ADAT Input Delay -- Demonstration Video

[MakerDP]

Here it is folks... yes you can hear it in this case... whether it matters or not in your situation is up to you.

Just some fun mostly because my PreSonus DigiMAX D8's delay times are not on Matt's spreadsheet. (24 samples at 48k is what I came up with.)

Using this technique you can see exactly what YOUR combo's delay numbers are.

[video=youtube_share;4ZUE86jYcRQ]https://youtu.be/4ZUE86jYcRQ[/video]

Let me know if you have any questions... I may even be able to answer them.

Matt's Spreadsheet: https://docs.google.com/spreadsheets/d/15dUWNZg-aN1248fZ4i7kxRx5Kg44d94oBzsxIMhDKh0/edit?usp=sharing

 

[Bruce_Sokolovic]

Yea, wait until you get 8-12 mics around the whole kit. Thats when the real fun begins! Dying art? Maybe. I wouldn’t do it. Send me a midi track any day of the week and I’m happy. Phase alignment is one reason of many.
As I’m sure you know, this really only matters when you have multiple mics on a single source and as you pointed out, using 2 different sets of converters can throw things off, too. I ran into this with an Apogee ADDA and my interface years ago. Happily, I rarely mic things anymore, and would almost never multi mic for anything.

 

[rodd]

Nice! Thanks [MENTION=46492]MakerDP[/MENTION]. FYI, the link to the spreadsheet is broken. I have a question. In your example it's clear that capturing audio from the same source using an Apollo and an ADAT interface is going to cause issues, and you can adjust for that. How would you find the latency if all of the audio comes from the ADAT interface? For example, I record all of the tracks into Luna (guitar, keys, etc), and then I want to record drums that are mic'd using an ADAT input. I don't think there will be phase issues like you demonstrated, but there should be a delay, I'm just not sure how to calculate it because there's nothing to use as a reference like you have with the Apollo mic. I guess I could loop some test audio out from the Apollo and back into the ADAT interface, but that would be round-trip latency and probably not as simple as halving it.

 

[Bruce_Sokolovic]

Nice! Thanks [MENTION=46492]MakerDP[/MENTION]. FYI, the link to the spreadsheet is broken. I have a question. In your example it's clear that capturing audio from the same source using an Apollo and an ADAT interface is going to cause issues, and you can adjust for that. How would you find the latency if all of the audio comes from the ADAT interface? For example, I record all of the tracks into Luna (guitar, keys, etc), and then I want to record drums that are mic'd using an ADAT input. I don't think there will be phase issues like you demonstrated, but there should be a delay, I'm just not sure how to calculate it because there's nothing to use as a reference like you have with the Apollo mic. I guess I could loop some test audio out from the Apollo and back into the ADAT interface, but that would be round-trip latency and probably not as simple as halving it.

In his test, the latency was behind .5ms or 24 samples of the latency introduced from the apollo, so figuring the Apollos latency plus 24 samples would be the total latency. If you’re recording a kit and use all your ADAT pres, you can still of course run into phase problems due to mic distances (which could have also been corrected in the example given instead of using a plugin). Moving a mic a few cm can fix or cause a phase problem.

 

[rodd]

In his test, the latency was behind .5ms or 24 samples of the latency introduced from the apollo, so figuring the Apollos latency plus 24 samples would be the total latency. If you’re recording a kit and use all your ADAT pres, you can still of course run into phase problems due to mic distances (which could have also been corrected in the example given instead of using a plugin). Moving a mic a few cm can fix or cause a phase problem.

Thanks Bruce, maybe I am overthinking it. It seems like if I am listening to the playback via headphones connected to Apollo while recording drums via an interface connected via ADAT, that this would be different latency than recording the same source via a mic pre in Apollo and an ADAT-connected interface at the same time. I am not sure if there is an effective way to measure this latency, or maybe it would be close enough to do it the same way as in the video? Maybe looping the audio out from Apollo and back through the ADAT interface would show it more accurately?

I understand what you mean about potential phase issues with mic placement, I just meant that there wouldn't be the same kind of phase issues as what MakerDP was demonstrating in the video.

 

[MakerDP]

The key to measuring the difference is to run the exact same source into two or more devices. If you don’t have access to another computer/interface like I did then you can use something as simple as your phone and a breakout cable for your headphone jack. Send the left channel directly to the interface and the right channel through the ADAT. Find a MONO sound source that has a sharp transient… a signal generator app using a square wave would be a great source. Or if you have a mic splitter you could mic a source. As stated using two or mics to measure the delay will introduce its own natural delay so that won’t give you the measurement.

 

[rodd]

The key to measuring the difference is to run the exact same source into two or more devices. If you don’t have access to another computer/interface like I did then you can use something as simple as your phone and a breakout cable for your headphone jack. Send the left channel directly to the interface and the right channel through the ADAT. Find a MONO sound source that has a sharp transient… a signal generator app using a square wave would be a great source. Or if you have a mic splitter you could mic a source. As stated using two or mics to measure the delay will introduce its own natural delay so that won’t give you the measurement.

Thanks, I figured out the error in my logic. Great post! I need to do this.

 

[Matt Hepworth]

Here it is folks... yes you can hear it in this case... whether it matters or not in your situation is up to you.

Just some fun mostly because my PreSonus DigiMAX D8's delay times are not on Matt's spreadsheet. (24 samples at 48k is what I came up with.)

Using this technique you can see exactly what YOUR combo's delay numbers are.

[video=youtube_share;4ZUE86jYcRQ]https://youtu.be/4ZUE86jYcRQ[/video]

Let me know if you have any questions... I may even be able to answer them.

Matt's Spreadsheet: https://docs.google.com/spreadsheets...it?usp=sharing

Nice!

Is that 24 samples with the Twin mk II?
EDIT: sounds like with the Silverface

 

[MakerDP]

Nice!

Is that 24 samples with the Twin mk II?
EDIT: sounds like with the Silverface

Well here's the one thing that confused me about this whole thing... I have the SF and the Twin mk II daisy-chained via TB of course and the D8 is going into the Twin. Three identical signals were coming out of the laptop ... one into the Twin, one into the SF and one into the D8. Now I see in your spreadsheet that the Twin and SF are not perfectly in time but my tests show the phase alignment to be spot-on perfect and the delay to be identical from the D8's signal... which leads me to believe there is some compensation going on between the Twin and the SF at the driver level??? Or the delay only exists via ADAT???

It just occurred to me that my clock source is set to the D8 because it does not have ADAT input only output and I don't have any BNC cables/connectors to set up a common clock for everything... don't know if that would make the Twin and Apollo match or not since they are both clocking off of the D8.

Maybe I'll have to experiment with turning off the SF and just going into the Twin via ADAT...

I've also messed-around in my OBS settings and vastly improved my screen resolution/readability while capturing.

 

[chrisso]

OK, I need to record a snare drum with two exact same mics, placed accurately the same distance from the drum head.
I have been recording drums all 2020 using a mix of Cranborne 500ADAT (into Apollo ADAT input), hardware mic/pres (into Apollo line inputs) and my four Apollo pres.
Of course every recording has phase issues, which one can chose to ignore (because it sounds good), flip the phase, or introduce an IBP.
I haven't ever accounted for ADAT delay. The drums sound good time and I've sent the recordings off to several other professionals and no one has made any comment on sound issues or phase issues.
But I'm definitely gong to test this now.

 

[rodd]

yeah I tried it out last night and can definitely see it. I've often wondered if my time was just off when playing, but now at least I can see that I am not fully to blame (but still mostly). Anyway, I am not sure what to do about it, because in Luna you can only compensate by adding delay in the utility row, and I need to subtract time. I guess I could add time to every other track? Also, are there any free plugins that anyone knows of to measure the latency difference? I'm not completely against buying the Waves plugins, but once I know what the setting is then I won't need to use it again. I checked Matt's spreadsheet (correct link here, [MENTION=46492]MakerDP[/MENTION] yours still throws a 404) and my ADAT interfaces are not on there.

 

[chrisso]

OK, I just tested my Apollo blackface (2017) and my Cranborne 500ADAT with two identically placed mics, one going into a Camden pre > ADAT, the other going into Apollo ch1 mic/pre.
24bit 96khz. The ADAT audio is 0.16ms late, or 15 samples. It's a sound thing not an audible timing thing.
So I guess Luna needs to be able to nudge in individual sample increments?

 

[UniversalAudio]

So I guess Luna needs to be able to nudge in individual sample increments?

it does.

 

[chrisso]

Ah yes, I've been searching for the specific video for half an hour.
I can set a custom nudge amount in the samples tab. In my case, 15 samples. Superb!

 

[MakerDP]

yeah I tried it out last night and can definitely see it. I've often wondered if my time was just off when playing, but now at least I can see that I am not fully to blame (but still mostly). Anyway, I am not sure what to do about it, because in Luna you can only compensate by adding delay in the utility row, and I need to subtract time. I guess I could add time to every other track? Also, are there any free plugins that anyone knows of to measure the latency difference? I'm not completely against buying the Waves plugins, but once I know what the setting is then I won't need to use it again.

Two things here:
1) You can use the "nudge" feature in LUNA to move an audio track forwards or backwards by a specific number of sample.

2) I am not aware of any free plugins that do what InPhase does... doesn't mean they don't exist of course. One way you could do it without InPhase is to get any kind of stereo field analyzer like the Waves PAZ that will show you if you have stuff out of phase. Put it on a bus and send your two tracks in question to that bus. Pan one of the tracks hard-left and the other hard-right. Use the LUNA "nudge" feature to shift the audio of the ADAT track forward in time one sample at a time until you don't see anymore of those out of phase "tails" in the meter. Then you can delete the bus with the analyzer.

I will say though that these days InPhase is my most-used Waves plugin and I have Platinum and some other stuff added in. I use it every time I mix live drums so I can tighten the snare in the stereo overhead mics by sliding the right mic forward in time so the snares match-up as best as I can get it. (Right-mic from drummer's perspective... the one further away from the snare drum.) It's a subtle improvement but to my ears well worth the effort of the less than 5 minutes it takes. I'll then print the results so I can remove the plugin from the session.

Two alternatives to Waves PAZ Analyzer I have that you may also already have:
Melda StereoScope - different graphs but still gives the needed info - Stereo Graph seems to give the most useful display
Izotope Insight (Stereo Analysis preset) - doesn't seem to show as much anti-phase content info as PAZ or Melda

 

[Matt Hepworth]

As of 9.14 or 9.15 (can't recall) these will align with the monitor unit when multiple Apollos are involved thanks to Apollo's new driver.

So, which is your monitor unit?

For users of only one Apollo the chart will be correct. For users of multiple Apollos, use the monitor unit as the reference.

Well here's the one thing that confused me about this whole thing... I have the SF and the Twin mk II daisy-chained via TB of course and the D8 is going into the Twin. Three identical signals were coming out of the laptop ... one into the Twin, one into the SF and one into the D8. Now I see in your spreadsheet that the Twin and SF are not perfectly in time but my tests show the phase alignment to be spot-on perfect and the delay to be identical from the D8's signal... which leads me to believe there is some compensation going on between the Twin and the SF at the driver level??? Or the delay only exists via ADAT???

It just occurred to me that my clock source is set to the D8 because it does not have ADAT input only output and I don't have any BNC cables/connectors to set up a common clock for everything... don't know if that would make the Twin and Apollo match or not since they are both clocking off of the D8.

Maybe I'll have to experiment with turning off the SF and just going into the Twin via ADAT...

I've also messed-around in my OBS settings and vastly improved my screen resolution/readability while capturing.

 

[Matt Hepworth]

OK, I just tested my Apollo blackface (2017) and my Cranborne 500ADAT with two identically placed mics, one going into a Camden pre > ADAT, the other going into Apollo ch1 mic/pre.
24bit 96khz. The ADAT audio is 0.16ms late, or 15 samples. It's a sound thing not an audible timing thing.
So I guess Luna needs to be able to nudge in individual sample increments?

When you're not using an identical signal to each unit reverse the cables and run a second test (mic 2 to converter 1, etc. instead of mic 1 to converter 1) and make sure the difference is the same (meaning you're aligned within 1 sample). If it's not, halve the difference between the two results and that should be your actual delay.

I normally use a DI and send the thru to one unit and the mic out to the other, or I use a splitter.

 

[MakerDP]

As of 9.14 or 9.15 (can't recall) these will align with the monitor unit when multiple Apollos are involved thanks to Apollo's new driver.

So, which is your monitor unit?

For users of only one Apollo the chart will be correct. For users of multiple Apollos, use the monitor unit as the reference.

Ah OK that makes sense... my monitor unit is the Twin mk II.

 

[chrisso]

When you're not using an identical signal to each unit reverse the cables and run a second test

I just saw on Gearspace that Cranborne report 15 samples at 96/24 the same as my result.

 

[Matt Hepworth]

I just saw on Gearspace that Cranborne report 15 samples at 96/24 the same as my result.

Excellent! Is that with the same Apollo version? Each is different, except when combined!