Lame quality questions 3.98 to 3.99.5 to 3.100

[JustaNewUser]

Hi there,
I am a bit worried, ever since ripping 128 cbr mp3s ten years ago and having to do it all again, about the quality of the mp3s. Now at least I have some safety with flacs but I still need some mp3s and they should be in great quality. This is why I tend to adhere to the phiosophy of: never change a running system, and try to use long tested versions, which I change only if they have been around for at least a few weeks.
So I still use version 14.2 (works great so far) with lame 3.98 (I believe) now poweramp uses 3.99.5 and I thought about updating but I just read for a changelog in lame that version 3.100 (not yet released) fixes "PSY model tunings, to improve audio quality of problem sample 'lead voice' " while 3.99.5 fixed "Bug fix for tracker item [ 3486753 ] Artifacts at the beginning of decoded file"
Now these two sound pretty important (lead voice and artifacts) and, looking back in the changelog, it seems that more red coded items (pertaining to serious issues that affect quality) come up in a shorter time than before (not to mention the versions increase faster than before).
So does this mean the above errors are in lame 3.98 thus me having to worry about the quality and possibly having to re-encode, or have the errors just been introduced?
Should I further wait till verision 3.100 is out because 3.99.5 has a big error?
Is this all even noticable?
Is an update needed, does it have noticable benefits (both dbpoweramp and lame)?
Thanks for your input.

[Spoon]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

It does not sound like a big bug, just an improvement, 3.100 might be out in 6 months time, I would not use a pre release version of Lame if you care about quality.

[JustaNewUser]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

Thanks for the quick reply. I guess I am wrong but things like "artifacts" and "problems with lead voice" sound a bit scary even more so if you don't know if it just involves the last version or everything before this.
Oh and a bit off topic: I heard about some drives mixing up left and right channels (I hope there are very few and much older drives) if I am getting pretty high accuracies all the time, does that mean my drive is ok, or will this not in a different accuraterip result at all?
Otherwise what is the easiest way to tell?
Thanks

[Spoon]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

High accurate rip results will not be from swapped channels, I also think it is only very old (and a handful of drives) which do this.

[JustaNewUser]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

Great! Thanks, now I can sleep better Also overall good to know that accurate rip does differentiate between such differences...

1. One more important question off topic that had me confused, I recently read in another post about C2 ripping that you recommend multiple rips; I always thought (and had it set) that one rip at an accurate rip of 5 or more (with rare or new edtions possibly as low as 2 or so) would be accurate and thus would not require a second rip to assure the accuracy. Is this wrong? Should I thus always do a couple rips even with high accuracies in the first run through? Does not seem to make sense.

2. I thought the general guideline (two years ago) was to better leave C2 off since drives do not uniformly implement it well enough, is it now better to enable it? I only had about 4-9 tracks that needed re-ripping (seldomly taking a very long time) out of probably 3000 tracks and that with C2 turned off, leading to maybe half inacurate results out of the few re-rips. I hope this was not in error as most tracks had high accuracies, only thing I noticed, there were seemingly other pressings/editions with sometimes much higher accurate rip results. So did my leaving C2 off lead to rips that were marked as accurate but in reality might have contained errors, would the re-rips have been faster with C2?

Thanks

[Spoon]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

An AccurateRip of Confidence of 1 is all it needs to be sure the rip is accurate (if you have not previously ripped and submitted that result yourself)

If c2 gives issues, then switch off, on some drives it is not implemented very well. AccurateRip is independent of any ripping settings you have, if you get a match it is right.

[ffm76600]

Re: Lame quality questions 3.98 to 3.99.5 to 3.100

Final version of 3.100 was release :

http://lame.sourceforge.net/index.php