Any recommendations for PC based IR capture software?

You can use Audacity (free) for capture if you like.

My way:
I use a 20sec 48k/32-bit Float log Sine Sweep. I load that onto its own mono track in Cubase (6.5 currently) in a 32bF/48k session. I set up four more mono audio tracks; one to test mic placement (input is from the mic), one to capture a DI (amp block ONLY) from the Axe, one to capture that exact same Axe patch through a neutral power amp + cab/speaker(s) being captured, and a final one for the Sweep capture. In addition to the sweep, I have an inverse of it, which is what I use for the deconvolution step.

With that session as a template, I can open Cubase and load it up easily, save as a new IR capture run, and go to work. I keep notes in the session regarding mic placement, etc. Cubase also allows me to record to the same track with each pass being a new take (and numbered to reflect this). Other DAWquencers offer this; I like Cubase and Pro Tools. With each track labelled to reflect the source.. such as, "Mic Test," "Axe DI," "Axe Cab," "Sweep Cap"... and with my notes, I can easily return later and know what I did. I will then come up with a patch, get the volume where I want, and start with mic placement. I enable the "Mic Test" track, and start recording and playing back, searching for a good placement; in my current situation, this is easier than having someone play while someone moves the mic until I hear a placement that sounds good to me. When I find a placement that sounds good "on tape," I will record a take on the "Axe DI" and "Axe Cab" tracks. This allows me to return with the resulting IR later, load it into an IR plugin, and compare the DI + IR track to the appropriate Cab + Mic track; IME, if done properly, these two will match very, very closely, almost indistinguishably.

Mic placement set, I make notes, repatch (I use my patch bay) an out from my ADC/DAC I/O box to the input of the same neutral power amp at the same level setting, mic still in the exact same place.. I then solo the Sweep track, and record a pass to the "Sweep Cap" track that exceeds the 20sec sweep length by a small amount. By following this recipe, I now have Axe DI, Axe Cab, and Sweep Cap take numbers that correspond to one another!

Next, I take my Sweep Cap takes, process them in Octave (using the Inverse Sweep I mentioned previously), and then either use the resultant trimmed (based upon the inverse sweep length) files, or use Audacity to trim them manually where I would like. Then I take those files and use Axe-O-Matic to convert them for use in the Axe-Fx II. With Axe Edit, I can then name, transfer, etc. Works great as an editor/librarian for me! Personally, I prefer to limit my IR use to plugins rather than the Axe; please note, though, that I no longer use the Axe for any amp (or front end) duties. For someone opting for Amp Block use, perhaps with an FRFR system, or DI recording, etc, the Cab Block plus your own IRs can be fantastic, and the extra Axe features afforded can be a plus IMO.

Since Octave is a pretty "deep" app, I would also suggest Voxengo's Deconvolver. Oddly enough, I had problems processing captures using sweeps NOT generated by Deconvolver. Unless one desires "deeper" options and versatility, I think that Deconvolver should be just fine. I would suggest using a sweep generated by Deconvolver for your first tries. Maybe someone else got non-Deconvolver sweeps to work with it?
 
Thanks nikki,

great info!

What is the math behind deconvolving the sweep with an inverse vs. the original sweep?

And MATLAB / Octave is not beyond me. I'm just looking for something to approach the quality of the JM far field captures.

Do you apply a windowing function in Octave or does Axe-o-matic or does the AxeFx iteself window the IR in the cab block? Or none of the above :)

Do you tilt the cab so the speaker center is spot on the mic capsule?

And I was trying some ground plane measurements. I was thinking for me its easier to iso the cab on an auralex isolation riser vs. iso'ing the mic. Thoughts on iso?

Thanks so much,
Richard
 
Convolution with the inverse sweep is equivalent to deconvolution with the original sweep.

The follow-on to axeomatic, aomDSP, does deconvolution (plus convolution with the inverse), among other things.

The AxeFx applies windowing in the cab block.

Putting the speaker on a short stand like the auralex should be OK if that is the sound you want to capture. The mic must be on a hard surface that reflects as much of the acoustic energy as possible; if you have carpeting you should lay down a sheet of plywood and put the mic on that.
 
Convolution with the inverse sweep is equivalent to deconvolution with the original sweep.

The follow-on to axeomatic, aomDSP, does deconvolution (plus convolution with the inverse), among other things.

The AxeFx applies windowing in the cab block.

Putting the speaker on a short stand like the auralex should be OK if that is the sound you want to capture. The mic must be on a hard surface that reflects as much of the acoustic energy as possible; if you have carpeting you should lay down a sheet of plywood and put the mic on that.

Thanks for the followup LMO!

If I don't use any shockmounts or isolation, will vibrations from the cab affect the IR capture if the mic and the cab are both on the same hard surface? I.e. phyically coupled?

Richard
 
Thanks for the followup LMO!

If I don't use any shockmounts or isolation, will vibrations from the cab affect the IR capture if the mic and the cab are both on the same hard surface? I.e. phyically coupled?

Richard
That shouldn't be an issue. Sound traveling through solids is subject to absorption and scattering, so along with the weak acoustic coupling you get from the speaker cab the isolation is actually pretty good at 2m.
 
LMO covered it.

As for decoupling... I highly suggest it.
With my first run of caps when working with AlbertA, I forgot to decouple both the cab and mic stand. While this can be *fixed* to a certain extent, I would much rather build and utilize good/*proper* routines from day one.

I forgot to mention metering/graphing. Octave has plenty of options there, as do most DAWquencers, either natively or via plugins, many free. Add in a good surgical EQ.

If done properly, what you capture with a mic on a cab can be nearly identically recreated with the same source captured directly + an IR captured with the same chain + placement. This is vital to understand IMO. Thus, if you find you love the sound you are getting with a certain chain and the mic or mics in a certain spot, stop and sweep it/them, take notes/pics, and you have a "snapshot" you can reuse later. Nothing you record will reproduce what you hear in a room IMO. Some techniques can get some/many people close. But for the linear, time invariant bits, an IR is quite capable. (not sure I expressed myself well/properly).
 
Back
Top Bottom