Merge branch 'master' of https://github.com/slmdev/sac

README.md

@@ -1,15 +1,18 @@
 # Sac
 Sac is a state-of-the-art lossless audio compression model
 
-Lossless audio compression is a complex problem, because PCM data is highly non-stationary and uses high sample resolution (typically >=16bit). That's why classic context modelling suffers from context dilution problems. Sac employs a simple OLS-NLMS predictor per frame including bias correction. Prediction residuals are encoded using a sophisticated bitplane coder including SSE and various forms of probability estimations. Meta-parameters of the predictor are optimized via binary search (or [DDS](https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2005WR004723)) on by-frame basis. This results in a highly asymmetric codec design. We throw a lot of muscles at the problem and archive only little gains - by practically predicting noise. 
-
-I've tried adding neural networks, wavelet decomposition and predictor mixing in various ways, but so far with not much success. 
-Future improvements have to mix several predictions on a bit-level by remapping residuals to probability distributions. 
+Lossless audio compression is a complex problem, because PCM data is highly non-stationary and uses high sample resolution (typically >=16bit). That's why classic context modelling suffers from context dilution problems. Sac employs a simple OLS-NLMS predictor per frame including bias correction. Prediction residuals are encoded using a sophisticated bitplane coder including SSE and various forms of probability estimations. Meta-parameters of the predictor are optimized with [DDS](https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2005WR004723) on by-frame basis. This results in a highly asymmetric codec design. 
 
 This program wouldn't exist without the help from the following people (in no particular order):
 
-Matt Mahoney, Dmitry Shkarin, Eugene D. Shelwien, Florin Ghido
+Matt Mahoney, Dmitry Shkarin, Eugene D. Shelwien, Florin Ghido, Grzegorz Ulacha
 
+## Technical Features
+* Input: wav file with 1-16 bit sample size, mono/stereo, PCM
+* Output: sac file including all input metadata
+* Decoded wav file is bit for bit identical to input wav file
+* MD5 of raw PCM values
+
 ## Benchmark
 |Program|Parameters|Source|
 |:-|:-|:-|