Doom9's Forum - View Single Post

tebasuna51 · 28th January 2019, 03:51

I can't understand the problem.

- The thd is not modified by eac3to (it can't recode to thd) then the thd+ac3 is perfect for your ISO

- If the thd have at least 1 sample of 24 bits must be considered and decoded to 24 bits all samples.
It is not a problem if some samples have the less significants bits to 0.

- Don't mistake bitdepth (precission) with level volume.
A 16 sample with 16 bits to 1 have a normalized volume of -1.0 or -0dB
A 24 sample with 24 bits to 1 have a normalized volume of -1.0 or -0dB
A 24 sample with 16 bits to 1 and the last 8 to 0 (equivalent to a bitdepth of 16) have a normalized volume of -0,9999696 or -0.000264 dB
You can't see volume differences in the w64 attached image.

28th January 2019, 03:51	#10 \| Link
tebasuna51 Moderator Join Date: Feb 2005 Location: Spain Posts: 6,915	I can't understand the problem. - The thd is not modified by eac3to (it can't recode to thd) then the thd+ac3 is perfect for your ISO - If the thd have at least 1 sample of 24 bits must be considered and decoded to 24 bits all samples. It is not a problem if some samples have the less significants bits to 0. - Don't mistake bitdepth (precission) with level volume. A 16 sample with 16 bits to 1 have a normalized volume of -1.0 or -0dB A 24 sample with 24 bits to 1 have a normalized volume of -1.0 or -0dB A 24 sample with 16 bits to 1 and the last 8 to 0 (equivalent to a bitdepth of 16) have a normalized volume of -0,9999696 or -0.000264 dB You can't see volume differences in the w64 attached image. __________________ BeHappy, AviSynth audio transcoder.