This video is a further investigation into making impulse responses of guitar speakers / cabs.

The Pro Tools session along with all of the IRs is available for download HERE: [ Ссылка ]

I already held the belief that room acoustics are of high importance, and one way to get around the issue of building / using a treated room is simply to record the sweeps outdoors, in order to minimize early reflections getting back into the close mics.

Obviously this technique is not without its drawbacks, i.e bird noises, cars, builders and all manner of suburban noise that can add unwanted 'spice' to your final IR files, however if you keep the sweep relatively loud and short (I used 1 second sweep throughout the video) and just redo it if something noisy occurs when you are recording, the results are often better than the phasey sound you will often get if using a bare-walled room in your house.

The other factors I wanted to investigate include using a tube vs solid state power amp for the sine sweep, and activating / deactivating options such as 'reversed technique' and 'minimum-phase-transform' (see [ Ссылка ] for a description of these parameters).

In order to A/B all of this, I created several IRs of different speakers and cabs I had, using two sweeps - one with a fade-in and fade-out for the reversed technique and one regular sweep with no fades.

I then used my BOSS Waza TAE and my Marshall SC20 (power amp only via FX return) to play both sweeps through each speaker, and deconvolved all of the recorded sweeps with MP transform on, and then off.

After this I created separate tracks in Pro Tools for each change in variable, so that I could go between each one and listen to how a given parameter affects the sound. For example, for the Celestion G12K-100, I was able to listen to the IR that used the TAE compared to the SC20, as well as the IRs that did and did not use MP transform, or reversed technique and so on and so forth.

From this I have concluded that MP transform and reverse technique do yield a better result from Voxengo Deconvolver and so I will probably just keep those options in use for further IR creation.

To summarize, in my opinion, the factors affecting speaker cab IR quality in order of importance are:

1. Room Acoustics
2. Speaker + Cabinet selection
3. Mic choices, blend amounts and positioning
4. MP Transform + Reverse Technique switched ON
5. Sine sweep duration / volume

Other Observations
- I do not think 500ms IRs are necessary at all, and you're better off simulating that extra resonance using a short reverb in mixdown. You only really need to capture the tiny bit of time that the cab resonates for once the sweep has ended, and as the sweep ends at 20kHz, it will likely really only be resonating for a very small amount of time (mostly from the first ~150hz of the sweep which occurred in the first 0.5s of a 1s sweep). Furthermore, with most setups, your noise floor will quickly rise beyond any audible resonance once the sweep has reached 20kHz.

- Longer (say 5-10s) sweeps don't really achieve any extra fidelity (and are particularly problematic to record outdoors!) however I will investigate this a bit more perhaps on a quiet Sunday evening.

- Whilst guitar speakers do tend to sound quite different with changes in volume, it usually isn't worth creating IRs at high SPL as any distortion in the speaker will not sound good in the IR due to the non-linearity. As a general rule of thumb I'd recommend recording them at a level that gets you nicely above the noise floor of your equipment (mainly your interface output and power-amp in use) whilst keeping below the speaker's distortion threshold - find this point by ear.

Hopefully you find all of this useful and entertaining, feel free to comment below if anything needs further clarification.

Timestamps
0:00 the plan
2:35 first sweeps
2:41 swapping out the G12 Vintage for the WGS Green Beret (timelapse)
2:49 green beret sweep
2:50 swapping out the G12 VIntage in the SC20 for the WGS Veteran 10 (timelapse)
2:58 veteran 10 sweep
3:01 setup for A/Bing the final IRs in Pro Tools
3:25 extended A/B of the IRs and changes in variables
15:04 trying a resistive load instead (Bugera PS1)
16:01 back to the TAE, testing some cleans and speaker blends
17:52 my favorite blend (K100 + Green Beret)

свернуть