Delay before Reverb? Or Reverb before Delay? What's your preference?

How do you guys who run them parallel deal with the volume jump when you bypass them? Or do you never bypass them.
 
It isn't about Fractal code, it is about the way sound works.

It's math.

Do you find it weird that 5 x 1 x 9 == 45 and 9 x 5 x 1 == 45?

What Cliff said is that for effects that are LTI the sound is commutative. Same thing as the math above.

To take your example, 3 x 50 == 50 x 3. Simple ;)
Math and sound are not the same thing so I can’t agree with that argument - only that the resulting sound will contain both components but it will most definitely sound different depending on their order. I do understand your math example but I don’t understand how it relates to an audio chain. I will concede that code can be written to where it doesn’t matter but that would be weird for those of us who are used to how an audio chain works in the real world. I feel like there is some misunderstanding of terminology happening here and possibly of Cliffs explanation although I dont want to put words in his mouth. Also this thread is older so does this even still apply at this point?
 
Last edited:
I agree that math and principles are used but that doesn’t mean that 2+1 and 1+2 will sound the same. Math is not sound.
 
How about a practical example?
if I put an eq after a reverb and cut all frequencies above 250 hz will it effect the sound (frequency response) of the reverb?
 
I agree that math and principles are used but that doesn’t mean that 2+1 and 1+2 will sound the same. Math is not sound.
Our world, including all the sounds in it, is governed by the laws of physics. The technology you rely upon to make music is a product of that immutable relationship. I sense you might not be comfortable with science intruding on art, but the fact is there is no issue to be argued or disagreed with here, because the proof is all around you. The very existence of the Axe-FX is proof. In particular, linear processes are commutative (depending on how you frame the question, time invariance doesn't matter). If you have any doubt, I'd suggest, instead of posting a reply to the contrary, first try swapping an eq and delay and you'll see it makes no difference.
 
Rereading the whole thread, I find it weird that people find Cliff's explanation cool. Nothing personal, but doesn't the way the code work rob us of a creative decision?
AFAIK, a reverb is a series of many delays. So how should a 3-tap delay going into a 50-tap (wild guess for reverb) be the same as doing it the other way around? Hard to believe this would be the same with analog devices, and thus equally hard to find it cool how it currently works...

No, that is just how it works. Unless one of them is clipping or otherwise nonlinear, it isn't different.

It's the same with EQs unless they're doing something fancy (like saturation) and a bunch of other things. But, it is the whole chain.

Delay > Drive > Reverb is NOT the same as Reverb > Drive > Delay.

Again, you can prove it to yourself by recording both versions off a looper (or do something similar in your DAW) and using ABX tester.

Honestly, getting used to how to use ABX tester (with appropriate level matching) is one of the things that will let you figure out the kinds of things that do or don't matter, even if it's often detailed/complex to actually make the comparison fair (e.g., it's insanely difficult to ABX two different cabs or two different speakers correctly).
 
No, that is just how it works. Unless one of them is clipping or otherwise nonlinear, it isn't different.
So, in other words, let's take a name that is known for using one delay into another, like Joe Satriani (not saying all these guys are necessarily tech geniuses) ...

If you would tell Joe, "Hey, just switch out both delay pedals — it will fit your board better," he should reply, "Oh, waw! Yeah, no problem. It sounds pretty much exactly the same anyway"...?

Seems I'm not the only one that can't wrap his mind around that... :oops: If one day I try this with a few pedals, I will be utterly stumped... — but this thread gave me enough reason to consider it, thanks :)

OOPS: We're talking reverb & delay here, not two delays, so I'm giving a wrong example?
 
Last edited:
If you would tell Joe, "Hey, just switch out both delay pedals — it will fit your board better," he should reply, "Oh, waw! Yeah, no problem. It sounds pretty much exactly the same anyway"...?
Just because someone is famous, respected, etc doesn't make them infallible or immune to perception bias.


OOPS: We're talking reverb & delay here, not two delays, so I'm giving a wrong example?
No... Both cases are the same (again, with the listed caveats about distortion and modulation).
 
Just because someone is famous, respected, etc doesn't make them infallible or immune to perception bias.

It might actually be even worse for well-respected experts in some contexts. If you're good at something, it make sense that you're probably right about a lot more things than you're wrong about....and that feeling will stay even if the things you're right/wrong about don't matter to how good you are. IDK....it's hard to explain.

OOPS: We're talking reverb & delay here, not two delays, so I'm giving a wrong example?

Delay is perhaps easier to understand, but it actually can be a little different depending on the order if the delays aren't the same.

First, let's look at the delay times. If you just click the strings by hitting them while you're palm or left-hand muting instead of sustained notes and you have a 100ms delay followed by a 150ms delay, each set to give 3 repeats.

With only the 100ms delay, you hear clicks at: 0, 100, 200, 300, and then it stops.

With only the 150ms delay, you hear clicks at: 0, 150, 300, 450, and then it stops.

With 100ms > 150ms delays, you hear clicks at:

Original sound: 0
Shorter first delay, from the original sound: 100, 200, 300
Longer second delay, from the original sound: 150, 300, 450
Second delay, from the first delay's first repeat: 250, 400, 550
Second delay, from the first delay's second repeat: 350, 500, 650
Second delay, from the first delay's third repeat: 450, 600, 750

So, the full set of what you hear is at: 0, 100, 150, 200, 250, 300(x2), 350, 400, 450(x2), 500, 550, 600, 650, 750

If you reverse the orders of the delays, you hear:

Original sound: 0
Longer first delay, from the original sound: 150, 300, 450
Shorter second delay, from the original sound: 100, 200, 300
Second delay, from the first delay's first repeat: 250, 350, 450
Second delay, from the first delay's second repeat: 400, 500, 600
Second delay, from the first delay's third repeat: 550, 650, 750

So, the full set of what you hear isat : 0, 100, 150, 200, 250, 300(x2), 350, 400, 450(x2), 500, 550, 600, 650, 750

At least as far as the delay times go, they're the same.

The first few repeats will sound different if, for example, the delays themselves sound substantially different. If one of them is significantly darker than the other, for example, then which exact repeat times are bright vs dark will change. In the way I listed the times above, the last 3 lines of each ("Second delay, from the first delay's....") will have the effect of both delay sounds (e.g., not filtered, then dark, sounds dark...or dark, then not filtered, sounds dark). But, the times listed in lines 2 and 3 (each delay from the original sound), will swap which ones are not-filtered vs dark, depending on which one is dark vs not-filtered.

e.g.,

If your 100ms delay is the dark one, then the sounds at 100, and 200 will be dark.
If your 150ms delay is the dark one, then the sounds at 150 and 450 will be dark.
Because 300ms comes from both of them (in this example), it'll be in-between because it's pretty literally half-filtered. It'll also be louder than the other repeats.

All of this assumes perfect clocks.

Swapping one of the delays for a reverb makes the times a lot more complex to actually list...but the same thing basically happens.

The "tone" things (which I exemplified by saying dark), are probably easier to understand with a Delay and a Reverb if you really try to imagine/visualize it. But, because each of the reverb and the delay have their own tones and they "combine" more simply, you don't get that weird thing with the first few repeats changing tonality.

You'll get similar changes in sound from changing the order if one of the delays is modulated or has a broken/flawed clock, but that seems like it makes it not time-invariant (another necessary condition for the order not mattering). Which means that it also wouldn't be the same for a modulated reverb or something like Shimmer reverb. Again, those are not time-invariant effects.

Hopefully, that all made sense.
 
Hopefully, that all made sense.
Since no one is attacking it, I guess it might ;) TY for the effort — I will add it to my notes.

I guess in the end the most important thing for me is probably that there is no downside or problem doing things in the way I intuitively feel is most correct, and that playing with the order could also be worthwhile or may have its benefits in some cases.
 
I am going to do some testing to see for myself but I believe you. What should I study to learn more about this phenomenon?
 
Back
Top Bottom