Stereo Tool https://forums.stereotool.com/ |
|
Stereo Tool 7.24 LINUX test https://forums.stereotool.com/viewtopic.php?t=4846 |
Page 5 of 19 |
Author: | 2Sense [ Sat Jun 15, 2013 4:35 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Quote:
Also: Are both input and output red? And are they not in the old version (7.23)?
I think just the output gave an error. Pipeline input was blue if I recall correctly. Hopefully I will get a chance to test beta 12 soon. |
Author: | hvz [ Sat Jun 15, 2013 8:50 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Quote: Quote:
Also: Are both input and output red? And are they not in the old version (7.23)?
I think just the output gave an error. Pipeline input was blue if I recall correctly. Hopefully I will get a chance to test beta 12 soon. About 012: I have not yet tested it myself either because my pc with ASIO is in my office - will try it Monday. But what I changed was definitely a mistake that could lead to strange behavior in the buffers. |
Author: | Bojcha [ Sun Jun 16, 2013 3:54 am ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Quote: I don't know if this solves anything, but there was an error in the ASIO block size calculation in previous versions that's now fixed:
Tested 30 secs on Air...http://www.stereotool.com/download/ster ... 24-012.exe This one is even worse. Constantly stuttering even if i turn off composite and declipper. Asio buffer fastly goes up and down. Till now, Last good beta is 009. |
Author: | hvz [ Sun Jun 16, 2013 1:13 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Looking through the code now, comparing 009 vs. current version. - AsioIO.cpp - no problematic changes - COMMON.cpp - suspicious change when setting sample rates, adding logging to check... --> Checked, is ok. Found SendToNormal twice??? --> That's ok, one with and one without multithreading. - radio_tool_SA.cpp - difficult. Need to re-check if there are still issues. Only Playback thread. - stereo_tool.cpp - Bypass mode should be ok for FM and normal, NOT for low latency. WantsUpdate() check must be improved to switch to/from ASIO immediately. --> WOW! I was looking through this radio_tool_SA.cpp file in more detail, and I found that I added a Sleep() if any of the inputs doesn't return ok. That's good - this avoids 100% CPU load (and unresponsive GUI) in that situation. However, it also did this if there was no 2nd (SCA) input enabled! (Actually, I think now that it should only do this if there is also no valid output available). I will run a new build without this line, no guarantees though - I cannot test this until I get back in my office tomorrow morning. |
Author: | hvz [ Sun Jun 16, 2013 4:26 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Untested (see previous posts)! http://www.stereotool.com/download/ster ... 24-013.exe |
Author: | Bojcha [ Sun Jun 16, 2013 4:34 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
- startup with bypass ON still gives no sound... - asio, seems ok now! |
Author: | hvz [ Sun Jun 16, 2013 4:42 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Quote: - startup with bypass ON still gives no sound...
Ah great! That's good news because testing the ASIO stuff requires my old desktop pc and - if performance is an issue - a build that takes more than half an hour for every change I make (even if it's just one line of code). Other things are much easier to test, and I can test them on my laptop.- asio, seems ok now! Biggest issues now: - Audio in Bypass mode - Changing sample rate (I expect it to fail, at least in non-ASIO mode) FIXED - Sound card open sequence, this seems to be a problem on the Marian 8 card. I will check the old sequence and make the new one identical. WAITING FOR FEEDBACK - I'll check the CPU load tomorrow. A few days ago it was a lot higher with the new code, not sure if that's still the case. EQUAL NOW |
Author: | hvz [ Mon Jun 17, 2013 10:48 am ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Quote: I think just the output gave an error. Pipeline input was blue if I recall correctly.
Could you check if this was correct in previous versions? (7.23 or older).I just checked the order in which the inputs and outputs are opened in 7.23 and 7.24, and this is the order: 7.23: 1. FM output at high sample rate 2. Low Latency output 3. Input 4. Normal output 7.24: 1. FM output at high sample rate 2. Normal output 3. Low Latency output 4. Input The point is, that they both start with FM output - so at least for the FM output the behavior should be identical. (This was different a few versions ago, so maybe the problem is fixed now?) |
Author: | hvz [ Mon Jun 17, 2013 11:39 am ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Bojcha, I've tested Bypass with and without ASIO and in both cases I do get audio on both FM and Normal output... Are you doing something special? 2Sense: I've been doing some tests here and it looks like problems occur on Windows versions after XP when the sample rate set in the driver doesn't match the one set in Stereo Tool. If both are identical I can open input and output simultaneously (although I'm getting a strange sound, still looking at that), if they are not identical I can open input OR output but not both. Edit: It works sometimes with the sample rates set at the same value, but not always - and I cannot really figure out when it does and when it doesn't. Ok, if this is not WORSE than before (I have similar problems using version 7.23 on this test system), I'll leave it for now - the new audio interface makes it easier to add other outputs such as Kernel Streaming in the near future. |
Author: | 2Sense [ Mon Jun 17, 2013 1:33 pm ] |
Post subject: | Re: Stereo Tool 7.24 LINUX test |
Will get the pc back in the next hour or so and will be able to test after that. Sorry for the delay in getting back to this, the computer has been in service so it's been a little challenging to try beta versions out. |
Page 5 of 19 | All times are UTC+02:00 |
Powered by phpBB® Forum Software © phpBB Limited https://www.phpbb.com/ |