallel like your drawing). It will stream with the level of the pre and postmixer settings but disregards the send device level settings.

[wahrk]

@Din

The only reason to adjust the PreFX fader (input gain, really) is for reining in incoming audio that’s too hot for the DSP chain, like VSTi’s with built-in feedback, flange, chorus, distortion/gain, and the like. Also, as I said in my first post, using the pre-fx level screws up dynamics processing DSPs.

Is there a reason you are using the PreFX fader for this instead of the instrument volume level?

You make a point on the dynamics processing, but that’s where I’d just toss a gainer in.

@Din

The problem is in mapping controls to a hardware controller.

That definitely poses a problem for your setup. That is a bummer.

@Din

Even leaving out hardware control, the mixer itself suffers for the lack of this basic functionality. If you use a gainer for any track-level control, why not use it for all track-level control? And then who really needs the post-fx mixer if everyone is using gainers instead?

Many folks do use gainers for primary track-level control a lot of the time. It gives them that added bit of control in between two sets of effects, one set before and one after. If you’re not using sends on a track though, the post-fx mixer works just fine.

@Din

I think the problem is that current Renoise sends only work as DSP inserts. They’re a part of the DSP chain. They can only be placed within that chain. Hardware mixers and DAWs integrate the sends into the mixer itself. The trade off is that they’re there whether you use them or not. It also imposes some limits. For example, Live can only handle up to 8 auxiliary tracks, and all are automatically available to all tracks; you just have to adjust the send levels. Renoise lets you use as many send tracks as your computer can handle, and you can have tracks using none, some, or all of them.

I think this is one of the awesome strengths of Renoise’s sends. You can toss them anywhere in the signal chain on a track. I do this all the time when I want some effects to only affect a portion of the sound. Multiband Send the low end to a send track and then use different stereo effects on the high and low end of the instrument.

@Din

You can do this by adding a gainer for the level control to the master and tick the “Keep source” button in the send.

Then basically use the gainer as your new post-fx gain. It doesn’t solve your controller problem though.

It sounds to me like the main reasons Renoise’s sends are problematic to you are:

You see using a gainer instead of a mixer fader when mixing as a break of workflow.

Is that correct? If so, then while there are plenty of ways to achieve the signal path you want, you won’t get the workflow you want with what’s currently available.

[details=“Click to view contents”]

@Din

Metaphysics doesn’t belong in computing.

Huh? [/details]

[vvoois (vV)]

@wahrk

Then basically use the gainer as your new post-fx gain. It doesn’t solve your controller problem though.

Regarding automatic assignments:no but assumed is one needs a specific workflow so this workflow has to be mapped manually first.
You can customize duplex to your needs. That it currently does not automap basic stuff doesn’t mean it isn’t capable of doing it:e.g. my BCF2000 (which is a hardware controller that is supported the way you desire out of the box) its faders and rotators are assigned to the currently first 8 tracks within the range of the current cursor position, if i go to the left with the cursor and tracks get into scope that are not yet controllable, the new tracks are automatically mapped to be controlled by the same 8 faders.
It would not assign faders to extra controllers like Gainers or send-device levels, but what not yet is implemented doesn’t mean it can’t be altered to behave that way. (Lua has enough notifiers to solve this, but yes, it would probably require a little extra programming)

Perhaps you could try fumbling with Duplex first and see what you do can achieve, perhaps adjust the BCF/BCR mapping to your controller?

[wahrk]

@vV

Regarding automatic assignments:no but assumed is one needs a specific workflow so this workflow has to be mapped manually first.
You can customize duplex to your needs. That it currently does not automap basic stuff doesn’t mean it isn’t capable of doing it:e.g. my BCF2000 (which is a hardware controller that is supported the way you desire out of the box) its faders and rotators are assigned to the currently first 8 tracks within the range of the current cursor position, if i go to the left with the cursor and tracks get into scope that are not yet controllable, the new tracks are automatically mapped to be controlled by the same 8 faders.
It would not assign faders to extra controllers like Gainers or send-device levels, but what not yet is implemented doesn’t mean it can’t be altered to behave that way. (Lua has enough notifiers to solve this, but yes, it would probably require a little extra programming)

Perhaps you could try fumbling with Duplex first and see what you do can achieve, perhaps adjust the BCF/BCR mapping to your controller?

I was just thinking about this as well. It wouldn’t be too hard to map your faders/pan pots to the last gainer in each track.

[Din]

If it’s a matter of tweaking the Duplex script, then I’m up for it. I still think these features are worthwhile (critical, if Renoise is ever to be very useful for mixing), but I’m willing to try a better workaround. Thank you both very much for this suggestion

@ wahrk, yes it is a matter of work-flow. That’s the main reason I got a BCF2000 in the first place. It’s more intuitive, faster, and more ‘real’ than sliding faders around with a mouse pointer. Plus I can move more than one fader at a time, when needed.

Also, I tweak the pre-fx fader more often for a few reasons. First, it’s already visible while I’m working of the Track DSPs. Second, not all VSTi’s output 32-bit floating-point, which means you don’t always want to turn them down using their own main output; otherwise you risk introducing quantization distortion and raising the noise floor on the signal. Third, if I’m using more than one instrument in the same track it’s easier to match their volumes to each other and use the track pre-fx mixer to mind the DSP chain input level.

Finally, I never said that the Sends as inserts was all bad, as I pointed out in my first post in this thread. I’d like post-mixer sends to be added as a new device, or the current sends to be modified to allow being put after the post-fx mixer. A new device might be easier. Of course, neither might be very easy (or even possible) depending on how things are currently coded.

[details=“Click to view contents”]

Sometimes: Things are the way they are supposed to be, and when we change them, we destroy their original purpose.

This is a metaphysical statement. [/details]

Again, thanks for your help. I’ll let you know how things turn out.

[wahrk]

Good luck and don’t forget to share if you get it working!

[kazakore]

@Din

This only works if all tracks are already volume-balanced before being routed to the group

As the Group could very happily have only a single track in it there are not multiple tracks to be volume balanced as you say! You could place each track in its own Group, have the Send Track in the Group and then use the Pre Fader for level adjustments which will affect both the Auxiliary/Sent signal and the kept/dry one.

This would to some extent solve your auto-mapping issue, but in likelihood it would overly confuse matters as you will likely find yourself only wanting to use alternate faders on your controller at best (even less if you use multiple, nested groups.)

As people have said you could always expand Duplex. A routine to scan the track, IF it find the last two DSPs in a chain are a Gainer followed by a Send then map to the Gainer, ELSE map to the Post Fader. Shouldn’t be too hard…

[vvoois (vV)]

@Din

That’s the main reason I got a BCF2000 in the first place. It’s more intuitive, faster, and more ‘real’ than sliding faders around with a mouse pointer. Plus I can move more than one fader at a time, when needed.

You might want to check out this topic as well if you haven’t got the feedback running yet…(applies to the BCF-2000.lua as well)