Stereo Tool :: Stereo Tool 6.00

I am at 512 samples mode last days. It's interesting challenge to make it work good! Around 25ms real latency with good asio soundcard, L/R output. I still need to measure FM out. With 1024 already can be made good quality preset.

Punch: Like it's not working at 512 (?), but interest thing happens with punch at 512. There is already good punch without punch filter. Maybe even too much, or it's because MB behavior at 512.

Currently only problem in 512 mode is MB and Bass Boost! All other filters seems good.
Basss Boost: Seem good at bass (steepness 30%) but playing eg. 500Hz sine, gives ringing sounds at bass Boost (check BassBoost 'diff').

I found that "steepness" is not good over ~30%, so, acceptable artifacts. But MB not working properly then. Band are too much mixed, and no control over lowest frequencies.
I knoq that Hans explained what's the problem with this very lowe latencies. Math is good. But then i ask myself, How Omnia works with 10ms latency (L/R input > MPX out), Orban, Breakaway, etc. Maybe there is different approach, or there is actually 1024 mode is used but processing latency lowered.

What should be lowest latency at 1024 ? Anyway i'll measure tomorrow.

Q: ST uses 1 core (?) but how many 'threads' ?

Note: For 512 and 1024 mode should be 'always' locked to "never-phase-linear in bandpass. There is just no reason for using phase-linear at this low latency settings.

Author:	Chris [ Tue Jan 25, 2011 6:47 pm ]
Post subject:	Re: Stereo Tool 6.00
Congrats Hans Big diff VERY NICE

Author:	phoenix [ Tue Jan 25, 2011 7:39 pm ]
Post subject:	Re: Stereo Tool 6.00
I am finding it kinda difficult to choose between beta12(the one I've liked so far) and beta22. It's like some tracks selectively sound better using one than the other. beta22, especially enabling 'Prevent artifacts' in MB, sounds like highs have been rolled off ever so slightly, but overall sound is tighter than beta12. Will take me a little more time getting used to this new beta.

Author:	Bojcha [ Tue Jan 25, 2011 11:22 pm ]
Post subject:	Re: Stereo Tool 6.00
1024 mode is very good But 512 still need that 1st band in MB. Somehow it can be adjusted to work anyway. For MB max "steepness" 30-35% = acceptable artifacts (512) I found some nice "phase rotator" settings too for 512/1024 mode. There is strange "tone" at LEFT channel (tested ST dsp), caused by Bass Boost, but not always! If i restart Winamp "tone" is gone.

Author:	hvz [ Tue Jan 25, 2011 11:35 pm ]
Post subject:	Re: Stereo Tool 6.00
Quote: Congrats Hans Big diff VERY NICE Hi Chris, Is this still valid? - Chris: "I still hear diff on Bass and kick between Beta 16 vs Beta 08 on 2048 latency. I hear more bass and kick on Beta 08. Pls check!" It's still on my todo list... Bojcha: I heard a strange tone once when I was playing with the bass boost settings - but that was in an intermediate debug build. I didn't think it would still be there... Anyway, I've heard it too, and I'll check it. What do you mean by: "Somehow it can be adjusted to work anyway."? Phoenix: Which latency are you using? And what type of settings (extreme multiband settings?) Artifact protection should have very little - usually no - effect on highs...

Author:	hvz [ Wed Jan 26, 2011 12:01 am ]
Post subject:	Re: Stereo Tool 6.00
BETA601-023: - Minor performance improvement (no effect on audio) - Reduced Composite Limiter latency from 1.7 ms to 0.9 ms at latencies 512 and 1024. Some (very small) effect on audio. Of course only when Composite Limiter is used, at these low latencies... Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 01-023.exe Stand alone version: http://www.stereotool.com/download/ster ... 01-023.exe VST version: http://www.stereotool.com/download/vst_ ... 01-023.dll VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 01-023.dll Command line version: http://www.stereotool.com/download/ster ... 01-023.exe Linux command line version: http://www.stereotool.com/download/ster ... ETA601-023 [not available] Linux GUI version: http://www.stereotool.com/download/ster ... ETA601-023 [not available] What remains for 6.01: - Multiband: Adjust band 3 (and others?) 'soft limit' behavior at very low latencies. Maybe: Also check clipping (also for band 3, might be re-enabled!). DONE! - 1 hour - Multiband: Adjust extreme EQ settings before processing to avoid issues caused by steepness. - 1 hour - Multiband: Add 'Enable' button for steepness. 1 hour - Performance: Multiband steepness: Moving UP can be done inside current loop, no separate loop needed. Maybe down too. Currently Multiband is FAR more expensive than before. - DONE, solved in a different way. - 1 day - MAYBE: Make Steepness smarter. That would much better preserve the audio, especially at very low latencies! How: Instead of setting all the levels at AT MOST the level of neighboring bands + a bit, combine adjacent bands, determine total output level, and then fix it such that this combined output level is approached more. (So one very low, one very high --> one a bit less low, one a bit less high instead of both low). - Fix NOISE GATE behavior in VST plugin - Fix 'red output bar' issue. DONE - 1 day? - Fix Highpass filter for higher input frequencies. - Dynamic audio processing window: I've tested this at latency 512, and there it helps. But I don't have a clue what the effect is on higher latencies. They might also sound better, but they could just as well sound worse. - 2 hours. Result: 512 and 1024 got better, 2048 and 4096 got worse. So only turned on for the first two. - Dynamic audio processing window: CPU load is probably a lot higher (haven't measured it yet) due to the dynamic adjustment of the behavior. The dynamic code was originally intended to be executed only once when a latency was selected, and it's not optimized at all. Fixed, CPU load reduced, and for latency 2048 and 4096 there's no difference. - Optimize, 1 day - Reduce downsampling frequency because currently very high frequencies (21-22 kHz @ 176.4 kHz input sampling rate) in the input can cause spikes even if Hard Limit is used. DONE - 1 hour - Vibrations caused by Very deep bass distortion protection at latency 1024. Also (but far less noticeable) in higher latency modes. No issue at latency 512 because there it's turned off. - Fixed, turned it on for latency 512, and turned other bass filter OFF for 512 because it caused distortion. - Check difference in behavior between 44.1 and 48 kHz input for multiband! This could potentially result in really big differences. - No - Performance: Multiband: Remove sqrt(sqrt(cos())), pow(x, .75) etc. - too expensive, replace by lookup table. - NOT FOR NOW, would increase memory usage and hence risk more page faults, so it's not sure that this would improve the performance. - 2 hours - Performance: Move chain variables to a single Struct DONE - 1 hour - Latency: Attempt to reduce Composite Limiter sampling latency - there is no audio anywhere near the filter frequency, so a much shorter delay might still work very well - DONE, composite limiter latency is now 0.9 ms at latencies 512 and 1024 (~1.7-2 ms at higher latencies)! It can be reduced a step further (to 0.5 ms) if I allow a bit more distortion - don't know if that's useful. - Chris: "I still hear diff on Bass and kick between Beta 16 vs Beta 08 on 2048 latency. I hear more bass and kick on Beta 08. Pls check!" - No longer true, apparently fixed in BETA020... (Which does not make sense at all, but anyway) - And I need to check how much the performance is impacted by the latency improvements, in the case where no upsampling and downsampling is needed. This seems to be impacted way more than I expected (could also be Multiband steepness --> Not anymore, fixed). Ah, got part of it: 1% is steepness. Which leaves about 3% to be explained. Het is NIET de FM Hard Limit - ook al bereken ik daar nu meer van. Wellicht chain2 calls? - No, wrong again. Decreased Steepness grain match from every 4 to every 16 samples, extra CPU load is now only 2% - acceptable. - Check/fix Bass Boost ringing reported by Bojcha for higher 'upto' frequencies. REDUCED it a bit, hope that suffices... Less steep filtering (ie. bigger difference between first 2 frequencies) helps. - Attempt to set GAUSS back to 0 - gives MUCH better processing of most filters (no high frequency noise). BUT: Loudness effect in Bjork - It's Oh So Quiet - can that be resolved in another way? BUSY, 1 day? - Frequencies between 60 and 75 Hz are not handled properly yet, and can still cause vibration effects at soft high frequencies in latency 512 mode. (But FAR less than in version 6.00). Fixing this will probably increase artifacts for bass in this frequency range. - Bojcha: "There is strange "tone" at LEFT channel (tested ST dsp), caused by Bass Boost, but not always!" Questions: * Multiband: Question: Is Steepness behavior ok? * Loudness: QUESTION: The changed Punch behavior, is that good or bad? Should I attempt to let Punch behave as it did in the past as much as possible, or not? At latency 512, Steepness can now be set a bit higher than before because there are far less artifacts.

Stereo Tool https://forums.stereotool.com/

Stereo Tool 6.00 https://forums.stereotool.com/viewtopic.php?t=2811	Page 34 of 72

Author:	Chris [ Wed Jan 26, 2011 12:27 am ]
Post subject:	Re: Stereo Tool 6.00
@Hans No... Bass and Kick its ok now, after you change the latencies on Beta 20 i think... 1024 latency is very nice, i did nt check any other latency so far, but i will. About Punch, i think you can try to make it work as past... Bojcha have to check more... i think its beyong my ears..!

Author:	Bojcha [ Wed Jan 26, 2011 1:39 am ]
Post subject:	Re: Stereo Tool 6.00
I am at 512 samples mode last days. It's interesting challenge to make it work good! Around 25ms real latency with good asio soundcard, L/R output. I still need to measure FM out. With 1024 already can be made good quality preset. Punch: Like it's not working at 512 (?), but interest thing happens with punch at 512. There is already good punch without punch filter. Maybe even too much, or it's because MB behavior at 512. Currently only problem in 512 mode is MB and Bass Boost! All other filters seems good. Basss Boost: Seem good at bass (steepness 30%) but playing eg. 500Hz sine, gives ringing sounds at bass Boost (check BassBoost 'diff'). I found that "steepness" is not good over ~30%, so, acceptable artifacts. But MB not working properly then. Band are too much mixed, and no control over lowest frequencies. I knoq that Hans explained what's the problem with this very lowe latencies. Math is good. But then i ask myself, How Omnia works with 10ms latency (L/R input > MPX out), Orban, Breakaway, etc. Maybe there is different approach, or there is actually 1024 mode is used but processing latency lowered. What should be lowest latency at 1024 ? Anyway i'll measure tomorrow. Q: ST uses 1 core (?) but how many 'threads' ? Note: For 512 and 1024 mode should be 'always' locked to "never-phase-linear in bandpass. There is just no reason for using phase-linear at this low latency settings.

Author:	hvz [ Wed Jan 26, 2011 2:25 am ]
Post subject:	Re: Stereo Tool 6.00
Quote: I am at 512 samples mode last days. It's interesting challenge to make it work good! Around 25ms real latency with good asio soundcard, L/R output. I still need to measure FM out. With 1024 already can be made good quality preset. That much, 25 ms at 512 samples latency? Before you reported 22 for version 6.00... And in the current BETA's it should be a few ms lower (and again almost 1 ms in the latest beta). I expect: 12 ms + 6 ms * CPU load (based on a single core) + ASIO latency (2-3 ms?). So a latency around 17-18 ms should be possible. Have you tried increasing Stereo Tool's priority? Quote: Punch: Like it's not working at 512 (?), but interest thing happens with punch at 512. There is already good punch without punch filter. Maybe even too much, or it's because MB behavior at 512. Yes, I think I need to build something in Loudness to protect the filter against very loud lows. Actually that's already there, but it doesn't work properly (it probably causes distortion instead of reducing it) at "extremely" low latencies. Fixing this should also improve the quality of loud bass sounds. Quote: Currently only problem in 512 mode is MB and Bass Boost! All other filters seems good. Basss Boost: Seem good at bass (steepness 30%) but playing eg. 500Hz sine, gives ringing sounds at bass Boost (check BassBoost 'diff'). Will check. Quote: I found that "steepness" is not good over ~30%, so, acceptable artifacts. But MB not working properly then. Band are too much mixed, and no control over lowest frequencies. I knoq that Hans explained what's the problem with this very lowe latencies. Math is good. But then i ask myself, How Omnia works with 10ms latency (L/R input > MPX out), Orban, Breakaway, etc. Maybe there is different approach, or there is actually 1024 mode is used but processing latency lowered. If I had infinite processing power (or actually, could use 256 times as much as I do now), I could completely remove all artifacts and at the same time reduce the latency by half - so to 6 ms. Or to 12 based on 1024 processing... But more realistically in the near future: I think I need to an extra MB stage that's not phase linear. I did that some time ago and it worked very well, although I couldn't use clipping. But that can be solved by using 2 MB's: One non-phase linear before all the other processing, then one that's phase linear for the final details and clipping. But I'm not going to do this for version 6.01 anymore - maybe some 6.02 version... Quote: What should be lowest latency at 1024 ? Anyway i'll measure tomorrow. I would expect 12 ms above the 512 latency, plus 6 ms * CPU load (so at 50% CPU load that would give 15 ms of extra latency). Quote: Q: ST uses 1 core (?) but how many 'threads' ? It uses 2 threads and potentially also 2 cores. But one does much more than the other. There's not much to be gained there in latency though - I cannot do much in parallel. At latencies 2048 and below only 1 thread is used to avoid 'unpredictable' behavior which could cause hiccups. Quote: Note: For 512 and 1024 mode should be 'always' locked to "never-phase-linear in bandpass. There is just no reason for using phase-linear at this low latency settings. Maybe... But the default is 'auto select', which I think does what it should do.

Author:	hvz [ Wed Jan 26, 2011 3:15 am ]
Post subject:	Re: Stereo Tool 6.00
Addition: I also have another idea that could make latency 512 sound much more like 1024 (including better lowest band separation!) at the cost of increasing the CPU load (I expect by almost a factor 2...). But that requires a bigger rewrite of existing code - so it will take some time to do it. I just realized that maybe adding a 2nd multiband step is easier than I thought - will get back on that soon (that's an easy solution that also helps for higher latencies).

Author:	hvz [ Wed Jan 26, 2011 3:16 am ]
Post subject:	Re: Stereo Tool 6.00
BETA601-024: - Minor performance improvement; CPU load is now only 2% higher than that of version 6.00. Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 01-024.exe Stand alone version: http://www.stereotool.com/download/ster ... 01-024.exe VST version: http://www.stereotool.com/download/vst_ ... 01-024.dll VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 01-024.dll Command line version: http://www.stereotool.com/download/ster ... 01-024.exe Linux command line version: http://www.stereotool.com/download/ster ... ETA601-024 [not available] Linux GUI version: http://www.stereotool.com/download/ster ... ETA601-024 [not available] What remains for 6.01: - Multiband: Adjust band 3 (and others?) 'soft limit' behavior at very low latencies. Maybe: Also check clipping (also for band 3, might be re-enabled!). DONE! - 1 hour - Multiband: Adjust extreme EQ settings before processing to avoid issues caused by steepness. - 1 hour - Multiband: Add 'Enable' button for steepness. 1 hour - Performance: Multiband steepness: Moving UP can be done inside current loop, no separate loop needed. Maybe down too. Currently Multiband is FAR more expensive than before. - DONE, solved in a different way. - 1 day - MAYBE: Make Steepness smarter. That would much better preserve the audio, especially at very low latencies! How: Instead of setting all the levels at AT MOST the level of neighboring bands + a bit, combine adjacent bands, determine total output level, and then fix it such that this combined output level is approached more. (So one very low, one very high --> one a bit less low, one a bit less high instead of both low). - Fix NOISE GATE behavior in VST plugin - Fix 'red output bar' issue. DONE - 1 day? - Fix Highpass filter for higher input frequencies. - Dynamic audio processing window: I've tested this at latency 512, and there it helps. But I don't have a clue what the effect is on higher latencies. They might also sound better, but they could just as well sound worse. - 2 hours. Result: 512 and 1024 got better, 2048 and 4096 got worse. So only turned on for the first two. - Dynamic audio processing window: CPU load is probably a lot higher (haven't measured it yet) due to the dynamic adjustment of the behavior. The dynamic code was originally intended to be executed only once when a latency was selected, and it's not optimized at all. Fixed, CPU load reduced, and for latency 2048 and 4096 there's no difference. - Optimize, 1 day - Reduce downsampling frequency because currently very high frequencies (21-22 kHz @ 176.4 kHz input sampling rate) in the input can cause spikes even if Hard Limit is used. DONE - 1 hour - Vibrations caused by Very deep bass distortion protection at latency 1024. Also (but far less noticeable) in higher latency modes. No issue at latency 512 because there it's turned off. - Fixed, turned it on for latency 512, and turned other bass filter OFF for 512 because it caused distortion. - Check difference in behavior between 44.1 and 48 kHz input for multiband! This could potentially result in really big differences. - No - Performance: Multiband: Remove sqrt(sqrt(cos())), pow(x, .75) etc. - too expensive, replace by lookup table. - NOT FOR NOW, would increase memory usage and hence risk more page faults, so it's not sure that this would improve the performance. - 2 hours - Performance: Move chain variables to a single Struct DONE - 1 hour - Latency: Attempt to reduce Composite Limiter sampling latency - there is no audio anywhere near the filter frequency, so a much shorter delay might still work very well - DONE, composite limiter latency is now 0.9 ms at latencies 512 and 1024 (~1.7-2 ms at higher latencies)! It can be reduced a step further (to 0.5 ms) if I allow a bit more distortion - don't know if that's useful. - Chris: "I still hear diff on Bass and kick between Beta 16 vs Beta 08 on 2048 latency. I hear more bass and kick on Beta 08. Pls check!" - No longer true, apparently fixed in BETA020... (Which does not make sense at all, but anyway) - And I need to check how much the performance is impacted by the latency improvements, in the case where no upsampling and downsampling is needed. This seems to be impacted way more than I expected (could also be Multiband steepness --> Not anymore, fixed). Ah, got part of it: 1% is steepness. Which leaves about 3% to be explained. Het is NIET de FM Hard Limit - ook al bereken ik daar nu meer van. Wellicht chain2 calls? - No, wrong again. Decreased Steepness grain match from every 4 to every 16 samples, extra CPU load is now only 2% - acceptable. - Check/fix Bass Boost ringing reported by Bojcha for higher 'upto' frequencies. REDUCED it a bit, hope that suffices... Less steep filtering (ie. bigger difference between first 2 frequencies) helps. - Attempt to set GAUSS back to 0 - gives MUCH better processing of most filters (no high frequency noise). BUT: Loudness effect in Bjork - It's Oh So Quiet - can that be resolved in another way? BUSY, 1 day? - Frequencies between 60 and 75 Hz are not handled properly yet, and can still cause vibration effects at soft high frequencies in latency 512 mode. (But FAR less than in version 6.00). Fixing this will probably increase artifacts for bass in this frequency range. - Bojcha: "There is strange "tone" at LEFT channel (tested ST dsp), caused by Bass Boost, but not always!" - ASIO latency: Add configurable ASIO granularity - ASIO latency: Add active output push instead of reactive. Questions: * Multiband: Question: Is Steepness behavior ok? * Loudness: QUESTION: The changed Punch behavior, is that good or bad? Should I attempt to let Punch behave as it did in the past as much as possible, or not? At latency 512, Steepness can now be set a bit higher than before because there are far less artifacts.

Page 34 of 72	All times are UTC+02:00
Powered by phpBB® Forum Software © phpBB Limited https://www.phpbb.com/