A Russian thread has about 74 pages on this player. On this page http://forum.doctorhead.ru/index.php?showtopic=24713&st=1650 someone is trying to explain this player with pictures. Run it through translator, see if it makes sense.
It is not simple to understand . The main difference is that the author, Hiroyuki , considering the source format 441000/16 not as a model , but as a lossy format . Accordingly, all these Black, Emerald , armor and other - these are very sophisticated algorithms restore lost when digitizing information. It uses spline interpolation , and as a model of multiple - model schemes recording and CD production . Accordingly, a huge load on the processor.
On a button - start when the first dialog box buttons and hold Shift + Z mode is activated Z-tune ( only on 64 bit systems and 8 gigs of memory ) , which optimizes the algorithm of memory , then the choice of ASIO or WASAPI output. First preferred.
After self window. Starting with version 2.92 of the dialogue removed buttons that are not harnessing the specific modes. Normal - mode output similar to Foobar and differs mainly in algorithms only work with memory. Modes with sound BufHead - a bit about them .
top row - selection X1 X2 X4 X8 - degree sampling transition , Basik, Emerald, ... Brownie- level and signal processing . LPF-free and under it - filtering the output signal different algorithms Free- unfiltered - important - if filtering is enabled it is necessary to ensure that the level would not exceed the maximum allowable .
Self sound optimaiser- create their own algorithms
1 - mode volume control output -AUTO player will reduce the volume if it is outside the permitted , pressing aotklyuchaet automatic control , then a box under a continuation of the arrow indicates the maximum output level for a particular algorithm - for example Browine 5x2 it will be 13 , he should be put on the volume
Next 2 - allows you to select the size of the cluster and the precision of the calculation for the spline interpolation , the higher accuracy and the more stages , the more accurate the result and the more demand resources. ( Workstation 6 on two nuclear Cheon loaded at 100% and believes the sound is 180-300 seconds at the maximum values â€‹â€‹of the image )
3 - selection phase of the output signal
4 - Select the algorithm clear memory 9 different types of memory are different algorithms provide the highest quality sound , empirically )
5 - choice player in the harnessing of cores and threads
6 - Select the processed sound Bagh- HS - output sound from Bug LV5- maximum BUG LV1- minimal -
mode Self sound optimizer has the ability to export the processed stream player in crude RAW file for export WAV data from it and the possibility to listen to any player . There is no support CUE!!!