All times are UTC+01:00




Post new topic  Reply to topic  [ 166 posts ]  Go to page 1 2 3 4 517 Next
Author Message
 Post subject: Stereo Tool 9.32 BETA
PostPosted: Fri May 10, 2019 9:31 am 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 9897
Latest version - Beta 024 (June 23 2019)

Downloads
Changes/bugfixes
  • Fixed OS X crash on sound card re-open, but something is still weird (there was a code bug, but the way it got there doesn't make sense)
  • Fix and add AM presets EU, US
  • Added and updated presets
  • MicroMPX: Add API call for streaming/silence/backup.
  • MicroMPX: Add JSON API for errors.
  • Update version number
  • MicroMPX check Valgrind hangup -> ALSA doesn't seem to like Valgrind, callback stops. Ignore.
  • MicroMPX clear error messages does not work anymore???
  • Test MicroMPX and Stereo Tool with Valgrind
  • Added Kantar headroom
  • Added Kantar update sample rate (was set to 48000 fixed before)

TODO's
  • Presets Matt, Wes
  • Fading in on start after sound card issue or startup. 1 second normally, 10 seconds the very first time Stereo Tool is started.
  • RELEASE
  • MicroMPX encoder bug
  • Paul Thompson 75us preset loading bug
  • NAN on startup bug
  • Add uMPX encoder level meter -> Was already there
  • Update RDS lib
  • Add logging to see behavior of clipper
  • Test 64 bit Pi 3/4 image
  • Reported crash in MicroMPX encoder -> Caused by bad compiler optimization, fixed
  • Optimize code for small L3 cache size for ARM builds
  • Update presets
  • Clipper default settings in 9.32 sound worse than 7.70. Figure out what changed. -> Bass Clipping strictness to 1, Reduce Asymmetric Bass on
  • Reported 0/1/2 status issue in uMPX decoder
  • Force crash in uMPX encoder and make sure ST gives correct offset and does not crash. -> Message changed to clarify that this is a different address.
    -
  • Command line version could try to save 32768 as maximum int16 value instead of 32767, causing loud clicks
  • White bars fix is breaking GUI behavior on some systems when moving the window. -> Annoying but not urgent, will be fixed soon anyway.
  • Turning MicroMPX on without FM Output sound card still sends pre-emphasized audio to Normal Output regardless of setting.
  • MicroMPX should also get "Normal output" selection.
  • PhoneBooster initial version
  • uMPX can hang on close, keeps license locked. -> Freeing license on close now, so if it hangs (caused by Asio4All!), it at least can start again with license.
  • ABDP expose more strictness values (especially for mids)
  • PhoneBooster load settings without using files - actually if we put this in statics, the multiple channels thing might not be needed anymore
  • PhoneBooster 1 object for multiple channels
  • PhoneBooster show before/after displays - make bigger if possible
  • PhoneBooster add more filtering stages
  • PhoneBooster difference mode
  • PhoneBooster generates something like IMD, but why? - Gone with better filtering
  • PhoneBooster improve filtering of 3500-8000 Hz area
  • PhoneBooster cleanup of 350-3500 Hz area
  • AGC meter doesn't move in certain combinations of settings (screenshot coco)
  • Add setting for AGC side chain bass band
  • Load/Save presets via web interface. First for built-in presets.
  • Load/save presets via web interface. With up/download
  • Reduce Pi 3 CPU load further
  • PhoneBooster make multi-threaded if needed (it probably is, esp. for Pi)
  • PhoneBooster make work on low latency setting
  • PhoneBooster make Pi version
  • Add "Advanced RDS" to license page
  • Finish Kantar
  • IBC
    -
    Moved to global TODO list:
  • Streaming output can hang - calls probably need to be moved to a separate thread.
  • MicroMPX needs exception handling, just in case.
  • ABDP remove stuff that's no longer needed? BUSY for low CPU
  • Optimize ABDP
  • Input and output gain for command line version (like for sound card)
  • Simple IF
  • Dynamic adjustment of priorities
  • Have Koek update interface for MicroMPX (bug fixes)
    -
  • L-R reverb
  • Crash ProppFrexx
  • Update Omnia.SST presets
  • AES67: Add WheatNet
  • Create file when input fails
  • Change clipper defaults (Dutch Chocolate Moose has some improvements that use the new features)
  • Installer: Make installer aware of programs like MediaMonkey, AIMP, .....
  • HD delay
  • Document JSON interface
  • Documentation: "Processing" page contains all kinds of BS412 info - why?
  • Declipper L/R noise
  • Console version: Read sample rate and bit size from WAV header
  • # bands selections should allow option to insert a band instead. Or ask what to do when you change it.
  • AES67 thread affinities and priorities
  • Add resampler
  • Regarding new slider behavior:
    • Double clicking the current position of the slider will open a dialog where you can manually enter a value.

Old versions
Quote:
Beta 022 (June 20 2019)

Downloads
Changes/bugfixes
  • uMPX encoder scope/spectrum display for 1-ch input and output was wrong
  • uMPX 0x0 message removed
  • Decide what to do with de-emphasis before streaming output to reduce highs to compensate for phones that boost highs. -> Looks reasonable, let's keep it. Reduced maximum to 25 instead of 50.
Quote:
Beta 021 (June 19 2019)

Downloads
Changes/bugfixes
  • MicroMPX: License code sometimes didn't work directly after startup, fixed.
  • MicroMPX throttle status 0 is not an error, no longer showing as error.
  • WatchCat Mac build broken - why?
  • Fix bug (that apparently had no effect) in clipper code that I can't reproduce anymore.
  • uMPX build warnings
Quote:
Beta 019 (June 15 2019)

Downloads
Changes/bugfixes
  • Weird crackle in Auto EQ when playing a sweep after about 15 seconds.
  • Auto EQ could overreact after big changes, for example bass kicks in quiet parts just after a big change could be pushed down too far.
  • uMPX encoder in Mac versions of ST
  • "Block heavy processing" mode didn't work fully in composite clipper, and added some simplifications to Leif's highs gap protection code in this mode.
  • Check 4096 vs 2048 performance on Pi. -> Checked, big difference for oversampled audio only, so ideally split into 2 different latencies. But that also causes other CPU load effects. Better solution would be to try to "fix" the issue. Keep 4096 for now for testing...
  • Pi: Add "force to 2048" mode.
  • Pi: Remove "safe drawing for extreme low latency", that's not interesting here.
  • uMPX performance - what happened? (Pi is 5% heavier) -> Hopefully solved.
  • uMPX decoder does not exit on Windows (process keeps hanging) -> Is a bug in the combination of PortAudio/ASIO4All. Let's forget about this one for now...
  • Reduced CPU load for Advanced Bass Distortion Protection in "No heavy processing" mode a bit further.
  • Linux Jack issue in latest beta's (Jack didn't work) fixed
  • Message about no audio being received delayed by 4 more minutes when audio has been seen. So if there has been audio it won't complain until it's been silent for 5 minutes.
  • Put EQ or de-emphasis before streaming output to reduce highs to compensate for phones that boost highs? Added, ready for testing.
Quote:
Beta 017 (June 10 2019)

Downloads
[*]Windows 32-bit standalone
[*]Windows 64-bit standalone
[*]VST2 32-bit
[*]VST2 64-bit
[*]Winamp DSP 32-bit

[*]Linux 32 bit ALSA Stereo Tool
[*]Linux 32 bit Jack Stereo Tool
[*]Linux 64 bit ALSA Stereo Tool
[*]Linux 64 bit Jack Stereo Tool

[*]Windows MicroMPX encoder
[*]Windows MicroMPX decoder BROKEN! Don't use this version, only decodes V2.
[*]Pi 32 bit MicroMPX encoder
[*]Pi 32 bit MicroMPX decoder BROKEN! Don't use this version, only decodes V2.

Changes/bugfixes
  • uMPX encoder in Linux version of Stereo Tool
  • Update displays for "Block heavy processing" (make clear that Delossifier doesn't work for example)
  • uMPX problem with too many keyframes Mathijs
  • uMPX check sync code for possible skipping of samples, and optimize it
  • Kantar time synchronization added (not tested)
  • Updated build mechanism to create x86 Linux builds much easier and faster.
Quote:
Beta 014 (June 7 2019)

Downloads
Changes/bugfixes
  • Crash in BETA013 fixed.
  • Remove GUI 0 lines mode after testing and making sure that this is as useless as it seems.
  • Moved #lines to "Safe GUI for extreme low latency" under CPU & Latency, and changed the setting to be in Mpixel/second.
  • White bars in GUI on some systems fixed.
  • Add popup in stand alone version if there's no input audio after 1 minute.
Quote:
Beta 013 (June 6 2019)

Downloads
Changes/bugfixes
  • Report of segmentation fault on Pi -> CPU detection code did not set so parameter set failed
  • Check Matt's Immersive Bass settings -> HPF to 30
  • MicroMPX: Compare CPU load on Pi for MicroMPX with new resampler (higher temperature reported, but should have been lower!) -> No.
  • Finish Pi low CPU mode
  • Test resampler for 1 channel and > 2 channels -> Works (tested 1 and 6 channel audio)
  • MicroMPX: 80008 do something with error message (add suggestion about raising threshold?) -> No, other bug will be fixed that makes this no problem
  • Continue improving performance of resampler code - rewritten partially, test with floats instead of doubles on ARM -> Almost no effect, let's keep doubles for safety.
  • The BIMP and SCA mode were broken in the last few beta's, fixed.
  • Optimized Stereo Tool framework code for performance.
  • Add GUI settings to test writing to screen in smaller chunks to avoid cracles at extreme latencies (< 5 ms, 1 < ms I/O).
  • MB limiters don't use side chain for HPF yet
  • Check what output is sent to streaming - pre-emphasized? - No, same as Normal Out, copied setting to clarify.
  • UMPX LICENSE CODE protection Done, but needs testing
Quote:
Beta 012 (June 1 2019)

Downloads
Changes/bugfixes
  • Segmentation fault on x86 Linux (caused by compiler bug)
  • Accidentally set new "Block heavy CPU processing" default to ON instead of OFF in 932-011.
  • Check crackle in Immersive Bass - move it after Phase Delay? -> Couldn't reproduce it for some reason but probably fixed now.
Quote:
Beta 011 (May 30 2019)

Downloads
Changes/bugfixes
  • Optimize performance of resampler (combine channels)
    Note: This is a quite large code change, and the main reason why we haven't had a build in over a week. The main goal was to reduce the CPU load for MicroMPX, that has not really happened yet, but the code changes made so far will make it easier to do this.
  • Resampler output is mono
  • Peq LF Slope mode not in Legacy mode
  • Version number in DLL's now really fixed
  • Add "Safe clipping for lossy compression" switch
  • Immersive Bass display didn't work for IB around MB1/2
  • Pi low CPU mode BUSY
  • Fixed Jack name (now follows the binary name)
Quote:
Beta 009 (May 18 2019)

Downloads
Changes/bugfixes
  • Immersive Bass between MB's - Updated display code
  • Highpass filter (phase linear mode) fixed frequency calculation, added Legacy mode for compatibility
  • Highpass filter (phase linear mode) now possible at start of chain
  • Parametric EQ didn't work like it should (errors in the calculation), added Legacy mode for compatibility
  • Parametric EQ added bass filter with slope in dB/oct instead of Q which is difficult to understand.
  • MicroMPX encoder CPU load reduced, Pi version CPU load down from 170% with dropouts to 110%.
Quote:
Beta 004 (May 13 2019)

Downloads
Changes/bugfixes
  • Basstardizer: Is is broken? Compare with older version. -> No, same.
  • Fix resampler right channel issue.
  • R128 doesn't use heavy resampler code anymore (difference in output below -90 dB).
  • Immersive Bass between MB's?
  • WatchCat crashes on Mac - Crash is solved, but testing needed to make sure nothing broke.
Quote:
Beta 002 (May 10 2019)

Downloads
Changes/bugfixes
  • RMS-based limiter with ratio based on input
  • Optimize performance of resampler (combine channels) (partially done)
  • MPX Node Encoder 2 channels 16 bit input


Top
   
PostPosted: Fri May 10, 2019 11:56 am 
User avatar

Joined: Tue Mar 17, 2009 2:56 pm
Posts: 3482
Quote:
Latest version - Beta 002 (May 10 2019)
[*] RMS-based limiter with ratio based on input
What limiter?

_________________
control point


Top
   
PostPosted: Fri May 10, 2019 1:06 pm 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 9897
Quote:
Quote:
Latest version - Beta 002 (May 10 2019)
[*] RMS-based limiter with ratio based on input
What limiter?
Ah. On Multiband 1. There's a new "Follow RMS/Ratio of Compressor" mode (under Limit/Compress settings).

The advantage of this is that it acts the same way as the compressors. In peak mode, that's definitely not the case (and standard limiters use peak mode), so this new one uses the same detection and ratio as the compressors and only acts if there would be issues otherwise.


Top
   
PostPosted: Fri May 10, 2019 1:34 pm 
User avatar

Joined: Tue Mar 17, 2009 2:56 pm
Posts: 3482
Ok thats not usefull.
So i cant use RMS limiter(s) with full ratio when compessors are in peek or sym-peek mode.
Can this be atleast in singleband?

_________________
control point


Top
   
PostPosted: Fri May 10, 2019 1:59 pm 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 9897
Quote:
Ok thats not usefull.
So i cant use RMS limiter(s) with full ratio when compessors are in peek or sym-peek mode.
Can this be atleast in singleband?
We could put it there too, but why would you want it in SB? (We could probably also make a peak-based version of this).

For RMS-based MB1 it's VERY useful.


Top
   
PostPosted: Fri May 10, 2019 7:55 pm 
User avatar

Joined: Tue Mar 17, 2009 2:56 pm
Posts: 3482
I still think that limiters always need to be inf:1 and surely not sharing same ratio settings with compressors. However since limiter is by some rule peek based i would realy atleast check what would RMS limiter do, and only place where i can check it is singleband or atleast MB2 (limiter only mode).
Also, similar RMS limiter(s) behaviour in MB1 is when "flat bands top" is ON, really wrong (BUG).

_________________
control point


Top
   
PostPosted: Sat May 11, 2019 7:12 pm 

Joined: Wed Jun 03, 2015 7:03 pm
Posts: 43
I've found a bug in the latest beta. There is NO audio in the R channel with Normal Output. 9.31 is OK


Last edited by beneyto on Sun May 12, 2019 5:51 am, edited 1 time in total.

Top
   
PostPosted: Sun May 12, 2019 1:54 am 

Joined: Sat Nov 10, 2012 2:16 am
Posts: 83
Location: Australia
Quote:
I found a bug in the latest beta. There is NO audio in the R channel with Normal Output. 9.31 is OK
I am seeing the same thing also. Thought it was initially an audio driver issue but after reboots and it happening across several audio devices I was about to send a bug report as well! :D

Ross


Top
   
PostPosted: Mon May 13, 2019 10:07 am 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 9897
Hi, I don't see this here - can you give me more details about your settings? I can't think of any logical reason why one channel could suddenly be silent.


Top
   
PostPosted: Mon May 13, 2019 11:49 am 

Joined: Sat Nov 10, 2012 2:16 am
Posts: 83
Location: Australia
Quote:
Hi, I don't see this here - can you give me more details about your settings? I can't think of any logical reason why one channel could suddenly be silent.
I thought it may have been an ASIO issue but no change if I move FM Output to a non ASIO device.
However if I opt to turn off FM Output it immediately restores the right channel. Turn FM Output back on - immediate loss of audio for approx 1-2 seconds and only left channel is present.

EDIT: Found if I turn off "Synchronize to input or FM Output" on Normal Output it fixes it and I can run with FM Output ON. It was set to AUTO. Forcing it ON also is the same as AUTO (loss of right channel).

Does that make any sense?

Ross


Top
   
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 166 posts ]  Go to page 1 2 3 4 517 Next

All times are UTC+01:00


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Limited