Why isn't there a modeler allowing for usage of longer IRs?

Sascha Franck

Rock Star
Messages
4,008
I absolutely don't get this. Longer IRs open up for tons of creative things, you can pretty much bake your own reverbs and what not.
There's been the Logidy EPSi, which allowed (it's discontinued for whatever reasons...) loading IRs up to 6 seconds length in 2014 already. To my knowledge the only hardware unit allowing for such things.
Then there's, say, my 2008 Macbook, a really outdated laptop, allowing me to run several (!) IR based reverbs simultaneously at lowest buffer sizes. And if it was just about IRs, a little additional predelay on just the wet side of things possibly wouldn't bother anyone too much.
So, why isn't it possible with any actual modeler? CPU-wise, they should all run circles around my old Macbook. Do people just don't like IRs outside of the cabsim realm?
Just wondering...
 
They require more processing power, memory and most of the time are unnecessary.

If you look at the Axe-Fx 3, which can load up to 1.3 second IRs, those "FullRes" IRs are really only useful for room mic simulation type stuff. They sound cool, but there's not a whole lot of space for them and not a whole lot of options on the market either.

Strymon Iridium is capable of loading up to 500ms IRs and again it doesn't manage to sound better. I can't tell a difference between say 2048 sample or ~42ms IRs used on Fractal. You can load the same IRs in Fractal and even vary their length and the result doesn't change a whole lot because the relevant data to reproduce the sound of a close miced cab is in that sub-100ms range.

Meanwhile for reverb, even top price studio units like the Bricasti M7 rely on algorithmic rather than convolution reverb because it's more musically adjustable than painstakingly replicating real world spaces.
 
I think modelers preload IRs on to DSP memory, that is very limit, something like 512k.

Ok - but then, why not just add more memory? It's rather cheap in comparison to other components.

and most of the time are unnecessary.

I beg to disagree here (I think I'll start an "IR fun" thread in the recording sub-forum...).

And fwiw, I'm not talking about longer cab IRs or whatelse - but about IRs used for all kinds of wicked spatial FX, resonant adventures and what not.

Meanwhile for reverb, even top price studio units like the Bricasti M7 rely on algorithmic rather than convolution reverb because it's more musically adjustable than painstakingly replicating real world spaces.

Well, for studio usage it's not all *that* relevant anyway as there's IR loading reverb plugins galore.
 
Longer IRs would possibly benefit bass guitar but maybe not even then. The effect of a shorter IR for guitar is lower resolution in the low frequencies. The shorter the IR, the less room info also.
 
Well, given how many people just drool about their Strymons, Eventides and what not, I'd expect that number to be vastly larger.
I think sub-1% is about right. People drool over Strymons and whatnot not because they do specialty fx but because they do really good versions of staples.

Those frogs do sound really cool tho.
 
I got some other IRs with mixed in shepard-tone-alike stuff, quite interesting too, you get wicked moving resonances and what not.
 
And in case your IR loader is offering some minimal tweaking options (such as an envelope and EQ), you can get extremely far with just a few baseline IRs.
 
I don't think those folks know how, or want to, make there own weird IRs.

Seriously, it's as easy as it gets. See my examples above. Record anything that has any kind of a mixture between noise, movement, "bubbliness", drag into IR loader, adjust envelopes (basically the same as adjusting decay on your reverb unit) and off you go.
Plus, there'd be a huge aftermarket (just as with cab IRs) after a while. Heck, I could supply a metric ton of IRs in a few days. Add to this that there's plenty of great free ones around, some pretty interesting ones among them (caves and such).
It's really not more of a deal than hunting down a few cab IRs.
 
Seriously, it's as easy as it gets. See my examples above. Record anything that has any kind of a mixture between noise, movement, "bubbliness", drag into IR loader, adjust envelopes (basically the same as adjusting decay on your reverb unit) and off you go.
Plus, there'd be a huge aftermarket (just as with cab IRs) after a while. Heck, I could supply a metric ton of IRs in a few days. Add to this that there's plenty of great free ones around, some pretty interesting ones among them (caves and such).
It's really not more of a deal than hunting down a few cab IRs.
However easy that may or may not be, it's not something the vast majority of guitar players care to think about, much less do.
 
How do you know when there's no options doing so?
Because most guitarists have the tools to easily capture quality IRs of their own speaker cabs, yet...they go and post a new thread on an internet forum asking who makes the best C. Rex IR, why there aren't any commercially available IRs of Weber speakers, etc., rather than just capture an IR of their cab.
 
How do you know when there's no options doing so?

Convolution reverbs in pedal form already exist. Go get a poly digit.

I think it’s a niche market that only a small handful of users care about. The digit got a fair amount of press from influencers as it started to make the rounds, the whole guitar market didn’t start freaking out and demanding convulsion reverb because it was so much better.

I don’t think the modelers will have it generally unless it becomes a widely adopted and coveted feature. If you look at how a lot of people wanted to use the digit, it was to get convolutions of famous/expensive hardware units more so than physical spaces (which if I made a modeler would tell me to focus more on nailing those sounds instead of trying to solve for convolution reverb).

D
 
Back
Top