On 2015-08-23 21:21, James Knott wrote:
On 08/23/2015 03:11 PM, Carlos E. R. wrote:
A second step would analyze this audio stream and, using its techniques for removing what the ear can not feel,
I know that's the claim, but I can hear the difference in some recordings. For example, in The Beatles "A Day in the Life". there's some low level vocals in the middle that seem to disappear in the ogg file. I can also hear artifacts that have been added when creating the file. BTW, I use K3b when making ogg or mp3 files.
Well, you have a good ear and memory :-) Yes, what you say is absolutely possible. It depends on the selected quality when the ogg or mp3 was made, though.
compresses it again to another lossy format. Two lossy conversions chained. Thus, worse overall result.
That may be the case when converting back to original and re-encoding. But what about taking the compressed data and then "repackaging" (for lack of a better term). Which does ffmpeg do? I suppose the only way to find out is to compare a converted file with one made from the source. However, beyond my ears, I have no way to do that.
Yes, I understand what you mean. That was the doubt I expressed. If it is possible to convert from one to the other by changing just the container format. I think it is not possible, ogg can not use the same decompress algorithm as mp3. However... yes, you can verify. Convert both versions to wav or some non-lossy format, then load both into a sound editor capable of displaying the waveforms with accuracy, and compare them... visually, yes. That would be very interesting to do, I think ;-) -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar)