And now something for techies only. I made a -60dB sine file, and went through the same process as above. The resulting files are analyzed by spectrum analyzer and presented as follows:
green: (*1) original 16-bit wav file synthesized by soundforge violet: (*4) wav file obtained by compressing and decompressing using lame red: (*3) wav file digitally captured from squeezebox, playing mp3 The original file (green) is mathematically perfect, all "artefacts" are 16-bit rounding. The mp3 file (violet) is surprisingly good, looking almost the same as the original. That is probably because a high-bitrate mp3 is more than enough for a single sine signal. The file captured from squeeze (red) is some 20dB worse, clearly showing the artefacts. In the real situation, with a complex music file, the induced noise is probably much higher, although not so easy to demonstrate. Hope this may help to easier find and analyze the bug. +-------------------------------------------------------------------+ |Filename: 1kHz -60dB spectrum.png | |Download: http://forums.slimdevices.com/attachment.php?attachmentid=3758| +-------------------------------------------------------------------+ -- damir ------------------------------------------------------------------------ damir's Profile: http://forums.slimdevices.com/member.php?userid=14119 View this thread: http://forums.slimdevices.com/showthread.php?t=40409 _______________________________________________ discuss mailing list discuss@lists.slimdevices.com http://lists.slimdevices.com/lists/listinfo/discuss
