3.6 When ?

[Sascha Franck]

OK, one last shot: this functionality is already supported.

No, it's not. USB access is only supported at the start and end of whatever signal paths. It might be technically challenging to place USB sends/returns somewhere in the middle.

 

[mbenigni]

No, it's not. USB access is only supported at the start and end of whatever signal paths.

I was speaking more broadly: USB audio is available. People are using it. In some cases, people are using it as an FX loop. They're just jumping through some hoops to do so.

No, it's not. USB access is only supported at the start and end of whatever signal paths. It might be technically challenging to place USB sends/returns somewhere in the middle.

Yes, it might be technically challenging. Or not. THAT WAS MY QUESTION.

But USB audio is subject to latency at those signal paths right now. (I know we can agree on this because you've been quantifying it and debating about it for two pages.) Assuming things would be any worse, any less useful, or any more difficult for Line 6 support if the Loop block exposed those audio channels is, for the moment, purely speculative.

 

[JiveTurkey]

Dear baby jesus @Digital Igloo come save us with usb routing

 

[Sascha Franck]

because you've been quantifying it and debating about it for two pages

No, I have not been debating at all. I have posted numbers, nothing else. Don't shoot the messenger.

Assuming things would be any worse, any less useful, or any more difficult for Line 6 support if the Loop block exposed those audio channels is, for the moment, purely speculative.

That is correct.

 

[paisleywookiee]

This is always a sore one, its technically correct, but that's usually also technically correct with windows too.

Put an RME card in your mac and see how much you like no drivers vs actually installing the drivers

That's entirely device-dependent. Some devices don't require additional drivers. Some do. It depends.

DAW reported latency is a very different thing than round trip latency

Except when the DAW reports round trip latency.

 

[pipelineaudio]

And fwiw, in case anyone is interested in all this and up for reading through a most monstrous thread, here we go:

Audio Interface - Low Latency Performance Data Base - Gearspace

Hey All, I am sure all of us have seen posts over the years where end users have purchased a new audio interface and have experienced varying degrees o

gearspace.com

DAWBench’s Vin Curigliano is the other voice in the wind telling how far off the reported latency numbers are from what measured latency is. I can’t for the life of me figure out how in the world Logic would know the converter latency when you could plug in any old converter to an ADAT port.

 

[pipelineaudio]

Except when the DAW reports round trip latency.

How does the daw know what converter you plugged into it’s digital input? The DAW cannot know the round trip latency. It can only reflect what the drivers report. Hidden safety buffers are also not reported to the DAW.

 

[Sascha Franck]

I can’t for the life of me figure out how in the world Logic would know the converter latency when you could plug in any old converter to an ADAT port.

Correct - these can't be reported properly. And hence aren't. But pretty much all recent interfaces report the correct values for their onboard converters. Really, you can check that for yourself. And it's not just Logic, it's all DAWs reporting proper numbers.
Once you connect external converters, you will have to re-measure and correct things manually. This is one of the reasons tools such as the Oblique one exist. And a reason why, say, Logic offers an option to adjust the recording offset globally (which, fwiw, also works great in case interfaces don't report their latencies properly).

 

[Sascha Franck]

Hidden safety buffers are also not reported to the DAW.

They are. Since decades already.
Feel free to ask Mathias Carstens (CEO of RME). We (some friends and me) once had an interesting conversation with him regarding these very issues at a Frankfurt Musikmesse.

 

[pipelineaudio]

it's all DAWs reporting proper numbers.

That is DEFINITELY not true. For instance, I know, beyond a shadow of a doubt that REAPER uses the driver reported latency, and that number matches what is shown in Cakewalk and Cubendo.

 

[pipelineaudio]

They are. Since decades already.
Feel free to ask Mathias Carstens (CEO of RME). I once had an interesting conversation regarding these very issues at a Frankfurt Musikmesse.

That's because RME are nearly the only honest ones, and have always reported their actual latency

 

[Sascha Franck]

That is DEFINITELY not true. For instance, I know, beyond a shadow of a doubt that REAPER uses the driver reported latency, and that number matches what is shown in Cakewalk and Cubendo.

Huh? In case it matches the numbers, that's proving what I just said.

That's because RME are nearly the only honest ones, and have always reported their actual latency

As said: Check for yourself with whatever interfaces you have at your disposal. These days, most report correct numbers (and yes, it has been very different in the past).

 

[pipelineaudio]

Huh? In case it matches the numbers, that's proving what I just said.

They don't report "hidden" safety buffers to the DAW, and the DAW numbers do not show them. That's why they're hidden

As said: Check for yourself with whatever interfaces you have at your disposal. These days, most report correct numbers (and yes, it has been very different in the past).

You are right, I could be very well tilting at windmills, I'll run some loopback on some TheSycon drivers. I have some focusrites at home with them. At work i only have RME, who report the correct number, however, that is ONLY for the interface buffers, and does not include converter latency as these are ADAT only, so again, no the DAW does not necessarily give you the right numbers.

But let's see what the focusrite drivers have to say

 

[Sascha Franck]

They don't report "hidden" safety buffers to the DAW, and the DAW numbers do not show them. That's why they're hidden

They do. Otherwise reported numbers wouldn't correspond with measured ones. They're only called "hidden" because usually there's no way the user has access to their settings (there's some exceptions, though, some Motu interface for example has something like a "risky" safety buffer setting you can access).

Fwiw, might very well be that Thesycon driven interfaces are a victim of wrong reported numbers more often as companies may just not care at all to do some required adjustments to the rather generic driver code.

 

[pipelineaudio]

When did they start showing hidden safety buffers? Certainly not in 2017, last time I really checked with an actual cable. Its good news if they started finally.

 

[Sascha Franck]

When did they start showing hidden safety buffers?

I don't know. It's just some information I got from another forum. Doesn't seem to be like that for all their interfaces, either. And apparently adjusting things there comes with a warning, too ("only do that in case you have a very powerful system" or so).

 

[mbenigni]

OK, so it’s a latency thread now.

 

[Sascha Franck]

OK, so it’s a latency thread now.

Thread derailment seems to be something new to you then?

 

[pipelineaudio]

Ok, I just put a cable on Helix floor and am posting up the pics as I speak, I need to do the math but I could very well be eating my words! Below is a REAPER screencap. Given what we know about certain DAWs, we know these will be the same numbers for Cakewalk (at least the Bandlab version) and Cubase/Nuendo, and Vegas, probably the rest as well, but not sure.

Line 6 Drivers are reporting 3.9 in and 8.6 out at 64 samples at 48khz for a reported 12.5ms (or roughly 600 samples) round trip latency

And I get.......600msec!!! Holy crap. Line 6 was honest with their reported latency, even including the converters. @Sascha Franck thank you I learned something and got some renewed faith in Humanity. Let's see what happens when I test the TheSycon.de drivers

Words eaten, and when Steven Pinker said things are getting better, he is sometimes right

@Jim Roseberry check this out!

 

[paisleywookiee]

I can’t wait to post my fake latency with my Audient tonight. I bet they’re lying liars too!