Would it be possible?

[tubetonez]

I am wondering if this could be possible. Say you run guitar cabs loaded with EV12L's or other fairly neutral speaker. Let's say that you already have a custom IR of your cab for FOH, splitting off your stage amp before the cab block. Suppose that you use RTA, SMAART or whatever to generate an EQ curve for flattest response from your cab. Then you take an IR in wav format of a different cab that you like, say a Greenback loaded one. Using a software plugin or whatever apply the EQ curve from the RTA to the wav file, and resave it as a new cab. Then you use two cab blocks, with the FX loop in between. The last duplicates your stage cab, and the first adds the EQ'd versions of other cabs. I intend to try this, just wondered if anyone had considered anything similar. Perhaps there could be some way load both IR wav files, of your stage cab and the cab you wish to model, into a software editor and do some kind of additive/substractive manipulation?

 

[Brian G]

Huh? Maybe it's just me not understanding this.

Could you perhaps clarify exactly what end result you're trying to get to here?

Is it that you want to "neutralize" your speaker / cab, so you can use other cab IR's without them being colored by your real cab?

 

[tubetonez]

Huh? Maybe it's just me not understanding this.

Could you perhaps clarify exactly what end result you're trying to get to here?

Is it that you want to "neutralize" your speaker / cab, so you can use other cab IR's without them being colored by your real cab?

Yes, so that you can use other cab models through a PA/guitar cab rig. EQ alone may not do it, but there are myriad ways to manipulate digital audio. It seems to me that one should be able to tailor the response of an IR for a specific cab setup. If an IR can make an electric guitar sound like an acoustic, I'd think it should be able to make an EV sound like a Celestion. The reverse might not work, if you're running cabs with speakers that break up early you couldn't add clean headroom to them.

 

[Brian G]

Given the sheer number and magnitude of the variables involved, I'm going to go out on a limb here and say that what might ultimately be possible is not the same as what's practical.

I've had some (reasonable) success in measuring, and then EQ'ing my guitar cab/speakers in such a way that I could then use the Axe's IR's. But my goal was only to apply EQ to compensate for the gross deviations of my cab, so that the IR's would then work more or less as intended. Put another way, I used a PEQ to flatten major peaks & valleys in the guitar cab's response to make it "pseudo-FRFR".

Don't think this is what you have in mind, though.

 

[Clawfinger]

There could be a way. A guy at guitarmodeling.com (Alu) did a nice little test (that I didn't find now), where he used a reamped guitar track played through three different speakers, and deconvolved one cab, with another cab as "sweep". The result was an IR that, when applied to the Marshal recording, sounded almost identical as the Mesa.

I'll try to dig it up.

 

[tubetonez]

Given the sheer number and magnitude of the variables involved, I'm going to go out on a limb here and say that what might ultimately be possible is not the same as what's practical.

I've had some (reasonable) success in measuring, and then EQ'ing my guitar cab/speakers in such a way that I could then use the Axe's IR's. But my goal was only to apply EQ to compensate for the gross deviations of my cab, so that the IR's would then work more or less as intended. Put another way, I used a PEQ to flatten major peaks & valleys in the guitar cab's response to make it "pseudo-FRFR".

Don't think this is what you have in mind, though.

I think it was your experiment that sparked the idea. Very precise EQ could be done via software, but would that be enough?

There could be a way. A guy at guitarmodeling.com (Alu) did a nice little test (that I didn't find now), where he used a reamped guitar track played through three different speakers, and deconvolved one cab, with another cab as "sweep". The result was an IR that, when applied to the Marshal recording, sounded almost identical as the Mesa.

I'll try to dig it up.

That sounds promising. Definitely something that I'd like to experiment with. There are so many exciting new innovations in software design, Cliff's being most welcome I'm fascinated with the Melodyne stuff, if you haven't seen the demo check it out http://www.youtube.com/watch?v=jFCjv4_jqAY
This may not be useful for editing IR waveforms, but it sure is cool.

 

[schnarf]

An impulse response is a very precise EQ.

Basically what you're asking is if you can generate an inverse filter for the guitar speaker/cabinet. This depends on whether the system is minimum phase, which in oversimplified terms is whether its frequency and phase response are related in a nice way. If they are, this can be done. If it's not minimum phase, which I suspect it might not be, you will not be able to perfectly invert it, though you can still get close.

You still have to remember that guitar speakers are highly directional (Jay Mitchell is much more knowledgeable about this than I am), so you would only be able to invert that speaker's response from one listening position. It might sound really bad from other positions.