L2 - Hardware VS Software

Posted: Fri Mar 08, 2002 8:49 am

....

Posted: Fri Mar 08, 2002 10:05 am

Brad quote:

Quote:

I'm glad I have the hardware version now.

Somehow I knew that this should be the test result. I don't know why I thought this.

I started eight years ago with plugin mastering (I couldn't afford those outboard tools). Later I compared a z-system EQ with the focusrite protools eq.
I was schocked by the sonically difference.
Later I checked tools from weiss.
Man, what a difference in sound comparing with plugins.

Now I dont use plugins and only dedicated outboard gear.

Thanks Brad that you did the test. It always takes time to do that.
I'm still working with the K6 and after some mistakes I begin to discover a new sound envirement, just like I did when I compared my plugins with the weiss and z-system tools.

Tell me when you are ready with the cranesong.

groet (dutch for cheers)
www.cut-n-clear.com

Errol Lem

Posted: Fri Mar 08, 2002 11:09 am

....

Posted: Fri Mar 08, 2002 12:32 pm

Good to read these results Brad. Thank you :tu:

Posted: Fri Mar 08, 2002 1:29 pm

Hey Brad,
I'm glad you ran this comparison. I am a Waves Beta tester and if you remember I just about bought the hardware version of L2 a while back. I ended up with the HEDD192 for A/D conversion and have been using the new L2 RTAS. According to our tech person we work with at Waves on the beta testing, the software version is supposed to be the same as the hardware. I'm glad you actually ran it through the test.
Allen

Posted: Fri Mar 08, 2002 1:37 pm

I just emailed the Waves guys about this. Here's the response so far.

Quote:

Eric,
Thanks for bringing this to my attention. Basically, I want to further put this to the test before I post an official reply.
The software L2 has a 56k Motorola 48bit on a 24bit fixed point processor function and an equivalent Naitve PPC or x86l 32bit floating point. These are essentially all coded from common math and the Motorola code that runs on Digidesign TDM systems. The TDM version of the software plug-in is identical to the L2 hardware that you own. Now, the native version can't actually be 1:1 but it's as close to -110dB floor. You can only now if you put it to the test, get the 2 weeks demo to try and cancel out the L2 vs. its software brother. Make sure their phase is sample aligned (bitscope?). Run the same sound through both, mix them, reverse the phase on 1 and listen to silence. The difference should be negligible, but it will be there. Normalizing the difference will probably set it out of context, you have to listen with your ears. Basically this difference is in the little fractions of different math used by the different processors.
I at this moment am not aware of an inferiority and if such exists then its a bug and I will try to track it and fix it. But, I don't believe there is one.
BTW a cancellation test as I suggest may have Jitter error that is a floor itself. I've done these a couple of times to assure the quality of the software L2 but Sadie is one setup I don't have.
With the Masters bundle you get the Linear EQ and Multiband in the same go. Now these you should have if you decide to hold on to your hardware L2, or not. The amount of value is truely high for this product, and we spared no effort, care and passion with years of research.
I think the difference, may be somewhat noticeable to some super sensitive ears but environment variables may sometimes challenge the quality of cancellation and A/B comparison or blindfold test. The same L2 algorithm may also sound different to itself, if its clocked from the AES, or Word Clock. I want a chance to check on this but the general idea is to post an invitation to test it yourself, get a scientific proof of equivalence to avoid voodoo.

Amir Vinci
Product Manager
K.S.Waves Ltd.

Interesting stuff, no?

Posted: Fri Mar 08, 2002 2:14 pm

....

Posted: Wed Mar 13, 2002 7:34 pm

Hello Brad, but it's still better than the L1 plugin right?(comparing it to the L2 plugin) :confused:

Posted: Thu Mar 14, 2002 6:27 am

....

This topic sponsored by: