A question about the audio engine in WL 10.0.20

I have been listening to a lot of high resolution files for a project. They are at 24 bits at 96 kHz. They are mostly acoustic music. I listened to them in WL 9.5.5 originally and now I am listening to them in WL 10.0.20. There is a very subtle but yet noticeable difference between what I am hearing in WL 9.5.5 and WL 10.0.20. The details in WL 10.0.20 seem richer and with more depth that when I am listening in WL 9.5.5. PG can you elaborate on the changes in the audio engine between the two version. I am really liking what I hear in WL 10.0.20. Thanks in advance for any insight and PLEASE keep up the GREAT work!

Although many things have changed since WaveLab 9.5, I don’t see a reason for a change in the acoustic results. In WaveLab 10, more CPU power is allocated to the audio engine, but unless you reach drop out limits, the result should be the same, at your speaker/headphone stage.

WaveLab 10 has a bit faster graphics and meters too, hence is a bit more reactive, therefore this is maybe psychological, like plugins with great animated graphics that seems to sound better (!).

Both WaveLab 9.5 and 10 support 64 bit sample processing for plugins (64 bit as default), but maybe you don’t have the same setting in the preferences? (Preferences > Global > Audio)

Anyway, I prefer you feel an improvement from 9.5 to 10 than the reverse :wink:

If you really want to compare, you have to record WaveLab 9.5 and 10 to some other computer, digitally, to get a digital copy of the streams. And then do a file compare (WaveLab file comparer).

Thanks for the quick reply. It is in the monitoring not the actual file that I am hearing a difference (so a file comparison would probably not show anything but I will try it) i will check to make sure both WLs are 64 bit but I assume they are. Thanks again for a superior product that meets or exceeds all expectations consistently. I just wish that the first installment was NOT so buggy but I understand.

Everything is set the same for both programs. I can definitely hear a difference between WL 10.0.20 and WL 9.5.5 . In 9.5.5 the transients are “blurred” where in 10.0.20 they are crystal clear. No plugins in the chain. I am listening to 96 KHz files at 24 bits. Why the difference??? I have no idea. Same sound card same monitoring setup. I just listened to over an hour of music, mostly acoustic. Maybe something changed in the audio engine or my setup just likes the new version of WL. Anyway I am happy with the changes.

FWIW

I will not try to argument against you in this case :wink:

I have had four people who’s ears I trust listen to WL 9.5.5 and WL 10.0.20 playing the same music and the consensus is WL 9.5.5 sounds good and warm and that WL 10.0.20 has more sharply defined transients BUT they are not as closely associated with the music as they should be. (they are off by a fraction of a second). My business partner, who is a GRAMMY nominated mastering engineer, says it sounds like WL 10.0.20 has some jitter problems. Please see and read this article http://www.positive-feedback.com/Issue43/jitter.htm FWIW. Please note that everything is equal between the two programs when it comes to monitoring. I am using a RME 9632 sound card, an Avocet monitor controller, a Bryson 4BST power amp and a pair of ALON IV speakers. I am running Windows 10 professional.

FWIW

they are off by a fraction of a second

What do you mean by this?

As the article mentions, Jitter is a clock issue. WaveLab 9.5/10 “posts” the samples to the audio driver, and Jitter, if any, can happen after this step, on the driver side, because WaveLab does not deal with a clock at any point of its code.

This being said, one thing which is different in 10 compared to 9.5, is that the audio in 10 is posted to the audio driver from a thread that has a “higher priority”, hence that is subject to be collected more “in time”
by the audio driver (compared to 9.5).

This being said, we are, in all cases, in a tolerance margin, that should not matter, because as long as there is no drop out, the data sent to the driver is the same.

In a system where you have high quality drivers such as RME, and on a machine CPU capable enough, I doubt there can be a difference. But well, I am not inside your computer system and even less have your ears :slight_smile:

Update as of 2-23-20

I updated the bios in my mastering room computer and then reinstalled the latest driver from RME and the difference are now very slight if at all. FYI

I guess there was something going on. Thanks PG for the help and guidance.

Have a GREAT weekend.