- Adapted to run with my new module "Audio Samplerate Changer"
- Optimized I/O scheduler tunables for POCO F6
- Added a support for Tensor G4 (zumapro)
- Tuned I/O scheduler tunables especially for "resampling-for-cheapies"
- Fixed an always 192 kHz USB audio output mode for Pixel 8 and 9 series
- Changed the USB offload driver configuration of Tensor devices to be automatic max. detection
- Changed a mirroring warning for incompatible Magisk v28.0; Please use Compatible magisk-mirroring and Magisk v28.0 together
- Added better re-sampling parameters for very high performance devices
- Added disabling Dolby control support for USB DAC's on HyperOS not to reboot
- Excluded "MotorolaSettingsProvider" on Motorala devices only for avoiding their bootloop
- Changed the USB period_us from 2250 usec to 2000 usec to optimize for 48 kHz Opus tracks (recent majority)
- Adjusted jitter optimizations and others for YTM's format change from AAC (141; 44.1 kHz & 256 kbps cbr stereo) to Opus (774; 48 kHz & 256 kbps vbr stereo)
- Added new props "audio.safemedia.force=false" and "audio.safemedia.csd.force=false"
- Fixed above disabling Dolby control support (manufacturer mismatch)
- Changed head ahead buffer sizes from 16960 kB to 17000 kB to reduce I/O jitter
- Adjusted NrRequests of I/O scheduling
- Changed dirty_ratio and dirty_background_ration to be 100 to reduce jitter
- Adjusted re-sampling parameters for old devices
- Changed adjustSoC_mq for A14 and later clover not to round I/O parameters
- Adjusted cfq I/O parameters for SDM69x devices
- Changed the way how to get the actual audio policy file path in the service phase because some ROM's fail to execute "dumpsys" in the phase
- Added "compatible Magisk-mirroring" message for incompatible Magisk variants
- Tuned the USB period size for SDM845 devices (2500 usec to 2250 usec)
- Tuned the USB period size for other devices (to 2250 usec)
- Fixed logically wrong selinux prop settings (no meaning for magisk's magic mount mechanism)
- Added checking incompatible Magisk variants
- Tuned the USB period size for Tensor devices (2625 usec to 2250 usec)
- Fixed for Pixel 8's
- Changed the re-sampling parameters for Galaxy S4 to the general purpose ones (optimized for 3.5mm jack; not USB DAC's)
- Reduced I/O scheduling jitter on Tensor and SDM660 devices
- Reduced CFQ I/O scheduling jitter on Qcomm devices
- Tuned tunables of I/O scheduler
- Added warning messages for unneeded magisk modules
- Stopped Tensor device's AOC daemon for reducing significant jitter
- Optimized "extras/jitter-reducer.sh" for reducing I/O scheduling jitter on Tensor devices
- Optimized for Tensor devices by tuning GPU & I/O scheduling and replacing their stock audio policy configuration
- Hid preinstalled "Digital Wellbeing" feature for reducing significant jitter (please uninstall this manually if remaining as a usual app)
- Optimized "extras/jitter-reducer.sh" for reducing I/O scheduling jitter
- "extras/jitter-reducer.sh" now confirms and sets the cpu scaling max freq to its available max (sometimes the scaling max freq has been lowered before by a controller on some devices)
- Added support for ColorOS (experimental)
- Fixed some SELinux related bugs for Magisk v26.0's new magic mount feature
- Diabled pre-installed Moto Dolby features for reducing large jitter caused by them
- Added support for Tensor devices to bypass their spatial audio feature for reducing jitter distortion
- Added support for MTK A12 vendor primary audio HAL for reducing USB audio jitter
- Fixed a bug of DRC removing on phh GSI's (umounted in the boot process of phh GSI's)
- Nullified volume listener libraries in "soundfx" folders for disabling slight compression (maybe a peak limiter only on Qcomm devices). Tuned for "mq-deadline", "kyber" and "none" I/O scheduler (latest kernel common schedulers?)
- Added support to the Nyx kernel 4.19
- Restructured source codes to be sharable with the jitter-reducer.sh in USB_SampleRate_Changer, audio-misc-settings and drc-remover
- Adjusted a USB transfer period of the USB HAL driver for reducing jitter
- Tuned kernel tunables by assuming an audio scheduling tunable "vendor.audio.adm.buffering.ms" to be "2"
- Adjusted a USB period to reduce jitter
- Disabled the Android doze system itself for reducing jitter considerably even though battery optimizations of app's are effective
- Tuned kernel tunables for SDM 8 series, MTK Dimensity's and Galaxy S4 (LineageOS 18.1)
- Changed resampling parameters for A12 or later (not having an aliasing processing bug)
- Tuned I/O scheduler tunables
- Improved tunable values of the deadline I/O scheduler
- Fixed a bug that DRC was not detected when updating from an old module
- Tuned parameters of the I/O scheduler
- Resampling quality changes from attenuation 140dB & length 320 to 160dB & 480
- Improved AudioFlinger's resampling quality (stopband 140dB, cutoff 91%)
- Enhanced the vm jitter reducer to handle "swap_ratio" and "swap_ratio_enable" for Snapdradons
- Optimized tunables of the I/O jitter reducer
- Added LICENSE
- The GPU frequency became to be fixed really at the max frequency for Qualcomm Soc and MediaTek SoC
- I/O scheduler tunables were optimized for audio clearness
- nr_requests was optimized
- Optimized I/O jitter reducer and fixed a bug disabling effects framework under PHH GSI's
- I/O tunables improved
- Initial public release
- Initial public pre-release
- Supported audio policy configuration XML files for "disable a2dp offload", "force-disable a2dp offload", and so on
- Moved scattered functions into "functions.sh" together, and treated IO Schedulers more rigorously
- Realme support (/proc/sys/vm/direct_swappiness)
- Stopped camera servers interfering in jitters on audio outputs, and reformatted source codes
- Stopped the EAS+ scheduling feature for MediaTek CPUs
- Initial limited release