Stereo Tool https://forums.stereotool.com/ |
|
Stereo Tool 9.71 BETA https://forums.stereotool.com/viewtopic.php?f=14&t=30748 |
Page 1 of 17 |
Author: | hvz [ Tue Jun 08, 2021 5:07 pm ] |
Post subject: | Stereo Tool 9.71 BETA |
Latest version - Beta 046 (August 28 2021) Downloads
Changes/bugfixes
PortAudio recently released a new version (the first official release in 4 years), and because some recent Windows updates caused issues for some people (certain hardware/driver/Windows combinations) we upgraded to that new release for version 9.70. Since then, some bugs have been found in the new code, which have been fixed. The new version has not officially been released, but when reading through the change log it seemed to be a good idea to update to the latest version. MicroMPX+ is a new, for now experimental version of MicroMPX for lower bitrates (just like AAC+, MP3Pro etc.). This still has most of the advantages that MicroMPX has at higher bitrates, but (obviously) at a bit lower bitrates. Testers needed. (*) As mentioned above, due to issues with PortAudio versions and recent Windows updates, it appears that some sound cards only work at very specific I/O block sizes. Sometimes these are different than those that are suggested by the driver! Because of this, the default block sizes can now be overridden. Typical values that appear to work are often powers of two (256, 512, 1024, 2048, 4096), but in some cases also multiples of the number of samples per millisecond (320, 441, 480). We hope to either find values that _always_ work, or a very limited set of values that cover all sound cards. Please report back to us which values work for your sound card. We need the following info: - sound card type - input or output - sample rate - block size(s) that work). TODO's
Quote:
Latest version - Beta 045 (Aug 26 2021)
Downloads Changes/bugfixes
Quote:
Latest version - Beta 040 (Aug 22 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 039 (Aug 19 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 037 (Aug 13 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 034 (Aug 8 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 030 (July 28 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 029 (July 27 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 028 (July 21 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 025 (July 16 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 020 (July 7 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 015 (June 18 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 014 (June 15 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 013 (June 15 2021)
Downloads
Changes/bugfixes
Quote:
Latest version - Beta 011 (June 8 2021)
Downloads
Changes/bugfixes
|
Author: | MrKlorox [ Tue Jun 08, 2021 5:19 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Regarding buffer configurations: Somebody on the Breakaway forum posted some thoughts about this subject a while back. Some points raised might apply. https://ceaforum.net/?topic=fine-tuning ... -and-rates Quote: 3) Buf Size/Buffers
Would a detected jitter percentage be an appropriate addition to the UI?As suggested above, this is the size of each workspace and the number of them that BAE needs to buffer the input from Winamp or other software, process it and pass it to the output device. If you are processing music which was ripped from a CD (sample rate 44.100 kHz), I would focus on a buffer size of multiples of 441 (ten times the sample rate). There are a bunch of technical reasons why, but it stands to reason that samples are going to “fit” better if the buffer is aligned in some way with the number of music samples taken. I have often seen 882 yield a better result, but the theory still holds – it’s a multiple of the sample rate. Similarly, if you have DVD quality audio sampled at 48kHz, I would experiment with 480, 960 etc This is probably where most of the tweaking is needed. You might find an input buffer of 441/3 will yield a low input jitter percentage, but the same sized or number of output buffers might produce a high output jitter percentage, so you might end up with 441/3 in, 441/4 out or 441/3 in, 882/2 out. I'm excited to mess with the speech detection. |
Author: | hvz [ Tue Jun 08, 2021 5:34 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Quote:
Regarding buffer configurations: Somebody on the Breakaway forum posted some thoughts about this subject a while back. Some points raised might apply.
This kinda looks like what I found. The weird thing that I saw was that if I open my webcam microphone for input (it only supports 32 kHz) and select ANY value that's not a multiple of 320, I get that number of samples but with zeroes in it. So the number of samples is completely wrong. So for example, if I select 340, I get 320 valid samples and 20 0's. The audio continues where it stopped in the next packet, so in fact I would get 34 kHz of data with chunks of silence instead of 32 kHz. What I don't know (yet) is whether this is some weirdness in my sound card, or something that more sound cards do. There is no such thing as jitter in this case (I'm not even sure what "jitter" really means in BAE).https://ceaforum.net/?topic=fine-tuning ... -and-rates Quote: 3) Buf Size/Buffers
Would a detected jitter percentage be an appropriate addition to the UI?As suggested above, this is the size of each workspace and the number of them that BAE needs to buffer the input from Winamp or other software, process it and pass it to the output device. If you are processing music which was ripped from a CD (sample rate 44.100 kHz), I would focus on a buffer size of multiples of 441 (ten times the sample rate). There are a bunch of technical reasons why, but it stands to reason that samples are going to “fit” better if the buffer is aligned in some way with the number of music samples taken. I have often seen 882 yield a better result, but the theory still holds – it’s a multiple of the sample rate. Similarly, if you have DVD quality audio sampled at 48kHz, I would experiment with 480, 960 etc This is probably where most of the tweaking is needed. You might find an input buffer of 441/3 will yield a low input jitter percentage, but the same sized or number of output buffers might produce a high output jitter percentage, so you might end up with 441/3 in, 441/4 out or 441/3 in, 882/2 out. I'm excited to mess with the speech detection. For Stereo Tool itself it would make much more sense to read and write in powers-of-two, since the processing itself happens in powers-of-two. And for many other sound cards, the value suggested by the driver is actually a power-of-two. Just not for my webcam microphone. I could just follow the driver suggestion, and that usually works. But it appears that that's not always the case. Which is why I want more info. Also, I can't think of any logical reason why multiplies of 441, 480 etc would make more sense than anything else, unless the sound card driver itself wants to send data per millisecond instead of per so-many samples. It might indeed work that way for some sound cards though. |
Author: | Daniel1988 [ Tue Jun 08, 2021 5:39 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
I've already reported this (I don't know if this was already known in BETA 008. AGC gain reduction is not visible on the AGC meter). 011 is still the same |
Author: | hvz [ Tue Jun 08, 2021 7:02 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Quote:
I've already reported this (I don't know if this was already known in BETA 008. AGC gain reduction is not visible on the AGC meter). 011 is still the same
Thanks! This is one of those things that are "under development" right now. Some large rewrites of the web interface code for SST, STXtreme and MicroMPX are underway, and one of the changes broke this in the native interface.
|
Author: | Bojcha [ Tue Jun 08, 2021 9:22 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
I already said i first tested with VAC and best options are 441 and 882 for 44.1kHz and 480 and 960 for 48kHz. Same as my USB headphones and my realtek onboard soundcard. However i found and tested that with BA One, where it shows small/acceptable amount of Jitter. Hans, You say your camera mic is 32kHz, to me it's logical that it works good only with 320 samples and 640 should be good or even better. In BaOne for sample rates you only have selection 44.1 and 48kHz (HD).. so Block Size List is restricted to only those numbers, multiple of 441 and multiple of 480. There is lower or standard numbers like 256 or 384 or half 480 but what i saw till now best/minimal block size for 44.1kHz-441 or double that, for 48kHz-480 or double that, or even 4*.. etc. There is also lower numbers in list but there is no 320 for example since it's not possible to set 32kHz. So, you can make it standard list (steps) multiply of 441 and 480.. and check box that enables any number. Sometimes for 48kHz 240 block size can work just fine. I also asked for jitter reading if possible. Easier to find lowest stable block size. What is jitter? Someone think it's myth because nobody can heat jitter error 10% or even more, but it's kinda audio distortion caused by clock deviation. So 0% Jitter is ideal but it's also ok at 4%.. 8% or more (error) since it's still not audible. Atleast I understand it like that. Gui errors, still that compressor meter in old agc is acting strange. About Timer Resolution. ST should fix it's usage at some fixed number, 1.000ms probably best number. Exactly what that Axia updated driver done. But to make that to work ST should enforce windows to same 1.0000ms timer resolution when it's started, since it's not permanent settings in windows. ISLC is example software that can do that. Also ST starts from 3/4 buffer again in this beta (mme used). Buffer bar is also bugged, sometimes settles to middle sometimes not. |
Author: | hvz [ Wed Jun 09, 2021 2:24 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
1 - I don't know if the behavior in BA One and Stereo Tool is the same, so I don't know if these 441/882 etc numbers make sense. It might, but basically, there's no good reason (except if drivers made things this way) to use the number of samples per millisecond as a basis for anything. In most cases, and also with how modern CPU's work, powers-of-two are much more efficient. For example with ASIO, which is the most optimized I/O layer in Windows, I have never even seen drivers that allow these types of values. 2 - I don't _see_ any jitter when using different values. The only thing that I see is that some sound cards with some sizes insert zeroes (silence) in the audio. 3 - AGC meter: Known issue 4 - Timer resolution for what exactly? We have never assumed anything about timer resolutions, also because older Windows versions already used 16 ms instead of 1 ms. So we have always assumed any timers to be not more accurate than 16 ms. (Something in PortAudio might assume it to be 1 ms though). 5 - Without changing any settings? That's odd... How big is your output buffer? |
Author: | Bojcha [ Wed Jun 09, 2021 7:44 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Quote:
5 - Without changing any settings? That's odd... How big is your output buffer?
I keep it 50ms tried 100 and 200.. random bar positions and fuller buffer then normal at start. My input sync need to go for from x1.0004 at start.Ahh.. also VAC mode "Cable range" is more adequate then "Cable format" now. It was vice-versa few betas ago. |
Author: | MrKlorox [ Thu Jun 10, 2021 3:40 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Seems like the LQLL won't push audio for me to monitor when the dehummer is enabled. Should I send a preset? 64bit windows stand alone, 9.71 b11 |
Author: | hvz [ Thu Jun 10, 2021 5:00 pm ] |
Post subject: | Re: Stereo Tool 9.71 BETA |
Quote:
Seems like the LQLL won't push audio for me to monitor when the dehummer is enabled. Should I send a preset? 64bit windows stand alone, 9.71 b11
Yes, please do. I just tried it here and there were no issues (there were a few beta's ago, but those issues were solved before BETA011). Please send your .ini file in a PM or to support@thimeo.com, that's usually the best way to reproduce it. You can find it under C:\Users\username, it's normally called something like stereo_tool.ini (depending on which exact version you're using).Before I look into it, some questions: - What does the buffer meter of LQLL indicate? - Does the normal output continue normally? |
Page 1 of 17 | All times are UTC+02:00 |
Powered by phpBB® Forum Software © phpBB Limited https://www.phpbb.com/ |