Old 08-08-2010, 01:40 PM   #1
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 5,483
Default Loopback testing, seriously confused...

New laptop (Dell Precision M4400), new OS (W7x64 pro), new drivers (ASIO4ALL 2.10, Lexicon Lambda ASIO driver 2.6.0), new Reaper version... so I decided to spend a few nights doing loopback testing. And I'm now seriously confused. Hopefully someone can shed me some light on this whole mess...

So, I connected the left output of my Lexicon Lambda audio interface to its left input. I set up a click source in a Reaper project and I recorded the clicks. Then I used Reaper to count the number of samples that the recorded audio was behind (and before!) the output clicks. This I did with both "Use audio driver reported latency" on and off. From the recorded audio I took readings at two markers, that's why the entries look like 771/770 in the table below; at one marker the latency was 771, at the other it was 770. ASIO4ALL has internal buffers that can be set, and I measured ASIO4ALL both at 0 as well as at 32 sample buffers, that's why there are two sub-columns there. Negative numbers in the table below means that the recorded audio came before(!) the output click.

I learnt how to do this from here http://forum.cockos.com/showthread.php?t=48884, btw. When I was already at it, I decided to test earlier drivers as well, all HW exactly the same only drivers un- and re-installed.

My results:
Code:
      |          ASIO4ALL 2.9             |     Lambda ASIO 2.5.0             |
      +-----------------------------------+-----------------------------------+
      | use audio driver reported latency | use audio driver reported latency |
      +-----------+-----------------------+----------------+------------------+
      |    off    |        on             |      off       |         on       |
      |           +-----------------------+                |                  |
      |           | latency compensation  |                |                  |
spls  |           |    0+0         32+32  |                |                  |
------+-----------+-----------------------+----------------+------------------+
64    |  827/827     259/259      195/195 |   1097/1097            461/462    |
128   |  771/770      74/74       -32/11  |    983/983            -104/-99    |
256   | 1027/1026     74/74       508/508 |   1074/1075           -535/-535   |
512   | 1539/1539     74/74       652/653 |   1695/1696          -1486/-1486  |
------+-----------------------------------+-----------------------------------+

      |          ASIO4ALL 2.10            |     Lambda ASIO 2.6.0             |
      +-----------------------------------+-----------------------------------+
      | use audio driver reported latency | use audio driver reported latency |
      +-----------+-----------------------+----------------+------------------+
      |    off    |        on             |      off       |         on       |
      |           +-----------------------+                |                  |
      |           | latency compensation  |                |                  |
spls  |           |    0+0         32+32  |                |                  |
------+-----------+-----------------------+----------------+------------------+
64    |  465/465     117/117        8/8   |   1046/1047            282/283    |
128   |  768/768     288/288      184/186 |    776/777             164/165    |
256   | 1024/1024    248/248      184/184 |    927/928            -716/-714   |
512   | 2108/2108    248/248      184/184 |   2290/2290          -1487/-1488  |
------+-----------------------------------+-----------------------------------+
Lots of questions now... I guess the main one is: Does the numbers show the Lambda ASIO driver to be as bad as I think it does? I mean, the numbers are all over the place, they are not nearly as consistent as for ASIO4ALL. With "use audio driver reported latency" off, the numbers should really increase with increasing buffer, right? With ASIO4ALL you can see the buffer size in how the numbers grow; not so with Lambda ASIO.

Also, does the negative numbers with "use audio driver reported latency" on, really mean that the Lambda ASIO driver reports incorrect lateny to Reaper?

How come ASIO4ALL 2.10 reports much worse latency than did 2.9? Is my measurements reason to talk to the dev?

Neither driver work very well at 64 samples buffer?

Computer spec:
Dell Precision M4400, 2 Core Duo T9600 2.8GHz, 2x2GB DDR2 RAM, 6M L2 Cache, 1066MHz FSB, Seagate Momentus 7200.4 SATA 3Gb/s 250-GB 16 MB cache, W7pro x64
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 10:32 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.