Odd crossover behavior

[DrNick]

I'm experiencing something I don't understand with the crossover (XVR) block.

Suppose I have XVR at the beginning of my signal chain with all levels set to zero, lows panned hard left and highs panned hard right.

• If I feed this into an AMP block with input set a "Sum L+R" then what I hear doesn't depend on whether I bypass the XVR block or not. This makes perfectly good sense to me.
• On the other hand, suppose I split the signal out of the XVR block into two IDENTICAL AMP blocks with one input set to "Left" and the other input set to "Right." In this situation, I hear a distinct difference when I bypass the XVR block (the highs drop out quite a bit). This confuses me.

Why should these behave differently? I've attached a patch with the latter setup in case you care to try this (and convince yourself that my AMP blocks really are identical aside from the input select - I even copied and pasted one block to the other to double check this).

 

[DrNick]

This is a shameless bump, but I'll use it to try to give my desire to understand this behavior some context.

There has been a good bit of discussion on how to alter the gain structure on different parts of the spectrum (lows/mids/highs). I think much of it was motivated by the Randall Satan amp and its multiple gain knobs with this intent. I thought that using the XVR block might be a natural way to achieve this, so I decided to experiment.

Then I noticed this odd behavior, which seems to point to something that I don't understand. Before I spend too much time mucking around and tweaking, I would like to understand what's going on here.

So... is anyone able to explain (or reproduce) this phenomenon?

 

[dr bonkers]

I found in the amp block you have to select "Left" or "Right" on the amp block itself after the crossover panning in order to do what you want.

 

[DrNick]

I found in the amp block you have to select "Left" or "Right" on the amp block itself after the crossover panning in order to do what you want.

Yes - this is what I'm doing. The lows are panned left, and I have "Left" selected as the input on one AMP, while the highs are panned right and I have "Right" selected on the other (identical, for the purposes of this question) AMP.

 

[jimfist]

Dont know if this applies to what you are doing... Instead, use the multi band compressor with all bands set to 1:1 ratio or thresholds set so that they are not tripped. This will give you control over 3 bands of EQ in one block. FWIW. This won't help with panning, though.

 

[DrNick]

Dont know if this applies to what you are doing... Instead, use the multi band compressor with all bands set to 1:1 ratio or thresholds set so that they are not tripped. This will give you control over 3 bands of EQ in one block. FWIW. This won't help with panning, though.

This would effectively just be a pre-EQ, right? That's one common approach: pre-EQ to cut lows and boost them back after the preamp stage.

I'm curious about really splitting up the signal into bands and putting them into separate amps. But, again, I really want to understand what's up with the XVR, because it seems to be doing something that surprises me.

 

[dr bonkers]

This would effectively just be a pre-EQ, right? That's one common approach: pre-EQ to cut lows and boost them back after the preamp stage.

I'm curious about really splitting up the signal into bands and putting them into separate amps. But, again, I really want to understand what's up with the XVR, because it seems to be doing something that surprises me.

I have a prog bass preset and Snarly Husky Bass for FW 17 posted to the Axe Change that use the Crossover block for bass. These presets may be helpful to see tome different things you can do with the block.

 

[Bakerman]

But, again, I really want to understand what's up with the XVR, because it seems to be doing something that surprises me.

Nonlinear things like amp blocks don't work how you seem to be expecting. Consider how a chord through one amp sounds, compared to each note recorded individually on its own track.

 

[DrNick]

Nonlinear things like amp blocks don't work how you seem to be expecting. Consider how a chord through one amp sounds, compared to each note recorded individually on its own track.

Aha! I think you've hit the nail on the head here. Assuming that the original signal is the sum of its constituents coming out of the crossover (which I believe it is to very good approximation), nonlinearity in the AMP block is precisely what would account for this sort of difference. You're comparing a processed sum of signals to the sum of processed signals. So this isn't an artifact of the XVR block at all, it's an artifact of the AMP block.

Man I feel better - this has been bugging me for days. I'm also slightly ashamed I didn't see this, given what I do for a living...

Thanks!

 

[jimfist]

Sorry I missed the mark on my previous post, DrNick. I made the mistake of killing time on the forum while recovering from anesthesia following a gastroscopy, and let's just say that I wasn't fully "engaged" in reality for a while afterward.

 

[DrNick]

Sorry I missed the mark on my previous post, DrNick. I made the mistake of killing time on the forum while recovering from anesthesia following a gastroscopy, and let's just say that I wasn't fully "engaged" in reality for a while afterward.

It was the drugs!

No, but in all seriousness, I appreciate the comment, and you do raise another interesting approach to underlying goal of structuring the gain differently on different parts of the spectrum. It might even be interesting to employ some actual compression as well instead of going 1:1 everywhere. It's worth playing with. If I do, and anything interesting comes out of it, I'll let you know.