The title says it all. I have tons of FLACs and every one of them has a copy in MP3 encoded in LAME 3.98.4 and some are in 3.99.3. 3.99.5 is the new latest version now, should this mean I have to re-encode those FLACs to 3.99.5? I would gladly want to hear opinions of others.
No.
May I ask WHY?
May I ask WHY?
Nobody can answer the question. Only You. Because You will listen your music. Not us.
It all depends of your ears, music, hardware, LAME's settings.
The goal of lossy compression like mp3 is to achieve perceptual transparency at bitrates a fraction of the lossless original. Barring the exceptional problem sample, lame, when used properly, has been producing perceptually transparent encodes for quite some time though (going back who-knows-how-many versions).
So if your current mp3s are satisfactorily transparent to you then there's no need to re-encode.
The title says it all. I have tons of FLACs and every one of them has a copy in MP3 encoded in LAME 3.98.4 and some are in 3.99.3. 3.99.5 is the new latest version now, should this mean I have to re-encode those FLACs to 3.99.5? I would gladly want to hear opinions of others.
Joining in the choir here ...
You haven't even specified what settings you used. There were changes from 3.98.4 to 3.99 intended to improve upon quality. But that is quality for a given bitrate. (Also, some changes were done to the presets, I think, effectively changing the bitrate on a few of those.) So in principle, you could get fewer and less annoying artifacts for given bitrate. Or -- probably a more fruitful point of view -- the same subjective quality at a slightly smaller file.
If there are audible artifacts, then you should change the
setting. By all means upgrade the version, but that's probably a smaller difference. (And if you do, check that the presets do mean the same, I think that the 'tuning on VBR scale and resulting bitrates' change to 3.99 means they don't.)
From 3.99.3? Only if you experience the issues addressed in the upgrade: http://lame.cvs.sourceforge.net/viewvc/lam...ml/history.html (http://lame.cvs.sourceforge.net/viewvc/lame/lame/doc/html/history.html)
From 3.99.3? Only if you experience the issues addressed in the upgrade: http://lame.cvs.sourceforge.net/viewvc/lam...ml/history.html (http://lame.cvs.sourceforge.net/viewvc/lame/lame/doc/html/history.html)
That link shows
LAME 3.100 not yet released and . . .
LAME 3.10 May 30 1999 ?(http://img803.imageshack.us/img803/9949/65673640.jpg) (http://imageshack.us/photo/my-images/803/65673640.jpg/)
Furthermore, the use of a higher bitrate setting may allow you to not have to re-encode with every new release. For example, I have a few 256kbps-320kbps FhG mp3 files from back in the day that sill produce transparent results for me. They are 8 years old now but I have absolutely no reason to re-encode the source material again (I can't find the CDs anyway) because they are still transparent. Don't take that to mean that encoding at such high settings will automatically allow you to just scuff things off or that the high bitrate settings produce better results. I also have a bunch of older (5 years) 160kbps VBR AAC files encoded with iTunes that I am not going to bother replacing despite Apple releasing a few updates to their AAC encoder between then and now. I will encode new material with the updated encoder but I don't see a point in going back when what I have is already transparent to me.
(Also, some changes were done to the presets, I think, effectively changing the bitrate on a few of those.)
IIRC, the changes in bit rate resulted in bigger files with the newest version; at least at -V0.
FWIW, I'm still encoding with 3.98.4. It works great for my purposes @-V3 so I haven't bothered to upgrade.
If your goal is to reach the same sound quality with smaller files, a way to have a practical answer to your question is to take a relevant part of your collection, say a tenth, and re-encode it with the same quality target settings as the first time. If the space gain satisfy you then go on, else don't...
And it would be very kind of you to let us know the results.
From 3.99.3? Only if you experience the issues addressed in the upgrade: http://lame.cvs.sourceforge.net/viewvc/lam...ml/history.html (http://lame.cvs.sourceforge.net/viewvc/lame/lame/doc/html/history.html)
That link shows LAME 3.100 not yet released and . . . LAME 3.10 May 30 1999 ?
…and…?
Both are true.
…and…?
And, you could explain that version numbers aren't decimal numbers. 3.9 < 3.10 < 3.100 because it's really 3 & 9, 3 & 10 and 3 & 100. The mark has no other meaning than to separate two numbers.
Oh, great point! My bad.
Yeah, they’re just running out of space in their original numbering scheme.
There could be few reasons why it would, literally, be "necessary". Perhaps if it was known that the version that had been used to encode your existing files introduced bugs or known sound quality problems, but that's not the case.
Then there's the question of whether or not it's desirable. That's debatable, but there arguably hasn't been much progress in sound quality from 3.98 to 3.99. A few members have even suggested the opposite.
For me, the main reasons for switching to 3.99 have been in usability. You can now embed artwork files larger than 128kB. You can now add TXXX Id3v2 frames when encoding. Since anything new that I transcode from FLAC will be done using 3.99.5, I chose to re-encode the entire existing FLAC library. On my server with about 35k tracks, it takes about 3 days running in the background, which isn't really any more work for me than if it took 3 minutes.
Personally I find such usage of electricity wasteful.
Regarding artwork, am I not able to embed a >128kb to my 3.98.4 Lame-encoded mp3 with a program like mp3tag?
Personally I find such usage of electricity wasteful. ...
"No trees were killed in the creation of this post. However, a large number of electrons were seriously inconvenienced."
Regarding artwork, am I not able to embed a >128kb to my 3.98.4 Lame-encoded mp3 with a program like mp3tag?
I'm sure you can. When transcoding a library from FLAC to MP3, though, the process (for me) needs to be totally hands-off. With LAME 3.98.4 I think my script was using metamp3 or a similar utility in addition to LAME. Version 3.99 makes things both a little easier and a little faster, since the file doesn't have to be rewritten to add the artwork tag.
An additional usability issue that remains for me is that I'd like the ability to write ReplayGain frames directly to the ID3v2 tag instead of (or in addition to) the LAME header. It's still not clear to me why anyone would bother adding the ability to compute RG values and then store them where few applications will see them. I don't need album RG, so LAME's RG calculations work well for my purposes.
And one more: The ability to copy a timestamp from the input file. When keeping a mirror library, it helps a lot if the resulting MP3 file has the same timestamp as the original. This would make it easier than having to call in a 'touch' utility to alter the MP3 file's mod time after the encoding process.
I have several old mp3 with lame3.90 -lame 3.97 cbr 192kbps etc. I 'd love to upgrade all of them to the latest version 3.99.5 ?
Which is the software that i should use? I don't want to lose any audio information. And only the version should be updated and nothing else should change.
Please tell me some software that does the above job accurately..thanks in advance
@damoresh, see "Lossy-to-lossy transcoding" in this wiki entry (http://wiki.hydrogenaudio.org/index.php?title=Transcoding). What you propose will most likely not be an "upgrade".