This tool makes it easy to convert back and forth between FLAC and lossy.flac + lwcdf.flac files, while maintaining tags and picture art. Required tools are flac.exe, metaflac.exe and lossyWAV.exe somewhere in %PATH%. Un-comment the line: REM set "dir=%~dp0" to use tools placed in the same folder as lossyflac.cmd instead of PATH.
Some examples 1) Convert all flac and wav files in current folder to lossy.flac files (keeps flac tags and art). lossyflac *.flac *.wav -o outdir -s o @ -6
2) Convert all flacs in folder into tagged lossy.flac and correction files (lwcdf.flac) lossyflac *.flac -o lfdir -C -f
3) Reconstruct (merge) all the original tagged flac files from lossy.flac and correction flac files. lossyflac lfdir\*.lossy.flac -M -o flacdir
Changes 1.2 - fixed argument parsing of - for stream in. - option -o outname corrected. Will now work with dBpoweramp and others. - temporary files now in %TEMP% folder, not output. 1.1 - fixed handling of special characters in file names.
15 September, 2016, 04:47:14 AM by Nick.C | Views: 4154 | Comments: 16
lossyWAV 1.4.2 is released.
lossyWAV is a near lossless audio processor which dynamically reduces the bitdepth of the signal on a block-by-block basis. Bitdepth reduction adds noise to the processed output. The added noise is adaptively shaped by default and can alternatively be fixed noise shaped or white noise depending on command line parameters. When lossyWAV processed output is compressed with certain lossless codecs (FLAC, Wavpack, Tak, LPAC, MPEG-4 ALS and WMA-Lossless) the bitrate of the output file is significantly* reduced compared to the lossless original.
*: on average, depending on content.
Changes from 1.4.0:
lossyWAV v1.4.2 15/09/2016
Modification to "-l, --limit" parameter. The upper bound for the upper calculation limit is now 45.35% of the sample-rate of audio processed. Entering a value that exceeds this will no longer give an error, however the selected limit will be reduced to the upper bound for the upper calculation limit appropriate to the sample-rate of the audio being processed. The lower bound for the upper calculation limit remains at 12.5kHz.
Introduction of a "hybrid" adaptive noise shaping "-s h, --shaping hybrid" that takes into account the results of all analyses performed and not just those for c.1.5ms and c.20ms.
The order in which additional analyses are introduced using the "-a <n>, --analyses <n>" parameter is changed for the "-A, --altspread" and "-s h, --shaping hybrid" options, The order of addition is broadly the reverse of the default, i.e. 512, 256, 128, 16 samples rather than 16, 128, 256, 512 samples.
The "-f, --freqdist" parameter now takes an optional "all" parameter and will display the frequency distribution for each of the calculated analyses. Replaces "--longdist" parameter.
Modifications to "-A, --altspread [n]" parameter. Now disables "old" integer bin spreading (still used in default spreading) and permits the user to reduce the effect that adjacent bins have on the calculated minimum result for each bin. Takes an optional parameter in the range 0<=n<=1 to six decimal places, default is 1.0.
Bugfix to "-c, --checking" parameter.
Reintroduction of the "--static" parameter to modify the minimum absolute number of bits to keep during processing;
Reintroduction of the "-D, --dynamic" parameter to modify the minimum dynamic number of bits to keep during processing.
Cosmetic issue regarding display of time corrected (thanks Atak_Snajpera for bringing this to my attention);
Bug in input scaling corrected (not an issue if scaling was not used previously while processing).