• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Is it possible to add this measurement or is it useless ?

versus666

Member
Joined
Jun 15, 2020
Messages
8
Likes
3
Hi !
Long time lurker I finally decided to register. ASR is a GREAT source of REAL informations, really professional and I doubt to fine enough words to say how much I love that place ! :D
I love the numerous tests and the characteristics measured but as an avid gamer ( everyone have a flaw ;) ) I'm looking for the most responsive hardware and I wonder if all DACs have the same latency, if any. I know there is a way to set latencies in ASIO and ALSA but that's on the software side so do DACs, whatever their qualities, induce any latency by the way they are made, just like mice do, and is it possible to measure it in each reviews ?
 

zermak

Senior Member
Joined
Jun 2, 2019
Messages
372
Likes
251
Location
Italy
This exact question was already asked before and my understanding/conclusions are that you don't have to worry about audio latencies because the main issues will always are the monitor input lag and mostly the pixels rise/fall times (response times) which are way higher time than a possible audio synch problem even with the best 240Hz monitor for gaming (best G2G I saw was 2,6ms and best input lag I saw was something line 2ms but on another monitor).
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
What you're thinking of is called propagation delay. Probably in the few dozen picoseconds range. The travel time in cabling will be more, and is in the nanoseconds range. Not worth bothering with.
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
Hi !
Long time lurker I finally decided to register. ASR is a GREAT source of REAL informations, really professional and I doubt to fine enough words to say how much I love that place ! :D
I love the numerous tests and the characteristics measured but as an avid gamer ( everyone have a flaw ;) ) I'm looking for the most responsive hardware and I wonder if all DACs have the same latency, if any. I know there is a way to set latencies in ASIO and ALSA but that's on the software side so do DACs, whatever their qualities, induce any latency by the way they are made, just like mice do, and is it possible to measure it in each reviews ?

I am not a gamer. But for people that like to make real time musik with there equipment thats a very importent question too.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I am not a gamer. But for people that like to make real time musik with there equipment thats a very importent question too.
Yes but he is referring to hardware latency. Not the more common software latency which can run into enough milliseconds (more than 20) to be a problem for recording.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,986
Location
Riverview FL
Last edited:

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,986
Location
Riverview FL
Ah...

My well documented DAC gives a "common case" number or two:

1592240406527.png



Maybe more than I expected - 59,976 60 samples at 44.1kHz, and 90,240 90 samples at 192kHz at the other end of its range..

I would have imagined a number in the hundreds. Maybe @John_Siau could comment, as this is a DAC2 HGC in my rack.

Equivalent to being another foot and half (or less) few millimeters away from someone/something with which you are trying to synchronize audibly/visually.
 
Last edited:

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,986
Location
Riverview FL
Oops.

Miscalculation recalculated above...
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
I love the numerous tests and the characteristics measured but as an avid gamer ( everyone have a flaw ;) ) I'm looking for the most responsive hardware and I wonder if all DACs have the same latency, if any. I know there is a way to set latencies in ASIO and ALSA but that's on the software side so do DACs, whatever their qualities, induce any latency by the way they are made, just like mice do, and is it possible to measure it in each reviews ?

DACs by themselves don't usually have latency issues but it doesn't come up once you get into stuff with builtin DSP like AVRs or some active speakers.

I think it should be measured on those devices, but not as many people here care.
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
Yes but he is referring to hardware latency. Not the more common software latency which can run into enough milliseconds (more than 20) to be a problem for recording.

I

At the end a lot of systems run with software.
20ms for real time music is fu** much.
But how do let's say dsp's in speakers add to this? For me thats a absolutly legitim question. I enjoy that gamer perspektive.
 

zermak

Senior Member
Joined
Jun 2, 2019
Messages
372
Likes
251
Location
Italy
I am sure you can test it yourself somehow, empirical way, with some audio and video sync tests (ala guitar hero test to compensate for monitor lag; better with high framerate video).
I am using EqualizerAPO with a FIR filter and 131k samples and I can't spot any problem while playing games (or for that matter on those youtube video I was talking about even if I found only 60fps video and my monitor is a 144Hz).
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
Ah...

My well documented DAC gives a "common case" number or two:

View attachment 69085


Maybe more than I expected - 59,976 60 samples at 44.1kHz, and 90,240 90 samples at 192kHz at the other end of its range..
RME specifies latencies between 0.024 ms and 0.14 ms for the AD and DA chips in the RME ADI-2 PRO, depends on filter type and sample rate. The total latency is higher by 11 - 23 samples, resulting in 23 samples at 44.1 kHz = 0.5 ms and 36 samples at 192 kHz = 0.2 ms. So nothing to worry about.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
It's actually an interesting question from a theoretical perspective.

Mathematically ideal D-A reconstruction (Whittaker-Shannon) is symmetric which requires looking ahead. Put differently: each sample has a "ripple" that influences samples both before and after; the "sinc(t)" function. The worst-case scenario is a single full-scale sample with zeros before and after it -- the single-bit impulse often used to compare different filters. Here, the ripple from this sample decays to -96 dB (16-bit resolution) in about 44,000 samples in both directions (forward & backward) in time. So theoretically, at CD quality (44-16) the DAC must have at least 1 second of latency for the ripple to decay into the noise floor.

That said, this latency is purely theoretical and much longer than is needed in practice, where DACs use different methods (like delta-sigma) which gives results nearly the same as the above mathematical ideal, but with much less computation and latency. In practice, latency from most DACs is in the microseconds or less so it's a non-issue. However, some DAC chips provide alternative filters having lower latency. A DAC designer might use one of these for near-real-time applications like recording and monitoring.

PS: my DAC uses the WM8741 chip which provides 5 digital filters. At 44.1k sampling the latency of these varies from 7 to 47 fs. I believe that's "femto-second" which is 1e-15 second. This is less than a single sample which is about 2.3e-5 second.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,550
Location
Seattle Area
To the extent anyone is looking at me to measure, Audio Precision software is not real-time and fully predictable. Nor does it have a simple latency measurement. I have tried to work around these to get some numbers for some reviews but they are not high confidence values.
 

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
3,060
Likes
4,416
Location
France
Hi !
Long time lurker I finally decided to register. ASR is a GREAT source of REAL informations, really professional and I doubt to fine enough words to say how much I love that place ! :D
I love the numerous tests and the characteristics measured but as an avid gamer ( everyone have a flaw ;) ) I'm looking for the most responsive hardware and I wonder if all DACs have the same latency, if any. I know there is a way to set latencies in ASIO and ALSA but that's on the software side so do DACs, whatever their qualities, induce any latency by the way they are made, just like mice do, and is it possible to measure it in each reviews ?
I'd first optimize the overwhelmingly important software side: do you use a CONFIG_PREEMPT=y kernel with a high CONFIG_HZ? Maybe with a less throughput oriented scheduler like MuQSS, PDS or BMQ. Knowing how and when to use nice/ionice is a good idea too.
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,209
Likes
2,675
Location
Salvador-Bahia-Brasil
I'd first optimize the overwhelmingly important software side: do you use a CONFIG_PREEMPT=y kernel with a high CONFIG_HZ? Maybe with a less throughput oriented scheduler like MuQSS, PDS or BMQ. Knowing how and when to use nice/ionice is a good idea too.

is this even possible on windows?
 
OP
versus666

versus666

Member
Joined
Jun 15, 2020
Messages
8
Likes
3
I'm happy to see my question gave so many answers and wasn't ditched. I love that open mind, even is some were a bit off. I admit I was a bit hasty and forgot to talk about other time sensitive uses like interactive music reproduction, with a keyboard or a guitar. And about any time delay effect of the various filters used by the DACs chips. I'm glad it was discussed nonetheless.

I'd first optimize the overwhelmingly important software side: do you use a CONFIG_PREEMPT=y kernel with a high CONFIG_HZ? Maybe with a less throughput oriented scheduler like MuQSS, PDS or BMQ. Knowing how and when to use nice/ionice is a good idea too.
Yes, Xanmod. Nice and ionice are know and used when needed.

So it seems the general consensus, even if, for once, it is not really measured is that the delay between the event in the computer (not screen) and the production of the analog signal exiting the DAC is well below anything a human can perceive.
Thanks you !
 
Top Bottom