Hey,
this method of using separate aux track for reverb is great - effectively minimizing resource utilization.
However as much as I would like to put it to use, it's not easy considering the tradeoffs, for which there might be workaround, but I just can't see it.

Let's say one instrument occupies one track. Altiverb 7 is used as convolution engine, with only one IR loaded per project.
To position the instrument correctly in the spatialization field, just panning won't do, as there's no knob for depth and height and also the pan must be sent pre-FX into true stereo engine to get good results.
What might work is processing the "direct" for each instrument separately and then sending the result to one reverb track which processes only ER and Tail, though not sure how to generate directs without loading Altiverb on every instrument track as before.

How do you handle this in your projects?

 

Apparent height differences you can get with the amount of high vs low freqs. The depth with amount of reverb, wet vs dry signal, but also with dynamics/timbre of an instrument, as in... if the instrument plays loud, with higher dynamics, but is quiet in the mix, it's percieved as if it's far away.

But what you are asking, I am guessing, is regarding panning/positioning the instrument in the stereo field. And that depends of the source signal, is it mono or stereo. For mono sources, you can pan your dry signal, and depending what DAW you use, do the same via separate controls for panning your sends (there is such control in Cubase, not sure about other DAWs). If not, you have to somehow pan the source pre send, obviously. For stereo sources, you can use plugin like waves S1, in cubase it’s called Stereo Combined Panner, and move around the whole field, both channels, narrow them down, or widen.Or you can delay one of the channels, the result our brain perceives like if the source's comming from one side.

Ofc, the best way to go is with prepositioned samples, like modern orch libraries which already have correct placement in the sound stage. But yeah, mixing other instruments, mono or stereo centered, can be tricky.

Hope I understood you what you are asking.

 

Last edited: Sep 24, 2017

Yep, that's great shit isn't it.

The problem with only one reverb could be the depth. If you want to place different instruments in different depth (distance to the listener) you, well, have to use differnt reverbs if you want to do it properly (because of the at least different predelay times).
You can cheat it with different amounts of sending levels and there surely will be situations where the decent listener wouldn't realize that, but that's nasty cheating.
I know people doing the classical stuff who always use three reverb tails and three early reflection plugins as send FX and even sometimes "blend" between two of them (means sending into two with different levels).
If you want be ultra realistic you would need one of those per "depth position" but that's not necessary in my opinion, but there might be others.

On the other hand there's a plugin with which you can position your instruments from left to right and front to back. Didn't ever work with, but some say it does it's job. Called Parallax-Audio VirtualSoundStage now aviable in version pro v2.0.

Hope this helps.

 

Last edited: Sep 24, 2017

create a buss, or aux which you send every track to, add your reverbs then make it 100 percent wet, then lower the volume of said buss to -20 db rms on master fader reading solo'd,
then you will have a master reverb channel with its own volume ,eq comp anything else you want to add to it , works best with stereo expansion and low volume compared to master track.

 

If I understand your question, you can use an ordinary send/return loop, you just need to make your sends post-fader, so the signal is sent to the reverb aux with the same pan signature as the dry track. If you want a pan position independent of the dry setting, it depends on your DAW - Pro Tools has separate pan settings built into the send controls. If your DAW doesn't, you can pass the signal through a separate aux just for panning, and send that aux into the verb return aux. If you're doing it for a lot of tracks it can get messy, but once they're set you can hide the panning auxes and forget about them.

 

OK, it seems rephrasing the original question is necessary.
Currently I use "Stage positioner" in Altiverb to set the correct position of the instruments, and Altiverb is currently loaded separately on every instrument track, as it has to be tailored for each individual instrument.
Apart from higher resource utilization it seems to be working very well, yet to optimize the project more, and free up resources for more tracks a workaround is desirable.

So I need to achieve similar results with dedicated reverb sends.
The instruments are recorded 100% dry with no additional reflections.
Altiverb is set to 100% wet.

The way this Stage positioner works is by modifying channel balance (not necessarily pan), channel delay and IR predelay (at least from what I gathered).
VSS is not directly comparable, as VSS does not process tails only ERs and Direct.

So far I can't think of any simple solution besides capturing Altiverb result using Impulse Response waveform and then reapply the IR file using different, more effective engine (ReaVerb ?).

 

i like bricasti haven't used altiverb in long time but i remember it was great

 

use as many reverbs as sends as you like. or as inserts. if global reverb doesn't serve you , use them as inserts. also, compress and eq the auxes or even deess/gate/distort/saturate/flange

 

Last edited by a moderator: Sep 25, 2017

yeah i like to def have different reverb each track and most time many reverbs each track but all at inaudible levels at least not overtly obvious. i like my reverbs and delays etc. to be 20 db rms quieter than reg signal.

 

I would record the result of the reverb and the placement from each instrument to a new audio track i would do this for every instument that need the chain fx you described, it can be tricky in the mixing process but keeping a copy of the original instrument is a good idea! This way you can free up resources but you are limited to some automations.

 

Yea, like @korte1975 says, use more reverbs. Usually three are enough, one room, one hall and one sort of a fx reverb like the blackhole.
Combine them or if you run out of cpu power, process plugin, you can do that in cubase.
@shomyca also gave good advice.

 

I do this too

Really? I've found it works fine. For depth, there's the send level, as for pan, there's pre & track pan. To get good localization, maybe less verb the better.

I don't have an expensive convolution verb, rather, just a cheap algo verb.

 

Every once in a while a new member comes that knows his shit. Glad to have you here.

 

Didn't know altiverb in practice myself so I visited their homepage and rtfm.

So, cool feature, hofa got the same and is excatly what

is supposed to do with only one instance, like I, ehm, mentioned before my dear friend . Still I don't know if the quality suits you.

If you wanna stick to high quality reverb like altitude, see if you can break down unlimited positioning concerning the depth to maybe three. In other words ask yourself if you really need to have so many different positions in the deep field.

If you're fine with for instance three positions (front, half back, full back) you will then only need three altiverb instances as send FX (until now you used it as an insert!) with different parameter setting according to the depth they should picture. The more away the shorter the predelay. Clear to here?

Then you only need to adjust the send levels in the different audio channels [for instance singer to front verb, drums to half back and choir to full back (cabasa completely outside, no, just kidding)] and violà ;-) you got them where you want to.

But be advised fine tuning the parameters of the, in this case, three different reverb instances can be time-consuming.

And to the stereo position: Frubbs explained it very well, nothing to add.

 

Last edited: Sep 24, 2017

If you wanna use not as many altiverb instances, you have to ditch altiverb's placement... not sure if anybody's using it like that. For post...sure. But for music, not recource friendly way as you already realized. Sends are the best way, multi verbs, as No Avenger described even better. Panning before sends... VSS 2 maybe...etc...

 

Oh, thank you and like Castle's mom said "right back to you, kiddo"

 

There is no such thing as a bad question, sadly that isn't always the case with answers!

 

Use Altiverb with a long tail on a return track and stock Ableton reverb with a short tail on each bus that you need, as they are very low on resources in comparison,

 

Last edited: Sep 25, 2017

I'm not sure if I know what you're meaning. I don't believe that altiverb changes that.

That you can easily do yourself with every track or even part if you want to according to sonic speed of 34,32cm/ms.

 

The reason I opted for such setup is because I'm working with Sample Modeling libs, which are notoriously "difficult" to get back to life in convoluted space. None of the other solutions proved to be satisfactory (tried with B2 and Valhalla) - the sound has to come through the IR being 100% wet and has to be modelled through IR's Direct.
Basically it must be 100% wet at all times.

Nevertheless, thank you all for your responses - I'll start experimenting with fewer 'collective' instances of Altiverb and move on from there depending on results.

I meant channel balance for Direct, not for the rest of IR.

 

Last edited: Sep 25, 2017