|
Avid Pro Audio CommunityHow to Join & Post • Community Terms of Use • Help Us Help YouKnowledge Base Search • Community Search • Learn & Support |
#1
|
|||
|
|||
"inverse lag" when using Pro Tools aggregate I/O
Hi, I just upgraded my Mac Book (Mac Book Pro 16 inch 2021 M1), running Monterey and Pro Tools Studio. I got everything set up and working great, but I noticed an "inverse lag" when I record audio. I.e. The recorded audio is shifted a 16th note at 120 bmp *ahead* of the bar line. I am using low-latency recording with delay compensation enabled, and I've tried a bunch of permutations of those off and on, but after some debugging I realized that it only does this if I'm using Pro Tools aggregate I/O. If I switch to just using one interface for the audio engine (Babyface) the "inverse lag" goes away. This is manageable I guess, but it's a hassle. I feel certain this is a settings issue - this laptop is so fast, and I'm testing with just a click track and one audio track, no plug-ins. Any ideas?
|
#2
|
|||
|
|||
Re: "inverse lag" when using Pro Tools aggregate I/O
I would not be surprised if there is an issue with how Pro Tools treats aggregate device latency. There is just enough reasons that is a potential minefield.
You are aggregating a babyface with what? What is the actual IO layout of the aggregate? What is the screenshot showing us? Can you reproduce this with pure audio signals and no MIDI/VI? since that introduces other stuff. In talking about this can you measure in and talk in samples (and clock rates) not notes. And make measurements accurate to a sample. That most usefully relates to the latencies in the system and can often make clear what is going on. I've spend a fair amount of time digging through CoreAudio and all the latency data it provides. CoreAudio Aggregate IO tries to report latencies of component devices to apps correctly, in all cases I've seen it seems to correctly report values obtained from the component devices. Who knows what Pro Tools does here, the only claim of official support for aggregate devices from Avid that I can recall seeing it they support "aggregate an interface with Built-In output. One interpretation of that might be (I have no idea) that if you use inputs on an interface that is part of an aggregate and the output is Built-In Output that Pro Tools might get this correct. Potentially complicating this is the Mac Built-In Output and input devices like Mic etc. report per-steam latencies and that carries through to the aggregate device, but it's unclear if Pro Tools correctly includes per-stream latencies in calculations, that's an area that has confused multiple developers/vendors so who knows... but doing some tests with aggregate devices might be a good way to uncover if this is being done correctly by Pro Tools. If you need to use an aggregate, you should be able to measure the latency and add/subtract that to the per track +/- compensation delay. There are some tests that would help finger Pro Tools or not as the fault. e.g. measuring RTL of the aggregate IO ports independently of Pro Tools and comparing what it does knowing what the aggregate device is reporting as latency. For me aggregate IO is just not useful so I've not spent time looking at this. I can point you are tools if you want to report latency components and make measurements, maybe more useful if you wanted to use that to then file a bug with details to Avid (--more likely, or less likely maybe a driver manufacturer?). But making measurement, which are not hard to get to be sample accurate is likely the best way to confirm there is something wrong here and who's fault it is. If you want help to look at this more contact me via a PM. |
#3
|
|||
|
|||
Re: "inverse lag" when using Pro Tools aggregate I/O
Quote:
Yes "reverse lag" is exactly Pro Tools thinking the latency is higher than it is and over correcting for that. And these issues are never performance related... if Pro Tools can't process stuff fast enough it deliberately crashes. Some of Pro Tools recording/I/O latency is compensated for even when delay compensation is off (or recordings would never even closely line up when you say play to a click). I know it's significant latency, but when calculating in samples you will find what exactly explains that offset to a sample of so... that way you can be confident what is wrong. Pro Tools gets latency data from CoreAudio, which gets it from the device drivers that have that data baked into them and/or stored in data in the interfaces. Measuring the exact number of samples you can then compare to all the numbers that CoreAudio is advertising to Pro Tools (it's a little non-obvious there are at least 4 different latencies to worry about for both input and output ports) and see if it's doing something wrong.. my guesses would start with ignoring the stream latency component, or just taking the latency of the first device in the aggregate etc. I can show you how to see all the internal latency data from CoreAudio, including with a utility I wrote, if you are interested. And yes if you are not, no problems, especially as it sounds like you can work fine without using aggregate IO. I'd like to know what Avid said to this, they might say it's all unsupported. Last edited by Darryl Ramm; 07-19-2022 at 11:54 PM. |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Installing from - "Setup.exe Application" or "Avid Pro Tools Windows installer"? | ihate100bees | Windows | 1 | 05-08-2021 08:35 AM |
"Pro Tools quit unexpectedly" when opening (holding "n" no longer fixes it)[RESOLVED] | andrew24c | macOS | 4 | 07-23-2014 07:21 PM |
Pro Tools bounce glitch with "Aggregate I/O" | rickwoody | Pro Tools 10 | 0 | 12-29-2012 12:38 PM |
Pro Tools 9 "Pausing" when I press "Play" with Focusrite Scarlett interface | eanderso22304 | macOS | 3 | 11-23-2011 06:45 AM |
Pro tools 9 crashes at startup "Bus error" in thread "MainThread", at address 0x0 | danball | macOS | 2 | 09-26-2011 04:49 PM |