| Author |
 Topic Search  Topic Options
|
Sonicsix 
Bobcat
Joined: December/11/2010
Points: 58
|
 Topic: Changing Buffer settings
Posted: March/24/2015 at 3:27pm |
Hi I'm having the dreaded Pro Tools error -6101 CPU overload with minimal tracks, that suggests raising the playback buffer. I usually can't go any lower than 512 for an 88.2 session without freezes. This gives noticeable latency. I'm curious if I should be adjusting the buffer settings in the Lynx mixer window to get better performance? If I try to adjust it, it doesn't change. Is this normal? Has anyone else encounter these issues? Setup is: Mac pro 2 x 2.8 quad-core processors 24 gigs of 800mhz ram 2 x Lynx aes16e cards (set to receive sync from Digital 8) 1 x Aurora 16, 1 x Aurora8 (set to receive sync via external bnc cables) 1 x UA 2192 connected via digital 8 on AES 16e ( sync set to internal and sending sync via bnc to Auroras)
Lynx mixer 1.0 build56A AES16e firmware version 13.1 Aurora 16vt and Aurora 8 firmware version 28
Pro Tools 10.3 and 11 settings in PT playback engine settings: PT Aggregate h/w Buffer size 256 samples Host Processors 7 CPU usage 90% Ignore errors unchecked Delay Compensation MAx Cache 14gb Plug in Buffer Default level 2 250 ms
|
 |
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: March/30/2015 at 4:16pm |
The buffer size should be set within ProTools, not from the Lynx Mixer. That shouldn't have anything to do with the performance problems though.
Your settings seem OK, you should be getting much better performance than that. Is it possible that there is a plug-in that is creating havoc? You might try instantiating them one at a time and see if any make the CPU number shoot up.
Can you test with just one card (instead of using the aggregate) and see if the behavior changes?
Also, with the aggregate device make sure that Drift is OFF for both devices. You can check that from Audio Midi setup. Generally drift is turned ON for the second card in an aggregate and that can cause problems.
|
|
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: May/29/2015 at 5:48pm |
I think this is something to do with my sync errors that I keep having. When I change sample rates for different sessions recorded at different rates I keep getting sync slipping out and the ring modulated audio sound constantly. Then after switching sync sources all around and back, it works for a little while and sometimes falls out of sync again, or sometimes works. I'm going to try to change my clock chain, before suicide.
Was set up as: UA 2192 set to internal (connected audio to digital channels 31/32 on digital in 8)> bnc cables to Aurora 16 set to external, and Aurora 8 set to external> both AES 16's set to external via digital in 8 (ua2192)
Now going to try: Aurora 16 internal > Aurora 8 set to external via bnc or aes?> UA2192 set to external via bnc or ads?> both AES 16e's set to external on digital 1. Does this seem like the best option?
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: May/29/2015 at 6:20pm |
I guess since there is only 1 bnc clock out connector on the Auroras, I'll try syncing the Aurora 8 and the UA 2192 via aes.
edit: OK so I have tried this and it seems to be working for the moment. An interesting thing is that on the Lynx mixer app, my preferrewd clock source digital 8 window is showing a clock sample rate of 33khz. Could it be that the clock chip is shot in my UA 2192?
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: May/30/2015 at 1:01pm |
 Sonicsix wrote:
I guess since there is only 1 bnc clock out connector on the Auroras, I'll try syncing the Aurora 8 and the UA 2192 via aes.
edit: OK so I have tried this and it seems to be working for the moment. An interesting thing is that on the Lynx mixer app, my preferrewd clock source digital 8 window is showing a clock sample rate of 33khz. Could it be that the clock chip is shot in my UA 2192? |
Ok that didn't work. I've now sync'd the Aurora 8 to the aurora 16 via bnc and will see how that works. Synchro lock seems on. Seems good so far, time to test....
|
|
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: June/02/2015 at 5:13pm |
|
Definitely keep an eye on the rate off of the 2192. If our frequency counter keeps showing errant rates like that there may be an issue with either a setting or a hardware problem to look at. Try different clock cables before going exploring to far down that road though...
|
|
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/04/2015 at 11:02am |
|
Thanks Paul. It appears to be randomly loosing sync and playback is gradually more ring modulated, even after I have changed the master to my Aurora 16 and have everything slaved to it. Sometimes I can get the ring modulated sync lost sound to disappear temporarily, by going into the mac audio midi setup and switching the clock source for the pro tools aggregate I/O 1: aes16e to 2: aes16e. I just changed the sample rate from 44.1 to 48k for a session and in the Lynx mixer the AES 16e card #2 is cycling a red bar under the syncrolock range. If I click preferred clock source to external ( sample rate not present) and back to Digital in 1, it locks on the sample clock screen in the mixer window. Does this seem like some sort of software switching problem? Is there any other tests I can do to check hardware or software issues? There are some gremlins here playing with me for sure.
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/04/2015 at 11:04am |
|
I will temporarily disconnect the 2192 completely to see if it is causing any problems also.
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/04/2015 at 11:17am |
|
Another thing I am noticing at the moment, is the 2: AES16e in the Lynx mixer's sample clock current source is set to internal and won't change. Even though the sync source is set to Digital in 1. Is this any clue?
|
|
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: June/04/2015 at 5:01pm |
|
That is a clue. The current source is what is actually being used. Chances are when you changed clock source, you had an audio application open. By default the card does not switch rates while open by an app. You can resolve this be either closing the app prior to changing the source, or in the Lynx Mixer you can click settings > advanced > Allow clock change if active. When this is selected, the card can change source even with an audio app open.
|
|
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/05/2015 at 9:21am |
|
Ok Allow clock change let me change the clock source. I am still getting a slip though. 2 minutes into a mix the ring modulated audio crept in. What is the recommended sync setup here. Should I purchase a AES16e header jumper and sync with the AES16e with the Auroras sync'd external to them?
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/05/2015 at 1:09pm |
|
Also just noticed the second AES16e and Aurora 8 being noticeably time delayed, but without the ring modulated distortion. I cycled the clock source in Mac os's from AES16e 1 to aese 2 and back to aes1 and it is back in time. .....
|
|
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: June/05/2015 at 5:13pm |
|
WHat happens if you use a single AES16e instead of the aggregate device. Any better?
|
|
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/06/2015 at 10:11am |
|
Yes it seems to be fine when not using aggregate device and only 1 interface.
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/06/2015 at 11:41am |
|
Today while switching from 44.1k to 88.2k (still set as master to Aurora 16 interface), the Lynx mixer doesn't change the sample rate in rate select. If i cycle the preferred clock source in the Lynx mixer, from internal on 1: AES16e (Digital 1 on 2:AES16e) to external and back it changes the rate to 88.2. Not sure what is hanging up here
|
|
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: June/08/2015 at 3:45pm |
|
Are you sure that Rate Lock is OFF in the Lynx Mixer for both cards? It should be
Also, for your aggregate device make sure that card #1 is listed first and that Drift is turned OFF for both devices.
|
|
|
|
Sonicsix
Bobcat
Joined: December/11/2010
Points: 58
|
Posted: June/16/2015 at 11:47am |
|
Yes, Rate Lock is off, Card 1 is listed first, and Drift is off. It has been acting up a bunch the last few days. I've connected the 2192 back up as a master clock only (no AES connections) and it seems to be cooperating at the moment. man this is frustrating.
|
|
PaulTech
Admin Group
Joined: August/13/2004
Location: United States
Points: 5495
|
Posted: June/16/2015 at 4:40pm |
|
You also might have a corrupted aggregate device. You might consider trashing your aggregate and starting over (remember to keep Drift OFF)
You also might consider getting the ICC cable so you can explore another clock option. With the ICc cable you could do:
AES16e #1 - Internal
AES16e #2 - Header
Auroras - AES A
Also, are you certain that the ring modulation does not improve when you increase the buffer size?
Lastly, open Console from Utilities. Set to show "all messages". Do a search for ">>>" . This will reveal any messages sent up from the Lynx driver. Copy a range where the problem happened and email those to support@lynxstudio.com, referencing this forum thread.
There may be a clue there.
Also, since card #1 worked by itself, what about trying card #2 by itself? Have the card on Internal and its connected Aurora on AES A. It would be good to know if there is a hardware problem and this step could narrow it down.
|
|
|
|
Changing Buffer settings