|
Avid Pro Audio CommunityHow to Join & Post • Community Terms of Use • Help Us Help YouKnowledge Base Search • Community Search • Learn & Support |
#1
|
||||
|
||||
Standard way to measure monitoring latency?
I have been considering moving from Pro Tools 10 to a Pro Tools 10 HD Native system.
I am looking at this comparison chart for monitoring latency and I am wondering how my current Lynx AES16 PCI/Aurora 16 with PT 10 (not HD) compares: http://www.avid.com/static/resources...ilyLatency.png Is there a standard way of measuring monitoring latency (the time it takes for a signal to undergo a AD conversion, be processed through the PT mixer, and then undergo a DA conversion)? Does Pro Tools report this figure anywhere? I have been using a single sample audio "spike" (drawn in with pencil) played on an audio track and routed to output 3 of my interface, then the analog signal is patched into input 1 on my patchbay, passed through a record-armed channel in Pro Tools (LLM is off) and passed out to output 1, and finally the once again analog signal is patched into input 3 which is recorded on another track. If you follwed that (!) there is a DA conversion followed by an AD conversion followed by a DA conversion finally followed by an AD conversion. The first and last conversions should be compensated for; it is the AD + DA in the middle that occurs during the monitoring on the record armed track that I think is what I am trying to measure. The ruler is set to samples and I then select the time between the original spike and its final recording as indicative of my monitoring latency. At 96kHz and a buffer of 64 samples, I am getting 153 samples which should be about 153/96,000 = .0016 seconds or 1.6 ms. I am wondering about that figure because it is comparable to that reported for HD Native. Is it possible for a purely native system to perform comparably to an HD Native one? I wouldn't have thought that to be the case. Am I doing this correctly? Or is there an error in my methodology? EDIT: If anyone using a Lynx interface wishes to try this, I have Memory Read Multiple and Double Buffer Output both unchecked in the Lynx Mixer and ASIO control panel.
__________________
Pro Tools HD Native 2023.12 | HD Omni | iMac Pro 10-Core 3.0 GHZ - 64 GB - 2 TB - Vega 56 | Ventura 13.5 | Sonnet Echo Express SE II for PT HD Native PCIe and UAD 2 Quad PCIe card | Avid Dock and S1 |
#2
|
|||
|
|||
Re: Standard way to measure monitoring latency?
Seems right to me. You we t through the hardware buffer twice, which is 64 samples x2 or 128 samples. The other 24 samples must be how long it takes your ad / da's to work.
|
#3
|
||||
|
||||
Re: Standard way to measure monitoring latency?
Thanks Brandonx1. That explanation makes sense.
One further question to those with experience with HD Native: On my system, as I increase the buffer size, the latency will worsen. Is this also true with an HD Native card? Or does HD Native somehow allow a low latency to be used irrespective of buffer size? I am trying to understand if there will be a significant benefit to me from going to a HD Native system as opposed to just purchasing the CPTK. Sorry for the noob questions.
__________________
Pro Tools HD Native 2023.12 | HD Omni | iMac Pro 10-Core 3.0 GHZ - 64 GB - 2 TB - Vega 56 | Ventura 13.5 | Sonnet Echo Express SE II for PT HD Native PCIe and UAD 2 Quad PCIe card | Avid Dock and S1 |
#4
|
||||
|
||||
Standard way to measure monitoring latency?
Unless you use the low latency monitoring option, I can't see how this would differ.
__________________
Can you please send yourself over fiber to go spam some other forum? Darryl Ramm |
#5
|
||||
|
||||
Re: Standard way to measure monitoring latency?
groundcontrol,
Are you saying that with direct monitoring, the experiences are similar, but with low latency monitoring (LLM), there will be a difference? My understanding from the Reference Guide is that on HD Native one can freely specify the outputs that LLM applies to whereas on PT 10 that is fixed at outputs 1&2. Is there more to the story than that, or was that the point you were making?
__________________
Pro Tools HD Native 2023.12 | HD Omni | iMac Pro 10-Core 3.0 GHZ - 64 GB - 2 TB - Vega 56 | Ventura 13.5 | Sonnet Echo Express SE II for PT HD Native PCIe and UAD 2 Quad PCIe card | Avid Dock and S1 |
#6
|
||||
|
||||
Re: Standard way to measure monitoring latency?
Tom, Shane, and Darryl, two of the three of you have HD Native systems in your sigs ... may I kindly ask if I am on the right track with this?
I'm trying to decide between CPTK or HD Native. Both will get me increased tracks and voices, VCA masters, disk caching, auto fades, etc. These are what I am primarily intersted in, although better latency performance is always of interest to me too. I trying to understand what else HD Native will offer me. To the best of my understanding, it involves higher I/O counts (does not apply to me), the ability to use an Avid HD interface (does not apply to me - I am sure they are good, but my Lynx is good enough for me), and possibly lower latency when monitoring through PT (although my testing above makes me wonder). That last one would be important for me if the latency is low even with a larger buffer. Is that pretty much it, or am I missing something? Thanks!
__________________
Pro Tools HD Native 2023.12 | HD Omni | iMac Pro 10-Core 3.0 GHZ - 64 GB - 2 TB - Vega 56 | Ventura 13.5 | Sonnet Echo Express SE II for PT HD Native PCIe and UAD 2 Quad PCIe card | Avid Dock and S1 |
#7
|
|||
|
|||
Re: Standard way to measure monitoring latency?
Quote:
PT automatically compensates for the playback buffer setting. On my system (SSL Alpha-Link AX) I get 81 samples round trip - regardless of sample rate or playback buffer setting. Remember, of course, that a given sample delay size still represents different latency times (in milliseconds) depending upon the sampling rate. But the size of the physical buffer (in samples) does not change. -Dan PS: Curiously, if I run the exact same test with PT Delay Compensation completely disabled, I get a roundtrip of 82 samples. Its an empty session with no plugins, so I'm not sure where the extra sample comes from. Weird - but then, I always have delay comp on.
__________________
Pro Tools 2020 | Mac OSX 10.15.x RME Ray-DAT | SSL Alpha-Link AX 2019 Apple Mac Pro 16 core / 96GB |
#8
|
||||
|
||||
Re: Standard way to measure monitoring latency?
Quote:
This is why I am performing a DA out of the DAW (should be compensated), then an AD into a record armed track, then a DA back out from that same track, and finally an AD back in to another track (this should also be compensated). My thinking is that the monitoring latency will be compensated for on the first and fourth conversions and that it will not be compensated on the two in the middle (LLM is off and delay compensation is on). I am still not entirely sure that what I am measuring is what I am trying to measure, however. That is why I was wondering if there was a standard, generally used method of measuring the latency of monitoring through the DAW. The other thing I am wondering is if HD Native handles this differently. Thanks to all who have replied thus far.
__________________
Pro Tools HD Native 2023.12 | HD Omni | iMac Pro 10-Core 3.0 GHZ - 64 GB - 2 TB - Vega 56 | Ventura 13.5 | Sonnet Echo Express SE II for PT HD Native PCIe and UAD 2 Quad PCIe card | Avid Dock and S1 |
#9
|
|||
|
|||
Re: Standard way to measure monitoring latency?
you need a 2 channel Oscilloscope and a plus generator
take the output of the generator and mult it to ch1 of the scope and inout one of the interface take output one from the interface to ch2 of the scope adjust the time scale of the scope to easily read the difference between ch1 and ch2
__________________
... "Fly High Freeee click psst tic tic tic click Bird Yeah!" - dave911 Thank you, Craig |
#10
|
|||
|
|||
Re: Standard way to measure monitoring latency?
Quote:
another benefit is automatic insert compensation. this is huge for me as i mix with a lot of hardware. i could never get the compensation as accurate previously. things like parallel compression or hardware on the drums other than the master was a phase disaster. as stated above, LLM is only for a pair of channels. that is fine for me as i never need it unless i am doing a punch in and i am way into the mix. you can punch in with LLM at higher buffer settings. i typically track at 64 and will leave it there till i am a ways into the mix. most of the time i can punch in with a moderate mix going without LLM and have no issues. i will also say that i am floored with the new 16x16 i/o. i came from an Lynx Aurora and while the different is not huge, the midrange was quite different for me, the Aurora felt scooped in comparison and translation on my mixes has improved quite a bit. seating vocals became a bit more simplified. i did not feel much noticeable difference in the highs and lows. dealing with the I/O setups is far simplified as well. previously had Lynx I/O software, RME total mix software, the PT I/O setups. it could be a PITA!! so nice not dealing with any of that now.
__________________
pro-tools-pc.com TRASHER Pro Tools Utility(updated 4-11-2024) HD Native, Avid 16x16, Eleven Rack, Focusrite Clarett 8preX, UA Quad Apollo TB. Intel I7 9900k Win 10 |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Low Latency Monitoring in Standard Pro Tools 9? | Dizzi45Z | Pro Tools 9 | 20 | 02-06-2012 08:55 PM |
LCD TV measure latency. | Mark Norup | Post - Surround - Video | 20 | 10-26-2009 05:55 AM |
how do you measure Latency properly? | skylab001 | 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Mac) | 7 | 10-30-2005 12:27 PM |
Measure latency? | Mululu | 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Mac) | 0 | 09-21-2005 02:24 AM |
Best Way to Measure MIDI Latency? | tomhartman | Pro Tools TDM Systems (Mac) | 4 | 04-14-2002 08:36 PM |