Concerning Reverb Inserts and Sends

B

bdam123

Member
So I've been thinking about reverb a lot lately and was curious about different techniques in general so I decided to run some experiments. It's common to advise producers and engineers to send tracks to one reverb instantiated on an aux tracks for the sake of CPU but I was curious if this technique actually sounds different. My experiments go as follows:

Experiment #1

Step 1) Two tracks (one vocal and one guitar) each with an identical instantiation of a reverb (set to 100%). Bounced.

Step 2) Those two tracks (100% send, pre fader, with tracks on mute ) sent to an aux with same reverb. Bounced.

Step 3) Conduct null test

FAILED

Experiment #2

Step 1) Two tracks (one vocal and one guitar) each with an identical instantiation of a reverb (set to 100%). Bounced.

Step 2) Those two tracks outputted to an aux with same reverb. Bounced.

Step 3) Conduct null test

FAILED

Conclusion

Reverb behaves differently when being sent audio in different ways (directly instantiated on a track, via sends, via output).

Is there something going on behind the scenes thats causing this?
 
bdam123 said:
So I've been thinking about reverb a lot lately and was curious about different techniques in general so I decided to run some experiments. It's common to advise producers and engineers to send tracks to one reverb instantiated on an aux tracks for the sake of CPU but I was curious if this technique actually sounds different. My experiments go as follows:

Experiment #1

Step 1) Two tracks (one vocal and one guitar) each with an identical instantiation of a reverb (set to 100%). Bounced.

Step 2) Those two tracks (100% send, pre fader, with tracks on mute ) sent to an aux with same reverb. Bounced.

Step 3) Conduct null test

FAILED

Experiment #2

Step 1) Two tracks (one vocal and one guitar) each with an identical instantiation of a reverb (set to 100%). Bounced.

Step 2) Those two tracks outputted to an aux with same reverb. Bounced.

Step 3) Conduct null test

FAILED

Conclusion

Reverb behaves differently when being sent audio in different ways (directly instantiated on a track, via sends, via output).

Is there something going on behind the scenes thats causing this?
Click to expand...
That's interesting. In experiment #1 you mention your pre-fade send is set to 100%. Does that mean the pre-send fader is maxed or is it set to digital unity? I've had some battles over the years with Pro Tools aux send faders not being accurate. We use them to bounce stems and levels are critical for broadcast standards. I always double-check what was sent.

My guess, and it's only a guess, is there's latency being introduced somewhere. If so you could nudge (by milliseconds) to see if the comparisons nulled at a particular value.
I'd be interested in which DAW is doing this?
 
Can you explain this. Instantiated is a term new to me and I don’t know why you’d instruct the producer or engineer to do this? I’ve never heard of this process?
 
it takes a while for the reverb send to go to the track, and return to the FX bus...
there are many ways to route, and they all have tiny differences that can add up at an incredibly small scale.
big picture stuff is, the mix sounds good/bad.

now, if you sent one track on a level of 4 (out of 10) and another track at 5 out of 10, those two tracks wont be processed exactly the same way.

in the end, none of that matters, except how the mix sounds.
whether or not you can get similar tracks to null isn't the goal.
 
rob aylestone said:
Can you explain this. Instantiated is a term new to me and I don’t know why you’d instruct the producer or engineer to do this? I’ve never heard of this process?
Click to expand...
You've never heard of sending multiple channels to a single reverb bus?

Anyway, the thing the OP forgot to do was run the tracks to a single reverb the same way twice and perform the null test. It will probably fail that too. Reverb is one of the types of effect that have enough randomness and timing differences that it will never null out.
 
Farview said:
You've never heard of sending multiple channels to a single reverb bus?
Click to expand...
It blows my mind to this day that people use additive effects in-line. If I had a single source and I wanted to add reverb (or any time-adjusted effect) to it, it'd be on an aux send. And that send would probably be compressed to some extent, band-passed (pre-effect) almost certainly, potentially EQ'd on the way back. If it was a delay on a vocal or guitar, all that PLUS the return would be sent to the same reverb aux as the source (to put it in the "same space" as the source).

Chorus on a bass...? Maybe. Flange on a vocal or percussive source...? Potentially. But something that's *added* to the source as opposed to something that's intended to *change* the source? It's going to be on an auxiliary send 100% of the time.
 
Oh, is that was going on? I’m afraid I lost the plot with the explanation. Where has ‘instantiated’ come from. In a lifetime of recording it’s a word I just don’t have in my vocabulary. Reverb for me is a post fade send out and either a dedicated return or a channel input or two. Maybe occasionally an insert loop. What is the null test, testing? Never done that either, unless it’s another new way of doing an old thing?
 
Instantiated is a fancy way of saying "an instance" of a plugin. When you add a plugin to a bus, you instantiate it.

A null test is when you take two different files of the same thing, invert the phase of one and add them together to see if they are the same. If they are the same, you will hear nothing. If they are not the same, you will hear the difference between the two
 
I knew what he meant when he said "instantiated"--yes, what is meant is "an instance of." Kind of unusual way to say that, but it is correct. Anyway, I don't get the whole experiment here. My advice is to not worry about the ""null" business and put the reverb on an AUX return, at 100% wet (here's where you might want to experiment with the wet level, but on reverbs I've always got them at fully wet). Then use the AUX send on the vocal and guitar to send to the reverb, adjust the reverb to where you like it in the mix, and then you're set. By all means dial down the actual return fader, or not, and massage the send level, or not. I don't know who might advise someone to save CPU resources by cutting a corner in this day and age. I use 10 AUX busses, with the same two plugs on them--one reverb, and one delay; sometimes, on AUX 10, I'll put a special effect plug, but you get the picture. I don't ever think about CPU because why?
 
Massive Master said:
It blows my mind to this day that people use additive effects in-line. If I had a single source and I wanted to add reverb (or any time-adjusted effect) to it, it'd be on an aux send. And that send would probably be compressed to some extent, band-passed (pre-effect) almost certainly, potentially EQ'd on the way back. If it was a delay on a vocal or guitar, all that PLUS the return would be sent to the same reverb aux as the source (to put it in the "same space" as the source).

Chorus on a bass...? Maybe. Flange on a vocal or percussive source...? Potentially. But something that's *added* to the source as opposed to something that's intended to *change* the source? It's going to be on an auxiliary send 100% of the time.
Click to expand...
Ha! :)

I dunno if it rises to the level of "mind-blowing", but I would agree that for time-based effects, a simple, fundamental, aux send and return is base-line. However, I would submit for many things in my world, it is essential to "insert" time-based effects on a track-by-track basis. It's uber-popular these days, particularly for ambient, film, and television scores, to create "rounds" using 100% wet reverbs inserted into a track. Further, whether you're in Cubase, Logic, Pro Tools, or Studio One there is a way to temporarily freeze the track with the ability to retrieve it later. The current go-to reverb is the Strymon Big Sky reverb and if inserted at 100%, all sorts of cool things happen. I saw a score the other day with 3 Big Sky's chained on one track and it sounded immense.

Granted this is not an everyday and or common use of time-based effects but I do believe it's at the very least, worth mentioning.
 
You must log in or register to reply here.
Back
Top