IR Capture & Aliasing Experts, Advice Needed

[Guitarcal]

Needing some advice and direction.

I primarily use Fractal, but I have a trusty Pod X3 Live and Pod Farm I will use on occasion. There is a specific cab, the Greenback 25 that I think sounds killer and works well in my mixes. I would like to capture this as an IR for the Axe FX and other modeling products.

I plan to capture via the sine sweep method / convolution plugin and also will capture utilizing Fractal's IR capture utility built into the Axe Fx.

1) If I capture by routing signal through the Pod X3 Live (all amp and effects disabled except the cab), will this also capture any aliasing the Pod X3 Live might introduce? Is aliasing at play with the amp block disabled for modelers such as the Line 6 Pod or Fractal Audio, or is aliasing only introduced when amp modeling is enabled?

2) If I capture by routing signal through POD Farm (all amp and effects disabled except the cab, will I also capture additional aliasing the POD Farm plugin might introduce?

3) For POD Farm, if I instead utilize the AD/DA converters in my Apollo Twin over the Pod X3 live, should I expect better results? I assume the converts are better in the Apollo Twin.

Thanks!

 

[Sascha Franck]

While you should not get any aliasing captured with (1) or (2), as aliasing usually is a result of some signal processing (whereas you'd only use things as converters), but anyway, the most important part would be: Are we talking about the plugin version of Pod Farm and the cab being present there as well? Because in that case, capturing an IR would be as trivial as it gets, no sweeps and deconvolving machinery required at all. You'd just place a dirac spike on a track with just the cab of the plugin running, bounce things down and possibly trim the resulting bounce accordingly. Done.

 

[Guitarcal]

While you should not get any aliasing captured with (1) or (2), as aliasing usually is a result of some signal processing (whereas you'd only use things as converters), but anyway, the most important part would be: Are we talking about the plugin version of Pod Farm and the cab being present there as well? Because in that case, capturing an IR would be as trivial as it gets, no sweeps and deconvolving machinery required at all. You'd just place a dirac spike on a track with just the cab of the plugin running, bounce things down and possibly trim the resulting bounce accordingly. Done.

Thank you, yes, I am using the Pod Farm plugin. Glad to hear aliasing will not be at play here.

Interesting, I have not heard of a dirac spike, I'll do some research.
Am I headed in the right direction using this link?

Homepage noisetime - How to create impulse responses

Homepage noisetime - How to

www.noisetime.com

 

[Sascha Franck]

Am I headed in the right direction using this link?

Absolutely!

As said, it's pretty trivial:
- Adjust guitar amp plugin (or rather the cab section) until it sounds great.
- Bypass anything but the cab.
- Place the dirac file on the track.
- Solo track and bounce.
- Trim and normalize.
- Load file into target device.

I've done that several times to consolidate IR mixes, to grab IRs from non-IR-ed cab sections (such as from the Helix Native plugin) and such.

 

[Sascha Franck]

Fwiw, the Zip-file on that site is missing a 48kHz dirac. You might however want to work at 48kHz as that minimizes the changes for further sample rate conversions (from all I know, the Fractal units are running at 48kHz internally). I converted one of the diracs of that very site (I'm using them since ages) to 48kHz for that reason, but it shouldn't be much of an issue as either your DAW or the Fractal editor would otherwise do the required conversion in case you used 44.1 or whatever.
Whatever, I simply added 48kHz diracs to that folder, just to be complete.

 

[MirrorProfiles]

Several second of a sine sweep has always sounded much better to me than using a dirac spike.

You shouldn’t have any concerns with aliasing for this set up.

 

[Sascha Franck]

Several second of a sine sweep has always sounded much better to me than using a dirac spike.

It should not matter as long as you stay strictly in the digital domain.
I once compared and things pretty much nulled.
Of course it's a completely different story once there's any analog signal paths involved.