Qwerty
Member
Hello,
...and thanks for having me here in this fine location! I know some of you guys have been bashing your head against Sonar and multi-UAD configurations so I thought I would add my experience into the mix.
I recently added a 2nd UAD-1 card to my system which uses a RME FireFace 800 and Sonar v5.0.1. In short, turning off hyper-threading decreased my useable latency settings for normal mixes down from 5.8 to 2.2msec.
A snippet from the email I sent to Cakewalk, Sonar and RME follows.
Ciao,
Q.
---------------------
...I am using an Asus P4P800 Deluxe socket 478 mobo with an Intel 3.04GHz CPU and 2Gig RAM which is updated to the most recent v21 BIOS revision. I am using a Lacie Firewire 1394.b PCI card, (Texas Instruments OHCI controller), and the RME FireFace running the most recent BIOS and driver revisions 2.47 and 2.43. I also have two UAD-1 cards running their v4.2 software and driver. I am using Windows XP SP2 with the Firewire 1394.b hotfix and associated SidSpeed=3 registry hack as per the Microsoft website.
If I run a single track in Sonar v5.0.1 and insert a single UAD-1 FX on the track I can use input monitoring with a 48 sample ASIO latency which gives me 1.1 msec + the FW-800 round trip.
The recorded track will play back with this same configuration.
If I add another UAD effect to a track, (same one or a different one, it doesn't matter), then I begin to get clicks and pops in the audio stream. The only way to resolve this completely is to increase the ASIO latency to 512 samples - 5.8msec + FW 800... This seems to be a threshold limit - track or plugin count didn't matter. Below this point snap, crackled and popped -- above this mark did not.
I then turned OFF hyper-threading in the BIOS, (which disabled the multi-processing engine of Sonar)...
Same situation still applied, but with different thresholds. I am stable playing or tracking with input monitoring on multiple tracks with multiple instances of UAD effects with the latency set down to 96 samples - 2.2 msec + FW...
There is obviously something up with HT + the VST adaptor's implementation of the UAD v4.2 plugs...
...and thanks for having me here in this fine location! I know some of you guys have been bashing your head against Sonar and multi-UAD configurations so I thought I would add my experience into the mix.
I recently added a 2nd UAD-1 card to my system which uses a RME FireFace 800 and Sonar v5.0.1. In short, turning off hyper-threading decreased my useable latency settings for normal mixes down from 5.8 to 2.2msec.
A snippet from the email I sent to Cakewalk, Sonar and RME follows.
Ciao,
Q.
---------------------
...I am using an Asus P4P800 Deluxe socket 478 mobo with an Intel 3.04GHz CPU and 2Gig RAM which is updated to the most recent v21 BIOS revision. I am using a Lacie Firewire 1394.b PCI card, (Texas Instruments OHCI controller), and the RME FireFace running the most recent BIOS and driver revisions 2.47 and 2.43. I also have two UAD-1 cards running their v4.2 software and driver. I am using Windows XP SP2 with the Firewire 1394.b hotfix and associated SidSpeed=3 registry hack as per the Microsoft website.
If I run a single track in Sonar v5.0.1 and insert a single UAD-1 FX on the track I can use input monitoring with a 48 sample ASIO latency which gives me 1.1 msec + the FW-800 round trip.
The recorded track will play back with this same configuration.
If I add another UAD effect to a track, (same one or a different one, it doesn't matter), then I begin to get clicks and pops in the audio stream. The only way to resolve this completely is to increase the ASIO latency to 512 samples - 5.8msec + FW 800... This seems to be a threshold limit - track or plugin count didn't matter. Below this point snap, crackled and popped -- above this mark did not.
I then turned OFF hyper-threading in the BIOS, (which disabled the multi-processing engine of Sonar)...
Same situation still applied, but with different thresholds. I am stable playing or tracking with input monitoring on multiple tracks with multiple instances of UAD effects with the latency set down to 96 samples - 2.2 msec + FW...
There is obviously something up with HT + the VST adaptor's implementation of the UAD v4.2 plugs...