Issues with Layla & 2.0

30 mins of 24/96 x 12 tracks worked fine here.  Admittedly using an Echo Mia stereo card, not a Layla.  No obvious problems with the recorded tracks.  If my calculations are correct that should mean I'd have no trouble with 48 tracks of 44.1/16.  Can't complain.

I haven't really tested how many tracks I can record for a long time in Audition, because I am wary of losing tracks/sync or other issues.

I haven't tried my limits in Sonar, either, but the longest I've ever recorded is 6 simultaneous tracks for 2 1/2 hours @ 96khz (128 sample latency) in Sonar, using realtime input monitoring with these effects:

Snare: Parametric EQ, FX Dynamics, Glaceverb, FxEq

Low Toms: FX Dynamics Processing, Parametric EQ

Top Toms: Parametric EQ, Compressor

Bass Drum: FX Dynamics Processing, FxEQ

Guitar 1: FX Dynamics Processing, FxEq

Guitar 2: Delay, FX Dynamics Processing, FxEq

and all was great.

FWIW, here I'm running, as I write, a Darla 24 in one machine and a Darla 20 using the demo AA2.0 on the other. Both are using ASIO and the 6.11 Echo drivers. Isn't the Darla just the Layla's older sister? Her looks may have gone but she still performs.

They are recording now, the Darla 20 on 6 tracks, the 24 on 12. Both have now recorded happily for well over 30 minutes. I've previously tried long sessions on the Darla 24 without any problems. Up to now (and probably forever) I only record straight with no effects.

I did have some glitching in the past on the Darla 24 when it switched to the WDM drivers without me noticing (as previously discussed - a session copied in from the non-asio laptop reset the drivers).

Edit Actually, I have now noticed some peculiarities with the 12 tracks on the Darla 24. At some random time into the recording the scrolling update of the waveform stops, although actual recording seems to continue. I'll try to do some more research into this and report back if I find anything definite.

Ok, a bit more.  I forgot to mention that I'm using two Layla24s.  The goal is 24bit 44.1k, as many tracks as I can muster.  I cranked up the Asio buffers to 4096 and got it to go for about an hour.  I realize that I'm demanding quite a bit out of my hardware here, but keep in mind that a similar test with Vegas did much better.  During the last try I left the windows performance monitor up too.  Interestingly the CPU usage began to climb after about 30 minutes.  It was at around 75% and then climbed to 95%.  I'll continue to report as I have another day or so to experiment before I have to pack things up for the road.
JTC

FWIW, here I'm running, as I write, a Darla 24 in one machine and a Darla 20 using the demo AA2.0 on the other. Both are using ASIO and the 6.11 Echo drivers. Isn't the Darla just the Layla's older sister? Her looks may have gone but she still performs.

Wildduck,
If you have a moment at some point, would you mind taking a look under the Audition Windows Sound control panel to see what bit depth it recognizes from your Darlas.
TIA,
JTC

Hmm, that's interesting. Both Darla's just list 16-bit in all the options.

It looks to me that when rewriting the engine, Adobe concentrated heavily on ASIO and just gave basic WDM support (my Layla says 16 bit as well in Audition Windows Sound).

We know that Audition 1.5 was able to incorporate WDM correctly.  

Just the name "Audition Windows Sound" is suspicious.  That alludes to AA2.0 using the windows mixer (which only supports 16 bit) rather than proper kernel streaming (for 24bit).  

Ironically, even with Adobe only incorporating lackluster WDM support, its push for ASIO only resulted in lackluster ASIO implementation, as well.

I notice that one of the features listed in the notes to the latest Echo drivers (which don't apply to my Darla) is that they have changed the text describing the drivers from saying WDM to PCI. This implies to me (but I'm often wrong) that Echo may have been responsible for a misnomer.

It does seem to need sorting out!

I notice that one of the features listed in the notes to the latest Echo drivers (which don't apply to my Darla) is that they have changed the text describing the drivers from saying WDM to PCI. This implies to me (but I'm often wrong) that Echo may have been responsible for a misnomer.

It does seem to need sorting out!

Actually, I have feeling that the PCI reference may have something to do with kernel streaming.  I remember a reference to that in something else, possibly in the documentation for asio4all.

In Adobe's docs they specifically state that the audio engine revolves around ASIO, the Audition Windows Sound driver is a wrapper, just as asio4all is.

FWIW: It just occured to me that while previous versions were fairly integrated into windows, this version, revolving around ASIO, must stand on it's own, consuming more CPU resource.......I think.

That said, with all of Audition/Syntrillium's history with Windows audio, I can't believe they wrote a wrapper that doesn't seem to correctly identify the bit depth of any Echo product.  I'd be interested to know if the same is true for any other manufacturer's gear.

My situation summarized is as follows:
Echo ASIO driver- minimum buffer 4096 to record 24 tracks at 44.1.  It gets flaky with any kinf of video change going on too.

ASIO4ALL- no better than above, perhaps worse.

Audition Windows Sound/Echo WDM- I can't use it as it doesn't recognize my Layla's as being 24bit.

Vegas 6 w/ Echo's ASIO driver- Chugs right along recording 24 tracks of 44.1 with no apparent problem.

JTC

The 16 bit issue with Audition Windows Sound may not be confined to any particular brand of sound card. My Mackie Onyx firewire interface also becomes 16 bit with this driver (wrapper). With ASIO4all or Mackie's own ASIO driver it is 24 bit.