Stereo Tool

- ok.
- It simply passes some 50Hz, and it's been there long time ago.
- already told you about that.. ~month after implementation. Easy to hear if you monitor last MB band.
However, filter works better then old one, meaning, really makes holes much less audible.

_________________
control point
control point2

Hey Bojcha, I too have not heard distortion. Do you a method to replicate this easily, ie, music content, etc?

Hi Hans,

Just noticed I can't save Exceed Release Rise settings in MB2, Release part, dsp version beta 044.

Fixed:
- ASIO latency back to what it was in 8.00, so settings from 8.00 should work again - Latency issue: Is it higher than before? And what to do with people who are updating? -> Just write in memory samples_per_step samples later
- ASIO crash (but I'm not sure if that was present in the previous beta or not) fixed. The code where GuntherM reported the crash is rewritten.

Windows 32 bit:
Windows stand alone: http://www.stereotool.com/download/ster ... 01-048.exe
Winamp DSP: http://www.stereotool.com/download/dsp_ ... 01-048.exe
Windows VST 32 bit: http://www.stereotool.com/download/vst_ ... 01-048.dll

TODO:
- Measure latency WEDNESDAY
- Check Wunjo's Exceed Release issue - Tested in 048, ok
- Check if old presets behavior is unchanged
- Remove ^2 mode
- Test Jesse's preset with BETA048 and current build.
- Build Windows, Linux, Mac, Pi
- Add Wes Keene's 15 preset
- Check 2 SQRT(2) value
- Fix multiband lowpass clicking artifact
- Fix End Release in Feedback^2 mode
- Test MicroMPX with Quadrature at 320 and above/below
- Headphone Jeroen
- Edwin vdP's True Bass idea
- "This sound is processed by" in versions with different names. How to handle that?

DONE:
- Crash with ASIO fixed (reported by GuntherM) (I hope)
- Sound card not-always-exactly-the-same-latency-on-startup issue solved. This should allow using lower latency settings!
- ASIO hangs on test pc... only with new code. why? -> Not new, driver issue.
- Several Koek OHW issues
- Crash at Koek
- Wes Keene's complaint about Feedback mode appears to be fixed, added (for compatibility reasons) a new Feedback^2 mode. In Feedback^2 mode, the ratio can really be 1:INF in Feedback mode Note: This does NOT (yet) combine well with dynamic ratio's (higher Max Radio than Start Ratio).
- Crash reported in Winamp version - 036 build failed, same as 035!
- Matt/Joop barrel shape thing: Split Ratio into Start and Maximum Ratio. Note: I don't know if this is good or not yet. I added this feature after talking to several people, and on one hand it seems to make sense (in dynamic tracks, short quiet moments will get louder and short bursts will get quieter), but on the other hand it might also be bad (making the loud bursts quieter might make dynamic tracks sound softer than non-dynamic tracks, which is the opposite of what we want). Waiting for feedback, if this is bad I'll remove it again. Note: With setting lower than the Start Ratio (for example the default value of 1.00), the Maximum Ratio is ignored.
- FB post
- Andere analyser installeren
- Sound cards! Arrived! WEDNESDAY TO RAI 2x
- Check Axia mixer OK! GET WEDNESDAY
- Fancy name tags ORDERED
- Fast pc! ORDERED Mathijs
- Connector tulp male naar XLR male
- Make full IBC setup
- SST 2 knoppen PNR
- "Force higher level action" must also be performed for server events!
- PNR RESET button does not get enabled/disabled
- MicroMPX bug: -a "Card name" does not work!
- Run Valgrind for MicroMPXMathijs
- Rename PNR Noise & Hum to Dehummer?
- Crash when zooming out in waveform fixed. (Has been present for a long time, but was almost impossible to reproduce).
- Declipper waveform was wrong for Quality > 100%.
- Crash Leif?? (Maybe bad server data?) - Not seen anymore, several bugs fixed - hopefully gone
- MicroMPX bug: -a "Card name" does not work!
- Run Valgrind for MicroMPXMathijs
- Myeasylog.log
- Rename PNR Noise & Hum to Dehummer?
- Crash with test tones -> Old beta (015), ignore.
- Crash when zooming out in waveform fixed. (Has been present for a long time, but was almost impossible to reproduce).
- Declipper waveform was wrong for Quality > 100%.
- Cleanup presets
- "Force higher level action" must also be performed for server events!
- PNR RESET button does not get enabled/disabled
- SST presets
- "Load processing" skips Clipping - No was old file.
- Help page with power user markings. Some settings (sound cards) should not be marked.
- STAND ALONE VERSIE in SST mergen
- Als saved en Load onder All Audio setting dan is Over Sampled Limiting and Clipping uit geschakeld -> Seems to work as designed (which is different from before though). -> Changed back for Stereo Tool
- Sync: Handle big difference between clocks. More than .01% is probably already too much. Then ignore. Add a parameter, set to 0.1%.
- Crash reported by Wouter Tietema, probably caused by low latency monitoring output. - probably fixed.
- Sync: Was adjusting new constant offset from Normal to FM output in wrong direction.
- Edit: I see occasional jumps in the output VU meter. They are *not* present in the scope or in the sound card output. Will check.
- Stand alone HTML5 GUI FRIDAY
- Linux Fedora issue
- Make MAC version HELP work. (tonight) - to be tested
- Remove LQLL HQ mode (non-SST) (tonight)
- Check VMware flag turning ON for unknown reason (tonight) -> Not changed recently, and seems not to cause problems, so ignoring for now.
- Fixed synchronization problem between meters, scropes and audio
- INDEX.HTML automatisch resolven
- Add no cache to objects.js!!!
- Declipper broken according to DJ_DOGGY, so far unable to reproduce!
- Change version number
- Crash in BETA029 fixed.
- Linux Rivendell version -> Returns silence in FM output, why??? Add logging! --> _ST_AUDIO define! But why does it work in Windows?
- Kill License SST
- Remove warnings
- Check red output bars for 'Rules are there to be broken' - overshoots and clicks present.
- Generic plug-in (not posted) FM output to host doesn't work.
- Make new separate Presets class ->
- Check and solve Linux Valgrind errors for Declipper
- Linux high CPU load
- Solved Low Latency mode crash
- Previous version could crash on startup. Should be solved now.
- Thimeo.js add DISABLED displays
- Test if clock moves in right direction (that could be the gaps that are reported) --> Clock jitter calculation effect removed for Stereo Tool (not for MicroMPX). I hope this fixes the issue. Waiting for feedback.
- White list IP's?
- Absolute Highs display broken. Only in HQ version. -> Weird only on Koek pc. Check tomorrow. Ah, only weird if there's real audio, not on background noise there. Maybe stereo effect? Still unable to reproduce on my pc... Copy settings. -> Was clipping in the Winamp output which was fed into VAC. Enabling the Declipper made it a lot better!
- Add agc drive, remove Pre Amp (not remove, default is already 1, so never mind).
- Show link to web interface, hide everything else! (Password)
--> Clock jitter calculation effect removed for Stereo Tool (not for MicroMPX). I hope this fixes the 'gaps' issue. Waiting for feedback.
- Crash reported by multiple people fixed.
- Claimbigf failed once ("Claimbigf failed: In use!") -> Added extra checks, should warn if it happens again. I'm guessing it was triggered by some other bug. --> Was caused by crash
- Split presets in repair/process
- VLC Input broken (resampler is BAD, should probably be disabled in this case?) - Also sound cards should collect MUCH more data before starting to resample.
- Turning pre-emphasis on/off give a big glitch! - Was caused by Phase Delay time shift triggered by having 'Ignore high frequencies' on.
- CPU load: Test if it's indeed lower in 021. Quick test: 45.5% to 44.8%. So, yes.
- Check that FM output can still be L/R stereo if Stereo/RDS encoders are disabled -> No. Needs to be fixed. -> Does what it should do now.
- Fixed some potential crashes - not sure if the one that was reported by 2 people is included
- Make better resampler! Test correct direction! Factor = input/output
- Quality 100-110 is not great, rest is.
- SSB DSB mode in enum, check if it works
- Enum 50/75 add OFF, encode Mu as non-HTML mu
- Add Declipper main view
- Bad pilot/RDS at certain Quality settings, only if Threads==1. Seems ok at high Quality settings... Different in HQ and non-HQ builds. Ah no, sometimes also at Threads > 1. Display only though. Ah, happens when Normal output buffer <= FM buffer size and FM output is disabled but I am displaying FM output. Must be looking to close to the point where I'm writing in OldSamples. (Correction for Normal Output is applied to FM output, probably?). O Caused by synchronization issues between threads... And at different numbers of threads everything is broken (note: Only scopes... but all of them!). I think we need a position in the display per thread index to solve this. Then the existing code might work again without changes. (Except that it needs to be triggered from the Framework thread).
- Weird time shift in bottom waveform (or in top waveform?) they don't stay synced at 10 second view size. -> Fixed with bad pilot fix
- Problem with threads is fixed, but still needs to be synchronized with the rest of the GUI and sound card (too much delay now on scopes).
- Phase Rotation before/after doesn't update
- FM Analyser was broken -> mostly fixed, some glitch still present after sound card restart on 2nd channel. -> Ignoring for now...
- Add Delossifier difference, also for Stereo, True Bass & Absolute Highs & Bass Boost, add Declipper, Equalizer, Bandpass, Clipper -> No, only Bandpass and maybe Clipper.
- Smooth preset switching: No more hiccups (if the same filters are enabled) and also faster.
- Quality between 100 and 150% improved (around 101-110 not great yet)
- Connection lost issue web gui
- Add window to AGC (wednesday)
- Make 'HEAR'/'Till here' work (wednesday)
- Test tones are broken (wednesday)
- Make output waveform visible again for non-FM output. -> Appears to be ok now.
- OS_COPYTHREADPRIOTO(td[ch].process_thread) (wednesday) - Was already ok
- Display CPU usage for LQLL threads (tuesday)
- Set affinity for LQLL (tuesday)
- Display needs to be updated for all processing threads (tuesday)
- Display updates must happen in correct order (weird noise jumps otherwise) (tuesday) -> No, was related to updates for all threads
- Don't calculate FM signal if FM is disabled (causes extra display in output too) (tuesday)
- Large pilot phase difference found (tuesday) Fixed with workaround. Reason still unknown.
- AGC gate in bars like compressors - now also in the normal versions
- Displaying CPU usage for all (primary) cores if you select more than 1 thread.
- Reduced memory usage to below what version 8.00 used!
- Gain works again
- AGC gate in bars like compressors
- Low Quality Low Latency monitoring (stand alone version) is working again
- Restart server if all clients are full for a while?
- Delossifier web display added
- Type IP in web site
- Quality below 100% in HQ build error

DONE IBC:
- Email NAB visitors
- Images for booth presentations Mathijs
- Images for ST presentation
- USB sticks!!! ORDERED ARRIVED
- Business cards Matt ORDERED ARRIVED
- Call Karlex CALLED, WILL BE SHIPPED.

Later:
- Update website LATER
- CHECK CRASH with VLC - VLC related?? Not seen anymore, later.
- Perform latency measurements
- Web interface export preset
- Generate UIDs and use link
- Password for login?
- Attempt Wes Keene's Feedback mode (see Facebook chat)
- Auto-determine best latency/affinity settings
- New idea: FM stereo trick might reduce MP3 artifacts as well! So add it to streaming output.
- Run as a service
- TODO: Add exception if ignoring too high VST parameter numbers (Audacity bug! -> Was fixed in later Audacity version, never mind).
- Websockets? Later
- However, MediaMonkey was quite insistent I upgrade to the later version 4.1.13.1801, so I did, and now what is happening is the Declipper screen is whited out and can't be configured, though it still works with the last set settings.
- Also when I close MediaMonkey, the Declipper screen hangs (everything else closes), and when I do a forced close, Windows says it's MediaMonkey hanging.
- Still a hack (enabled with a special setting) is needed for some people who want RDS with stereo output... It doesn't make any sense but on some transmitters that apparently works... However I only know of one person who uses this. Not urgent now, this person can use an older version for now.
- Remove inputOldSamplesFixedCopy, should not be needed if I split WAVE from VU meter calculations.
Show MicroMPX block size and quality

Try to listen output while monitoring last MB band.

_________________
control point
control point2

Bojcha - Are you suggesting that all FM users would be better off (cleaner, less distortion) disabling this feature right now?

I released a new Minnesota for beta 48.
viewtopic.php?f=10&t=6048

This (and beta 40) might be the only version with a "Digital" mode multiband that allows operation nearly this close to the sound I've been wanting to get from this preset for over 4 years now. It only happened because of an accident, so hopefully this version works for you.

A few people told me that "Digital" mode is considered depreciated. Well, listen to this happy accident and see if you can match it with "Analog" mode. Sure Take it as a challenge.

Finally you can hear what I've been wanting to do with Minnesota during the 4 years of its development!

The input is calibrated to -23 LKFS / 0 LU. If you're using a different loudness at the input, please adjust the "Pre amplifier" to match the difference in your loudness normalization. You should be seeing the main AGC's band 2-3 doing 6 to 12 dB gain reduction most of the time, and seeing a decent amount of partial gating.

Sounding very nice. I TOTALLY agree with your comment about the Digital compressor, and add its not depreciated. Its just it has a 'similar' sound to the commercial vendors that ST has to compete against.

I think the deprecated "jab" was aimed at me, in part. To be clear, I love Digital Mode. Strictly speaking, it is deprecated though. If you go to Analog mode, you won't see the option present anymore, unless you check the option to show deprecated settings.

Just wanted to make sure people know I still like Digital Mode. I always have.

In part.

I mainly get the feeling that it's not going to ever be allowed (optionally) to operate linear like it almost does in Beta 48 (I've been suggesting the idea to Hans for years now). It still releases slower when it gets close to 0dB, but that's not a bad effect since it makes it easier to have the transition from having GR/range to none/flat easier to make transparent/inaudible. It would be nice to be able to adjust that effect too, but ehh I'll take all the happy accidents I can get. But other than that, it's responding linear in this beta.

Ahh ok. So it's considered depreciated for sure then.

I would like to see someone figure out how to make the new compressors have the kind of acceleration & deceleration that I'm getting from Digital mode's attack/release flatness (gamma), and release inertia. When I tried with Analog the last time, there was some kind of limiting happening to the way that it changes speed, even when changing certain settings into the millions by preset editing.