x50 v2 cab movement vs recab 3 dynamics - PART II

[HOFX]

The last test had some surprising results, specifically the tracks that had no movement/dynamics seemed better because they had more weight in the chug. This effort is more about understanding what is happening with the movement vs dynamics.

The file is made up of the same riff looped a bunch of times, with Lecto as the amp sim and God's Cab 57 1 inch cone near pres 4 as the impulse, to keep everything impartial and so we're simply comparing IR loader vs IR loader and movement vs movement.

The track goes:
1st: x50 100 % movement, 100 % IR length
2nd: x50 0 % movement, 100 % IR length
3rd: Recab 0 % dynamics
4th: Recab: 50 % dynamics
5th: x50 0 % movement nulled against Recab 0 % dynamics
6th: x50 0 % movement nulled against x50 100 % movement
7th: Recab 0 % dynamics nulled against Recab 50 % dynamics

I've uploaded the file as a wav to preserve the high end and not blame things heard in the null tests on mp3 artifacts.

2014-x50-Test-3.wav

 

[HOFX]

What I hear:
- x50 0 % sounds more mono on the low freq resonance in the chugs than x50 100 %
- I cannot explain why recab 0 % nulled against x50 0 % revealed so much difference. I adjusted volume, IR length on the x50 loader, no idea what's happening here.
- Recab 50 % nulled against Recab 0 % dynamics reveals scratches. That's been revealed before, which is why I hold off on going 100 % dynamics on Recab, because I can hear it going scratchy.
- x50 100 % nulled against x50 0 % shows it's applying very subtle difference.

I guess it's up to each individual to choose the option that sounds best to them. I'm not trying to pick one superior over the other, but to reveal the differences, and as a punter, understand what's going on a bit better.

 

[TRUIE]

- I cannot explain why recab 0 % nulled against x50 0 % revealed so much difference. I adjusted volume, IR length on the x50 loader, no idea what's happening here.

I'm not able to explain this on a technical aspect, but to me it's normal that each IR loader have his own sound, even with the same IR.

Thank you for this test BTW !

 

[deLuther]

IF IR loader only does convolution without additional fancy things, than it`s shouldn`t sound different.
For example Knufinke SIR2, Voxengo Pristine Space and KeFIR have almost identical result, difference is beyound 32 bit level.

 

[TRUIE]

Isn't the IR lenght playing a big role on this ?
Some allows you to tweak this parameter (like kefIR, X50).
Some might have fixed value (max or intermediate, don't have a clue).

Also, from the Pristin Space user guide :

Pristine Space was mainly created for impulse reverb perfectionists: it does not implement various `combo' approaches which try to save CPU cycles by creating a synthetic reverb tail instead of performing a full convolution. Because such approaches in many cases give unsatisfactory results if you expect them to sound exactly like your original reverb impulse sounds, we have decided not to use them. Instead, we have optimized to the maximum both the efficiency of convolution processing and the plug-in's latency.

Maybe these "combo approaches", to some degree, and the "efficiency of convolution processing" can explain why different loaders don't sound the same.

 

[onqel]

If the length isn't exactly the same you will never get full phase-cancellation.. (infinity)
Different implementations of FFT-algorithms could also yield different results if numerical truncation is present etc