Auditon 3.01 input latency

Ok….I did a similar test, and took four screenshots, with smaller and bigger zooms, at the 50 seconds point of the recorded tracks.
(Lets see if  I got them all attached….)

As you can see from the picture “latencytest_infra_zoom_verybig.jpg”, the latency of the first recorded track is 6 ms. The track is recorded with the internal metronome and a 64 samples buffer.
Track 2 is recorded from track 1 with a 64 samples buffer, and the latency is around 10-11 ms from the 50.000 seconds point. Ok, that’s locigal. The latency increases with that 5-6 ms everytime you record the next track from the previous track (in my case).
Track 3 is also recorded from track 1 with a 512 samples buffer, and here we can see the real problem. The latency is now around 43 ms from the 50.000 seconds point.

I can’t zoom in your picture enough to see the milliseconds, but it looks pretty good = not much latency.
But could you take one more screenshot with a bigger zoom? ( I mean, first zoom in the 50 seconds point in Audition, and then take the screenshot)

But ok, all ready at this stage it seems that the explanation must be that the Edirol has some kind of latency compensation function in itself. With the max asio buffer settings there should be a notable delay, if no compensation is used.

It's pretty clear that the 2048 samples buffer with wdm-drivers causes a latency so big that, even if there is a latency compansation in the Edirol, it's too much. And probably the latency compensation works only with Asio-drivers.(?)

This also explains why I never had problems with my Lexicon Lambda - Asio drivers - AA3. I think the Lambda also has this compensation function, like your Edirol. And that's missing from my new Infrasonic-card.
I can not find any information about this issue on the net. I've been looking for at least something about "Input latency compensation in soundcards" but I can't find anything. So this is only an assumption at this moment.
But soon I think I got enough facts, so I can discuss this issue with Infrasonic.

Very much thanks!

P.

No...but here is the "verybig" picture

If I zoom in on the same Audition screen, this is what I see. It looks as if there is no latency at all with the 3 asio tracks, but the wdm latency puts the waveform off the screen.

I don't understand this, and am sure I must be doing something wrong. The Edirol documentation just has the standard statement that a larger buffer means more latency, and I would have thought any compensation would have to be in the recording program rather than the drivers.

If no-one comes along and suggests something, I will test with a different audio interface in the next few days. The fact does remain that I hear no problem when recording guitar. It's just that I'm intrigued that I can't see any latency at all.

Well, I think you should be lucky! The Edirol buffers seems to be great.

When they talk about latency, they usually mean "the listening latency", not this "input latency". I'm shure you get a listening latency, if you put the buffers high and use "Audition mix"-monitoring.
In my opinion, the "input latency" is a thing that the user should not have to worry about. The drivers should fix it, just like your Edirol-driver probably does.
So I think everything is all right in your system.

P.

Hello Folks!

Input latency problem solved! I disabled the Infrasonic PCI-card (totally removed all hard- and software), and connected a Lexicon I-Onix U22 usb-card ( = audio-interface).
The input latency is now 0.8 ms regardless buffersize. All buffersizes from 49 samples to 1024 samples gives the same input latency of only 0.8 ms.This is how it should work!

I don't want to say it it public, but everyone can draw a conclusion of this .

Now I'm gonna record some rock'n'roll!

Peter