Avid Pro Audio Community

Avid Pro Audio Community

How to Join & Post  •  Community Terms of Use  •  Help Us Help You

Knowledge Base Search  •  Community Search  •  Learn & Support


Avid Home Page

Go Back   Avid Pro Audio Community > Pro Tools Software > Pro Tools

Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 05-18-2022, 10:10 PM
amishsixstringer amishsixstringer is offline
Member
 
Join Date: Oct 2017
Location: Athens, Ohio
Posts: 34
Default Hardware delay comp ahead lynx

I recently switched from a lynx aurora using the aes pcie card to a lynx aurora (n) using thunderbolt 3. Windows 10. Most recent pro tools build. All firmware is current.

With 44.1khz sample rate my h/w insert latency is 71 samples AHEAD. I have never had this issue with the old lynx. The comp just worked.

In the i/o settings box where you can modify hw offsets, all channels are set to zero. Putting any positive values in here makes the problem worse. So, I tried inputting a negative value, but pro tools does not allow this.

I have done the test multille ways and always 71 samples ahead. At 96khz sample rate the h/w insert is 61 samples ahead.

I mix hybrid using inserts and have for years, but this really throws a wrench into my spokes. I already have been in contact with lynx and they seem surprised by this issue. I have found a handful of people with the same problem, but zero solutions.

Please help!

Neil
Reply With Quote
  #2  
Old 05-18-2022, 11:35 PM
Darryl Ramm Darryl Ramm is online now
Member
 
Join Date: Nov 2010
Location: USA
Posts: 19,510
Default Re: Hardware delay comp ahead lynx

Make sure you test this all out in a trivial test session with no plugins at all (a rare plugins reporting incorrect delay comp numbers to the delay compensation system could cause problems).

I would not check ignore errors (I don't see Pro Tools correcting for that increase in latency on my system)--hate working with that ever checked anyhow. Checking ignore errors adds latency, which will move things the direction you want but I'd still not do that. Fix that with manually changing track +/- delay comp.

Pro Tools (at least Studio/Ultimate) seems adept at querying the latency values reported by the interface driver and applying them to hardware inserts, even when H/W Insert Delay is set to/showing 0 (i.e. Default). At least it does that well for me with PT Studio 2022.4 on macOS and a RME Fireface UFX+. Where it automatically sets what looks to be ~sample accurate alignment for hardware inserts (i.e. with H/W Insert delay set to 0.00 ms).

With the H/W insert delay set to 0 and delay comp enabled what latency do you measure? With comp delay disabled what latency do you measure? What latency does the track delay comp show (with it turned on obviously) is being caused by the HW Insert? (with H/W Insert delay set to 0.00 ms) -- that's Pro Tools calculation here and hopefully includes latencies the Lynx driver is reporting. These three numbers should be consistent, i.e. the track delay comp number subtracted from the non-compensated latency you measured should added to the compensated latency (i.e. what seems to be the negative number here).

The Reference Guide documentation says conversion latency is not included here except in Digilink interfaces, which does not jive with how it seems to be working for me, but that would rely on the interface vendors drivers correctly reporting latency.

Once you have both the measured and calculated by Pro Tools latency numbers if they still don't agree you can ask Lynx if this is because their driver is reporting incorrect latencies? I'm assuming their Thunderbolt driver is basically their existing PCIe driver, maybe they need to update the latency numbers they are reporting.

And for now you can use the track +/- compensation value to fine tune things. But hopefully if you play around more you might find some other issue. It's all a PITA, the UI/UX for all this is so badly done. The documentation does not appear to reflect reality. Sloppy stuff. But with care I expect you can get any system to work, but Avid really ought to clean this mess up.

Changing sample rate, IO buffer size, ignore errors, reduce additional latency, ... all has an effect on insert latency, and changing sample rates and IO buffer size will cause the interface to report/Pro Tools to calculate a different correction value. So set this up carefully for the actual settings you will use. Or measure a small set and keep a handy table of settings.

Last edited by Darryl Ramm; 05-19-2022 at 09:11 AM.
Reply With Quote
  #3  
Old 05-19-2022, 03:41 PM
amishsixstringer amishsixstringer is offline
Member
 
Join Date: Oct 2017
Location: Athens, Ohio
Posts: 34
Default Re: Hardware delay comp ahead lynx

Thanks for taking the time to make that detailed reply. I'll do my best to answer each section.

The session is clean and has no data in it. No plugins.

I never have ignore errors checked.

I have had no real issues in the past with the old lynx aurora. It always just worked. There were times when I could hear it flamming and I'd restart the computer and it went away.

With the h/w comp set to 0.00ms at 44.1Khz sample rate I have 71 samples of PRElatency (is that a word?). If I turn off delay compensation it's 4131 samples Regular latency (so, a difference of 4202 samples). The calculation on the bottom of the track in mix window is 2154 and I'm at 1024 buffer rate. I'm not sure if these numbers add up. I'm honestly not following 100% in this section, but those are the numbers I have.

I've been talking to Lynx and they seem to be quite surprised/confused. They've basically had me do the normal trash prefs, update drivers, software etc...which was already updated. They now want me to test it in another DAW to make sure it's an AVID issue...which let's be honest....it probably is an avid issue...

The lynx driver is indeed the same for the TB3 card and the PCIe (AES16) caards. I'm assuming they aren't going to update the reported latency if I'm on of like 5 people with this issue.

I can get the H/W insert to line up by putting a positive 71 sample value into the +/- on the bottom of the track, but this is horrible. If I make the insert inactive (in order to bypass) it keeps the 71 samples of compensation and plays back out of time. This is a real problem for parallel processing. Basically, I'd need to make a duplicate of whatever track or bus I have the hardware on and have one with and without the insert and its proper delay and then toggle between muting them. This is NOT how I want to work.

Honestly, I've been in Pro Tools 17 years now. I've made my living with this software exclusively for the last 5 years. I'm very close to bailing on AVID. In all my years with this software I don't think there has ever been a setup that just worked without doing workarounds. I just spent 30k updating the studio with a new interface, summing amp, monitor controller, outboard gear...and I am mixing my current project ITB because of this.

Sorry to vent/rant, but you are correct. AVID needs to clean up this mess.

Neil
Reply With Quote
  #4  
Old 05-19-2022, 04:42 PM
Darryl Ramm Darryl Ramm is online now
Member
 
Join Date: Nov 2010
Location: USA
Posts: 19,510
Default Re: Hardware delay comp ahead lynx

Quote:
h/w comp set to 0.00ms at 44.1Khz sample rate I have 71 samples of PRElatency (is that a word?).

If I turn off delay compensation it's 4131 samples Regular latency (so, a difference of 4202 samples).

The calculation on the bottom of the track in mix window is 2154 and I'm at 1024 buffer rate.
Ugh my head hurts. Edit: after measuring 2022.4 at 96 kHz with my interface and seeing it so accurate I might have overestimated what Pro Tools was doing. I've made more measuemntrs since then and will post a more complete wtite up. But in the meantime it would be great if you could measure the actual RTL with the RTL Utility I mention below.

And I'm not sure exactly what is happening here but I still want you to exclude or not if Lynx is reporting wrong latencies and compare those to what Pro Tools hardware insert delay it's applying.

Lets make this easier/do it a different way. Download the RTL Utility app from Oblique audio here: https://oblique-audio.com/rtl-utility.php. It is easy to use you loopback two interface ports with a direct cable and tell it what output and inputs you have connected then you run the test. It tells you both what RTL latency it measures and also what latency it calculates based on what the driver is reporting.

Edit: We'll see if the driver reported values affect this for you or not but if Pro Tools does only internal calculations of latency and does not read the driver reported latency then how can it get it so wrong for you?

Just run the RTL Utility test at a few sample rates and buffer sizes, including specifically whatever settings you have been using.

The "Measured" and "Reported" (i.e. calculated by RTL Utility from what the driver reports) RTL numbers that RTL Utility produce should be close, like exact or within a few samples.

<edit: comments deleted>

Can you share your test/measurement session?

Last edited by Darryl Ramm; 05-19-2022 at 11:15 PM.
Reply With Quote
  #5  
Old 05-19-2022, 11:25 PM
killah_trakz's Avatar
killah_trakz killah_trakz is offline
Member
 
Join Date: Sep 2008
Posts: 126
Default Re: Hardware delay comp ahead lynx

Quote:
Originally Posted by amishsixstringer View Post
I recently switched from a lynx aurora using the aes pcie card to a lynx aurora (n) using thunderbolt 3. Windows 10. Most recent pro tools build. All firmware is current.

With 44.1khz sample rate my h/w insert latency is 71 samples AHEAD. I have never had this issue with the old lynx. The comp just worked.

In the i/o settings box where you can modify hw offsets, all channels are set to zero. Putting any positive values in here makes the problem worse. So, I tried inputting a negative value, but pro tools does not allow this.

I have done the test multille ways and always 71 samples ahead. At 96khz sample rate the h/w insert is 61 samples ahead.

I mix hybrid using inserts and have for years, but this really throws a wrench into my spokes. I already have been in contact with lynx and they seem surprised by this issue. I have found a handful of people with the same problem, but zero solutions.

Please help!

Neil

You are not the only one going through this. My APOLLO based rig does the same CRAP!

All PROTOOLS needs is a PING FEATURE!

That’s ALL I need! I have to use other daws because of this and I HATE IT.

Smh


Sent from my iPhone using Tapatalk
__________________
WENT FROM HD TO M-POWERED
Reply With Quote
  #6  
Old 05-20-2022, 09:29 AM
amishsixstringer amishsixstringer is offline
Member
 
Join Date: Oct 2017
Location: Athens, Ohio
Posts: 34
Default Re: Hardware delay comp ahead lynx

So, thanks again for your replies. I'm going back into tracking mode for the next 4 days, so I won't be able to work on this issue for a bit. BUT, last night:

I talked to Lynx again and they asked me to write down the latency numbers for each playback engine buffer size. I already did this once, but decided to humor them by doing it again, and the results were strange...

64 samples = 0 Latency....ummm...what? since when?

128 samples = 256 samples of DELAY...

256 samples = 0 Latency

512 samples = 0 Latency

1024 samples = 0 Latency...

So...okay? What just happened? I changed nothing. So ran thru them again.

This time all settings were sample aligned. I don't understand what's going on here. It seems each time the playback engine cache is reset I get different values. This is unsettling. Do I need to keep restarting and testing every session to feel confident the system is working?

Does this give any clues?

Thanks again,

Neil
Reply With Quote
  #7  
Old 05-20-2022, 11:17 AM
Darryl Ramm Darryl Ramm is online now
Member
 
Join Date: Nov 2010
Location: USA
Posts: 19,510
Default Re: Hardware delay comp ahead lynx

Neil

So the numbers you are seeing are the DLY latency comp numbers show for a track with a single insert on it right? And you have H/W insert set to 0.00 in all cases?

And to repeat BScout's question he asked recently in another thread on insert latencies do you have low latency monitoring enabled? -- Keep low latency monitoring turned off while you try to understand what is going on.

When you get time...

Again, what I'd do is measure the actual and driver reported RTL numbers using the RTL Utility app. You may well be able to to this with Pro Tools open if the drivers are all nicely multi-client but you may need to close Pro Tools and run the RTL Utility separately.

Disk cache will have nothing to do with this. They are just orthogonal parts of Pro Tools.

If (big if) Pro Tools is using any driver reported latencies as part of it's calculation then crazy driver reported numbers might explain stuff, it's not really clear to me what Pro Tools does here but I'm doing some measurements to try to work it out, I'll just swap interfaces and see if the numbers change. And it might be that other utilities reporting more driver latency detail that RTL Utility might help but lets start there. I'm not sure Pro Tools actually uses any of the multiple latencies that drivers report but I just want to understand. Turns out that at 96kHz my RME UFX+ latencies and what Pro Tools was calculating was sample accurate, making me think they had started using the driver reported data. But at other clock rates it turns out it's not.... but what I get is not horrible, and I can get minor (~20 sample ~ 96kHz) negative offsets when say using digital loopbacks (AES, ADAT, MADI etc.) not a crazy situation [digression on digital outboard: even if the driver is providing latency data since they can't communicate different latencies for different I/O ports. And the hardware has no way of getting the latency of devices on the other end of a AES, ADAT or MADI connection even if the driver could pass this back to application]. I often use digital outboard gear... but it has hundreds of sample of it's own latency, so I'm always correcting with a positive time for those.

So while the UI/UX for all this sucks, and the low level calculations likely could be done better, and a ping function would be nice... I still have not ever seen crazy large negative offsets. I've sure seen problems in the distant past with related things with causes like plugins reporting wrong latencies and folks setting up complex routing problems with sends and aux, and low latency monitoring rearing its ugly head.

I would try to get stuff working in as simple as a possible setup, I know you are trying to do that. But again, don't enable ignore errors (now I do see Pro Tools tries to compensate for that, but just don't). Don't enable Low Latency Monitoring. Don't have any plugins inserted anywhere in the test session. Don't have any sends or aux channels (the send worries me more than the aux). Just a couple of mono audio channels. Signal routing though sends and stuff running through the large (1,000 or 2,000 sample) disk playback buffer and not properly being accounted for might introduce additional latency.

Can you send me your test session, post it somewhere with your negative waveform measurements in it. I'll run some additional measurements here and send back to you. I'm do recall one weird thing with a very? corrupt? session in the past was showing crazy total system delay and other stuff and deleting tracks and routing etc did not clean it up. So curious just to check that the session system delay (in Setup>Session) makes sense (in a trivial one hardware insert test session should be the H/W insert delay shown on a track along with output and possibly input delays).
Reply With Quote
  #8  
Old 05-20-2022, 03:18 PM
Dax Productions Dax Productions is offline
Member
 
Join Date: Mar 2008
Posts: 108
Default Re: Hardware delay comp ahead lynx

Quote:
Originally Posted by Darryl Ramm View Post
Neil
. but what I get is not horrible, and I can get minor (~20 sample ~ 96kHz) negative offsets).
Darryl this is the problem. 20 samples is pretty close. It’s close enough that I can insert something on a one track instrument and not be worried at all. I have my 1176 on my vocal bus on every mix I do. In my case it’s 62 samples early at 44.1. Not enough to sound out of time, and for that situation I don’t have to worry about comb filtering with parallel elements.

The problem is as soon as you want to do any parallel processing. 20 samples will have bad, glaringly obvious comb filtering if I want to use my adr compex as a drum back buss. It needs to be zero samples.


I did try the RTL app and it does confirm that the driver reported latency is slightly more than the actual recorded latency. 62 samples at 44.1, and similar, but not far off numbers for other rates.

My current interface is a Behringer 1820 but I have had similar “over compensation” with UA, Motu, and focusrite interfaces.
Reply With Quote
  #9  
Old 05-20-2022, 04:37 PM
Darryl Ramm Darryl Ramm is online now
Member
 
Join Date: Nov 2010
Location: USA
Posts: 19,510
Default Re: Hardware delay comp ahead lynx

Quote:
Originally Posted by Dax Productions View Post
Darryl this is the problem. 20 samples is pretty close. It’s close enough that I can insert something on a one track instrument and not be worried at all. I have my 1176 on my vocal bus on every mix I do. In my case it’s 62 samples early at 44.1. Not enough to sound out of time, and for that situation I don’t have to worry about comb filtering with parallel elements.

The problem is as soon as you want to do any parallel processing. 20 samples will have bad, glaringly obvious comb filtering if I want to use my adr compex as a drum back buss. It needs to be zero samples.


I did try the RTL app and it does confirm that the driver reported latency is slightly more than the actual recorded latency. 62 samples at 44.1, and similar, but not far off numbers for other rates.

My current interface is a Behringer 1820 but I have had similar “over compensation” with UA, Motu, and focusrite interfaces.
I am not defending it, I'm just saying what I measure. And I understand the impact. But those negative offsets only happen for me with interface digital ports, where the I/O is faster than Pro Tools is either naively just calculating or is calculating maybe with help of driver latency info. It *seems* Pro Tools may be just calculating 2 x IO Buffer size + 157 samples for my measurements with a Fireface UFX+ at multiple sample rates and IO Buffer sizes - with ignore errors *not* checked). I'm curious if you see the 2 x + 157 (remember not to check ignore errors). And it's not like I'm running an interface that is a slouch at conversion times.

What *I* don't see is wildly negative numbers for an individual insert, or problems with multiple inserts chained, or behavior that is inconsistent and changes and I know other folks are reporting seeing that, and it would be really good to try to work out why or reproduce the insanity on other systems. For me Pro Tools can be off a little, but all consistent with the likely speed of different conversion paths in the interface, and its all self consistent whether I'm using RTL measurements or letting Pro Tools do it's thing then measure how far out it is/correct it with the funky H/W insert menu. Again for all practical purposes I don't get negative values overall for an inset... e.g. on the rare occasion I say use an Eleven Rack as an insert it's latency far overwhelms the ~-20 samples Pro Tools sits at though all the digital connections to get to the Eleven Rack, so I end up sliding it forward again with the H/W insert delay setting. Again, just saying what I see, I'm very curious to try to work out what the heck is going on for other users.

And regardless the UI/UX is awful here, it all needs a serious rework.

If I shared a spreadsheet with you would you be willing to add your measurements to that sheet using Pro Tools and RTL Tool for your interface?

Last edited by Darryl Ramm; 05-20-2022 at 06:12 PM.
Reply With Quote
  #10  
Old 05-21-2022, 09:39 AM
Dax Productions Dax Productions is offline
Member
 
Join Date: Mar 2008
Posts: 108
Default Re: Hardware delay comp ahead lynx

Sure, want to post it here and we can all contribute?
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Carbon Delay Comp with external hardware? tomhartman Pro Tools | Carbon 2 09-04-2021 08:41 AM
issues: HDX-2 (PT10) delay comp…. RTAS & hardware inserts MixerGuy Pro Tools 10 8 10-06-2014 03:21 PM
Pro Tools 9 Hardware Insert Delay Compensation with Lynx AES16e and Aurora 16 SolutionRoom macOS 0 04-08-2011 10:54 AM
Hardware Insert Delay Comp Issue therecordinghouse Pro Tools 9 7 11-17-2010 06:01 AM
Hardware delay comp blackbirdstudio Pro Tools TDM Systems (Mac) 1 01-03-2010 09:35 AM


All times are GMT -7. The time now is 10:24 AM.


Powered by: vBulletin, Copyright ©2000 - 2008, Jelsoft Enterprises Limited. Forum Hosted By: URLJet.com