Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Transcoding - how bad is it? (Read 2496 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Transcoding - how bad is it?

Hello Peeps,

I've searched and read old posts but am wondering is there an research on how much of a difference transcoding makes?  As much as I'd like to avoid it I have some files in AAC that I need in MP3 and some MP3s that I need in AAC.  I'm talking about going from APS to Nero AAC Normal or vice-versa.  This is for occasional recoding purposes for personal use.  (I don't file share).

So please supply some links or tips so that I can evaluate how bad transcoding is.  Yes, I will also do some testing.

Regards,
Fairy

Transcoding - how bad is it?

Reply #1 –
i don't know of any research on transcoding per se, but you can work off some assumptions.

in theory (with a perfect psy-model) a song that's on the threshold of transparency will have all it's q noise "shaving" the threshold of hearing, ie below it but only just.

if you were to decode this to PCM, and feed it back into the same encoder at the same settings, it will not have any record of what is q noise and signal, so it will add more noise to the bands it originally added noise to.  this will obviously raise the quantization noise above the threshold of hearing, and so it will break transparency.

of course, no psy-model is perfect, and so some parts will be over conservative, other parts will have too much noise.

with AAC i'd make sure i'm transcoding from a bitrate that's at least "twice" transparent... (something similar to --preset extreme or insane).

Transcoding - how bad is it?

Reply #2 –
1 )ABX the original with the lossy

2) ABX the transcoded lossy against original

Or:

ABX the transcode against the other lossy if you don't have the original.


I used this method to test MPC Q5 conversion to Lame --preset fast medium

In a few short tests I couldn't abx MPC or the transcoded file. I would say the transcodes are at least near transparent which is good for portable listening.

Another interesting thing is that a sample failed on lame 128k cbr when encoded from original achieved transparency when transcoded from MPC q5 to --preset fast medium. Obviously vbr helped things but it might show the very high quality of MPC
even at standard setting.

I don't know about aac-mp3 or vice versa though.

Transcoding - how bad is it?

Reply #3 –
transcoding between difference encode schemes probably works better.  you'll be trading off different kinds of quantization noise.  if you were to transcode --aps to --ap fast medium you'll have a different result.

Transcoding - how bad is it?

Reply #4 –
Quote
I have some files in AAC that I need in MP3 and some MP3s that I need in AAC.

If you don't have a choice, then there's no point worrying about it!

You can ABX and listen carefully if you want. But if you have to transcode, I'd suggest that's the last thing you should do! Listen casually only, and hope you don't notice any problems!

Cheers,
David.