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: --r3mix 3.95 vs. 3.93, bigger files? (Read 12063 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

--r3mix 3.95 vs. 3.93, bigger files?

Hello.

I couldn't help noticing LAME 3.95.1 --r3mix creates much bigger files compared with 3.93.1.  Any idea why is that?  (Too many bits for me...)

Zn.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #1 –
Have a look at the HydrogenAudio recommended LAME switches here... --r3mix hasn't been in development for some time now (although this doesn't help explain how it changed between those LAME versions...) and also has been proven to be less useful and efficient than the more commonly used switches here.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #2 –
Isn't r3mix dead with 3.95?  I thought it was automatically mapped over to one of the presets now instead.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #3 –
Yes, it is. It now equals -V3 --vbr-mtrh.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #4 –
So if you want to use less bits and retain similar encoding speed use the next Vx number, ie. -V4 --vbr-mtrh (which is actually equivalent to --preset fast medium).

--r3mix 3.95 vs. 3.93, bigger files?

Reply #5 –
And Yet everyone still uses 3.90.3 Who wants to be scared they will have some unforseen problem with their files in the future. Heres to testing :up:
r3mix zealot.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #6 –
--r3mix is now "-V3 --vbr-mtrh"?!

It used to be -v1 and produce smaller files, and now -v3, supposedly a lesser quality, produces larger files?  What's up with the new -v? mapping?

I thought some time ago people were happy with --r3mix, considering it to be a very good choice.  Whatever happened to that?  I'm just not happy with the bitrate going up instead of down.  It beats the whole idea of compression, and here, look at some quick results I got from encoding a short classic guitar + piano part:

  3.93.1 --r3mix:          156kbit
  3.95.1 --r3mix:          174kbit
  3.93.1 --preset standard: 195kbit
  3.95.1 --preset standard: 208kbit

For a quiet piece, I'd be happier with 160kbit, and not 210kbit.

So what's the best equivalent to the old --r3mix, getting reasonable bitrates and taking advantages of whatever improvements added to LAME in the last few versions? Speed doesn't matter, just final bitrate and quality.

Ah... can't they just keep it simple?

--r3mix 3.95 vs. 3.93, bigger files?

Reply #7 –
Quote
--r3mix is now "-V3 --vbr-mtrh"?!

[...]

So what's the best equivalent to the old --r3mix, getting reasonable bitrates and taking advantages of whatever improvements added to LAME in the last few versions? Speed doesn't matter, just final bitrate and quality.

Ah... can't they just keep it simple?

Did you actually read the previous posts?

Now it is indeed really simple. Just choose a -Vn where you get the best ratio between filesize and quality. Is V3 too big? Go for V4... What's so difficult about that?

--r3mix 3.95 vs. 3.93, bigger files?

Reply #8 –
Quote
So what's the best equivalent to the old --r3mix, getting reasonable bitrates and taking advantages of whatever improvements added to LAME in the last few versions? Speed doesn't matter, just final bitrate and quality.


With an emphasis on quality, the answer you will get to this question is a variation of the "--alt-preset standard" command line. The reason for it goes to this:

Quote
Ah... can't they just keep it simple?


That's really the problem. The --r3mix command was deceptively simple, too simple. It was the first tentative step toward producing a (argued) transparent-all-the-time LAME mp3 encoder setting. However, even in this goal it was compromised. The originator of the --r3mix option did not take into account some of the more important vagaries of LAME, and that others could hear better than himself. Public testing of --r3mix was limited at best, and tended to be eschewed for generalizations of typical human hearing.

Flash forward an eon or so. Several capable individuals, including some of those directly involved with coding LAME, put forth a massive amount of effort to create a transparent-all-the-time setting for LAME (within the limits of LAME and the MP3 spec to do this). They were focused on dealing with LAME 'gotchas' that the --r3mix creator was not aware of or did not account for, and utilized verifiable public testing to determine on a larger scale the threshold of 'transparent' to the general audience.

In the end, this meant that the bitrate tended to be a bit higher, but the results are generally accepted to be more true to the source material. As stated above, if you really need lower bitrates and are willing to trade off potential quality for size, you might want to try the "--preset medium" (intended for portable players) or "--preset (insert target average bitrate)" with compatible versions of the LAME encoder. Do note that some of the LAME 'gotchas' dealt with by the "--alt-preset standard" command can result in easily detectable and possibly annoying artifacts in lesser settings. You are likey to get them with --r3mix as well.

Yes, I have been lurking for that long.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #9 –
right, i don't know as much as these guys but i'm gonna just try to be helpful. try:

--alt-preset standard -Y
or
--alt-preset medium

one of those is probably what you want.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #10 –
It's time that the LAME developers either disabled the --r3mix switch or make it an alias for --aps. There are thousands of people out there who use this switch.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #11 –
Quote
It's time that the LAME developers either disabled the --r3mix switch or make it an alias for --aps. There are thousands of people out there who use this switch.

agree. much harm hast -r3mix has allready done :|

--r3mix 3.95 vs. 3.93, bigger files?

Reply #12 –
Quote
It's time that the LAME developers either disabled the --r3mix switch or make it an alias for --aps. There are thousands of people out there who use this switch.

I agree. I think that it should be mapped to --preset medium, instead, though; since the bitrates tend to be equivalent and average-joes won't probably notice the change. Map it to APS, however, and you'll end up with a swarm of them complaining about why the update produces 25% bigger files with the "same" quality.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #13 –
Hey now, I got started with r3mix, and it was the only audiophile command-line switch for mp3 that I knew about at the time... it was a good starting point, but I agree, needs to be mapped to --preset medium or something.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #14 –
Quote
Hey now, I got started with r3mix, and it was the only audiophile command-line switch for mp3 that I knew about at the time... it was a good starting point, but I agree, needs to be mapped to --preset medium or something.

Dude WHY are you using -Z in your sig's lame command line? That is bad! Don't ever use that on 3.90.3 or 3.95.x! And for that matter, don't mess with the mid/side stereo switching in MPC either, it's tuned properly by smarter people than you!

--r3mix 3.95 vs. 3.93, bigger files?

Reply #15 –
Quote
Quote
Hey now, I got started with r3mix, and it was the only audiophile command-line switch for mp3 that I knew about at the time... it was a good starting point, but I agree, needs to be mapped to --preset medium or something.

Dude WHY are you using -Z in your sig's lame command line? That is bad! Don't ever use that on 3.90.3 or 3.95.x! And for that matter, don't mess with the mid/side stereo switching in MPC either, it's tuned properly by smarter people than you!

Please stop zealing at others...  he has to listen to his files, not you. Or did he force you to pay for and download his music?

BTW I doubt many people here slinging at --r3mix could ABX the difference to Darin's presets on their music.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #16 –
Quote
And for that matter, don't mess with the mid/side stereo switching in MPC either, it's tuned properly by smarter people than you!

In fact, --ms 15 is a recommended switch for those who encounter soundstage problems with the default setting in MPC's presets.
Of course, I would strongly recommend that he ABX the soundstage deficiency first, if he has not...

--r3mix 3.95 vs. 3.93, bigger files?

Reply #17 –
Average bitrate of a preset should not be measured using a single encode, otherwise it is not average.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #18 –
Quote
So what's the best equivalent to the old --r3mix, getting reasonable bitrates and taking advantages of whatever improvements added to LAME in the last few versions? Speed doesn't matter, just final bitrate and quality.

Ah... can't they just keep it simple?

Try this post

--r3mix 3.95 vs. 3.93, bigger files?

Reply #19 –
Quote
Quote
Quote
Hey now, I got started with r3mix, and it was the only audiophile command-line switch for mp3 that I knew about at the time... it was a good starting point, but I agree, needs to be mapped to --preset medium or something.

Dude WHY are you using -Z in your sig's lame command line? That is bad! Don't ever use that on 3.90.3 or 3.95.x! And for that matter, don't mess with the mid/side stereo switching in MPC either, it's tuned properly by smarter people than you!

Please stop zealing at others...  he has to listen to his files, not you. Or did he force you to pay for and download his music?

BTW I doubt many people here slinging at --r3mix could ABX the difference to Darin's presets on their music.

I'm not sure how this is zealotry... The -Z switch was added to 3.90.3 to change the noiseshaping model, but 3.95.x does not need it. Just giving him a heads up, and frankly the less -Z shows up on these boards the better, as confused people are still adding it to everything.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #20 –
*sigh*... so sad.

At the beginning I used FhG 128kbit, but after seeing what people had to say (more than actually hearing - possibly because I'm yet to have good quality sound gear and listening conditions [aka quiet computer]) I switched the LAME --r3mix.  I lived happily with the increase is file size, but now, is it all about to shatter?! ;)

Perhaps it is time for newer formats (OGG? AAC?) to take over.

But returning to that "table" of mine (see earlier), does it really take 210kbit to encode what I thought was a simple sound clip (classic guitar and piano)? This is a bit of surprise.

// ZiNC

--r3mix 3.95 vs. 3.93, bigger files?

Reply #21 –
duh kode54 beat me by several days..  --r3mix is indeed -V3 -vbr_new, nothing else anymore.

--r3mix 3.95 vs. 3.93, bigger files?

Reply #22 –
In the grand scheme of things, ~200kpbs is no more arbitrary than 128kbps. It just happens that better testing and focused tuning has resulted in a generally transparent setting (within spec and encoder limitations) at a cost of increased size. Transparency at any level was never a goal of the 128kbps settings, it merely represented a fair tradeoff of quality for size at the time.

With that said, there is nothing that requires you to use these settings. These merely represent the current 'sweet spot' of smallest size for the highest quality available, something you may not need. It is up to you to decide what quality/size ratio suits you best.

Ask yourself these questions:
Do I have enough space? Can I add more conveniently if not?
Do I need hardware compatibility, either in bitrate (CBR vs. VBR) or overall size?
If I do find I can hear the difference and find it annoying, is it worth the time to re-rip and re-encode?

Perhaps before the others, you might consider this:
Do I really need MP3 at all, or could I do just as well with a different and possibly more space-efficient format?

--r3mix 3.95 vs. 3.93, bigger files?

Reply #23 –
Quote
I couldn't help noticing LAME 3.95.1 --r3mix creates much bigger files compared with 3.93.1.  Any idea why is that?  (Too many bits for me...)

Nobody's forcing you to switch from 3.93.1 to 3.95.1. So, what's the problem? 

--r3mix 3.95 vs. 3.93, bigger files?

Reply #24 –
Ahh.. The vicious --r3mix switch has hit again.. I used to use this switch, but thanks to the admins here, I found a much better setting..

--preset extreme -b32 -B320

It works great, and keeps just about everything (unless you CAN hear above 19.5kHz). I am honestly quite pleased with this setting (you don't HAVE to add the -b32 -B320. I just like flexability). Thanks to TrNSZ and tigre for shifting me to better LAME settings.

..and I'm using LAME v3.95.1 and EAC v0.95pb4 (Secure Mode) to rip and encode.