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: conversion from 3.97b to 3.97 possible w/o qualityloss? (Read 7072 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Hi.

I am using LAME 3.97b2 for quite a time now, and i just thought i switch to 3.97 since it's the final version. Now, 3.97 was the former 3.97b3 so i don't think i missed much on improvement since the changelog names only one change since 3.97b2. Well, my only concern is that a reconversion reduces the actual quality of my mp3s.. Now, i took 3 example mp3s to check on the actual loss. I reconverted both mp3s with the highest average bitrate i could find in my db and the lowest (all my mp3s are encoded with the LAME 3.97b2 codec at "-v 2 --vbr-new"). The mp3s with the highest average bitrate had the biggest loss, as the first run differed between only 10, over 30 and even 60 kbp/s from the original (variation depends on the kind of music). The second run (converting the  already converted mp3s) showed at difference of 3 - 4 kbp/s. And the third run even less - 2 kbp/s of difference. Note: the codec that was used didn't make a difference either. i reconverted the original example mp3s 3 times, too and there was no difference at all. The mp3s with the lowest average bitrate didn't seem to change at all on reconversion.
My question: is there any way to convert my mp3s without any loss of kbp/s? Or is there no difference after all? Mean: since the difference gets smaller with every reconversion it seems like it's narrowing down to a "true" bitrate - the bitrate that is actually really needed to preserve the "true" quality of the mp3?
I couldn't find any thread approaching this subject in a way that would have helped me so i hope someone can help with this.

P.S. sorry about my english, it's not my native tongue.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #1 –
Quote from: akuma on
My question: is there any way to convert my mp3s without any loss of kbp/s? Or is there no difference after all? Mean: since the difference gets smaller with every reconversion it seems like it's narrowing down to a "true" bitrate - the bitrate that is actually really needed to preserve the "true" quality of the mp3?
I couldn't find any thread approaching this subject in a way that would have helped me so i hope someone can help with this.

P.S. sorry about my english, it's not my native tongue.

MP3 is a lossy encoder, so converting an MP3 to an MP3 is a lossy process.

If you are lucky, you won't be able to HEAR the quality loss in the new MP3. However it is never ideal converting from one MP3 to another MP3.

Also note. You seem to be directly associating the bitrate of a file with the sound quality of the file. Within reason this makes sense. But remember that a good MP3 encoder can produce higher quality files at lower bitrates than a poor MP3 encoder.

So it is impossible to encode an MP3 that was encoded with 3.97beta 2, to 3.97 (final) without POTENTIALLY reducing the quality. At best, you just won't hear a difference. At worst, the new file will sound worse than the 3.97b2 file.

I know I haven't completely answered all your questions, but I hope this helps.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #2 –
Out of curiosity... What are you really trying to archieve??? Is it the "b" that bothers you?

You will not gain anything (at least positive) when you encode from lossy sources, and chances are the difference would still be very small if you used the original source to encode over again.
Can't wait for a HD-AAC encoder :P

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #3 –
Don't do that.... stick with the 3.97b as there is no pratical difference... notice that b3 is what became 3.97 final... you already noticed that... think of transcoding MP3 to MP3 as sucking some orange that has been sucked by someone else... :-)

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #4 –
Quote from: ShowsOn on
...so converting an MP3 to an MP3 is a lossy process..

..If you are lucky, you won't be able to HEAR the quality loss in the new MP3..

..a good MP3 encoder can produce higher quality files at lower bitrates than a poor MP3 encoder..

So it is impossible to encode an MP3 that was encoded with 3.97beta 2, to 3.97 (final) without POTENTIALLY reducing the quality. At best, you just won't hear a difference. At worst, the new file will sound worse than the 3.97b2 file.


Hmm, well that helps alot though.. I mean, the fact that the bitrate doesn't change at lower bitrates shows that the codec obviously preserves a certain level of quality i'd say.. but i don't know if this level is to be considered good.. So what does it mean then when the bitrate of the reencoded files is lower than the one of the originals? Can this be considered better then? And what does it mean if it does not change at lower bitrates?


Quote from: odyssey on
Out of curiosity... What are you really trying to archieve??? Is it the "b" that bothers you?  ...


Well, for one thing i wanted to start using the final LAME 3.97. For another thing, i wanted all my other files to be encoded by the same codec. That said: yes, the "b" kinda bothers me..

Quote from: Bourne on
Don't do that.... stick with the 3.97b as there is no pratical difference... notice that b3 is what became 3.97 final... you already noticed that... think of transcoding MP3 to MP3 as sucking some orange that has been sucked by someone else... :-)


Yeah, i guess i'll do that.. Seems like the most rational thing to do if i want to save the quality of my files. I mean, as long as i don't know what what it means if the bitrate goes down when i reencode the files, i can't say if the quality goes down or up.. At least i wouldn't have to worry about the lower bitrate-files..

Also, I probably wouldn't be able to hear the difference between a V 2 cd-rip and a V 1/V 3 cd-rip, but doesn't the bitrate play a higher role the higher the volume is i play them at? For instance, shouldn't i be able to hear the difference at the lower frequencies? Hmm, well that question is offtopic though..

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #5 –
Quote from: akuma on
Hmm, well that helps alot though.. I mean, the fact that the bitrate doesn't change at lower bitrates shows that the codec obviously preserves a certain level of quality i'd say.. but i don't know if this level is to be considered good.. So what does it mean then when the bitrate of the reencoded files is lower than the one of the originals? Can this be considered better then? And what does it mean if it does not change at lower bitrates?


Bitrate does not change the fact that if you use an mp3 as source in order to re-encode it, you can not increase quality. You should always encode from the original wav file, not from an mp3.

Imagine xeroxing a sheet of paper: if you xerox the xeroxed-result, and then xerox again the 2nd generation result, and so on...
Paper size will be the same, but you will have reduced quality by a good margin, accumulating losses at each generation.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #6 –
Quote from: Gabriel on
Bitrate does not change the fact that if you use an mp3 as source in order to re-encode it, you can not increase quality. You should always encode from the original wav file, not from an mp3.

...Paper size will be the same, but you will have reduced quality by a good margin, accumulating losses at each generation.


OK. But what about the fact that the decreasing of the bitrate stops (or almost stops) at some point and at lower bitrates doesn't change at all? I mean, sure, that's only the bitrate again, but doesn't it have any validity at all then? Or is it just showing the bitrate the codec used to encode the file without saying anything about the quality?
Also, i won't increase the quality but i just wanted to save preferably 99% of it when reencoding..

P.S. Thanks for all the answers so far and for any in advance.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #7 –
Quote from: akuma on
OK. But what about the fact that the decreasing of the bitrate stops (or almost stops) at some point and at lower bitrates doesn't change at all? I mean, sure, that's only the bitrate again, but doesn't it have any validity at all then? Or is it just showing the bitrate the codec used to encode the file without saying anything about the quality?
Also, i won't increase the quality but i just wanted to save preferably 99% of it when reencoding..

P.S. Thanks for all the answers so far and for any in advance.


I think, every re-encoding substitutes some part of music by noise. So, every re-encoding is an additional quality loss.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #8 –
It's just not about bitrate really... LAME removes 70% to 90% of stuff human ears cannot pratically perceive. So imagine how much of it is removed from the WAV file. Bitrate is just bitrate, and there are other factors that determine the file quality other than bitrate.

@Gabriel: The photocopy machine was the perfect metaphor.

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #9 –
Quote from: akuma on
But what about the fact that the decreasing of the bitrate stops (or almost stops) at some point and at lower bitrates doesn't change at all?


MP3 encoding mostly doesn't remove parts of the music, but rather reduces precision, introducing some noise.

Back to our photocopy machine: when you photocopy a written sheet of paper, this doesn't remove sentences or words, but rather introduces some noise (easily noticeable after a few copy generations).

conversion from 3.97b to 3.97 possible w/o qualityloss?

Reply #10 –
I just wanted to thank everybody who replied to this and/or gave it a thought. All of the answeres here helped me alot, thanks!