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: Is LAME 320k CBR any better than V0? (Read 1578 times) previous topic - next topic - Topic derived from Re: Lower algorithm q...
0 Members and 1 Guest are viewing this topic.

Is LAME 320k CBR any better than V0?

I downloaded Fatboy sample of LAME 3.100  V0 and CBR 320 from here. https://hydrogenaud.io/index.php/topic,121579.0.html

320k CBR sounds really strange and worse than V0 or V2.  No need to do ABX as sample is really hard for all encoders. Here is a log of blind test between different encoders.

Code: [Select]
ABC/HR for Java, Version 0.53a, 20 October 2024
Testname: 

Tester: 

1L = LAME MP3 V0\13 fatboy_30sec.wav
2L = LAME MP3 192 kbps\13 fatboy_30sec.wav
3R = Apple AAC 192 kbps\13 fatboy_30sec.wav
4R = LAME MP3 320k\13 fatboy_30sec.wav

Ratings on a scale from 1.0 to 5.0

---------------------------------------
General Comments: 
---------------------------------------
1L File: LAME MP3 V0\13 fatboy_30sec.wav
1L Rating: 4.4
1L Comment: 
---------------------------------------
2L File: LAME MP3 192 kbps\13 fatboy_30sec.wav
2L Rating: 4.2
2L Comment: clipping
---------------------------------------
3R File: Apple AAC 192 kbps\13 fatboy_30sec.wav
3R Rating: 4.8
3R Comment: 
---------------------------------------
4R File: LAME MP3 320k\13 fatboy_30sec.wav
4R Rating: 3.6
4R Comment: 
---------------------------------------

ABX Results:

 Could it be related to this -q issue ?
Edit: Probably not, as files have been encoded with default -q 3 option. 
Edit 2 :Encspot shows  that 320 CBR has lower  "Big values" of bit reservoir during first half of sample, that's where all artifacts occured.

Re: Is LAME 320k CBR any better than V0?

Reply #2 –
Resampling to 48000 is a phenom progress.
I see no reason for me to stick to 44100.
 

Re: Is LAME 320k CBR any better than V0?

Reply #3 –
Quote from: Kamedo2 on
More likely because of more aggressive use of the bit reservoir in VBR. https://hydrogenaud.io/index.php?topic=85135.msg864674#msg864674
Thanks for the link.

I've performed test to check whether it's not issue with  high -q0-3 being inferior -q4.   https://hydrogenaud.io/index.php/topic,125216.25.html
-q4 isn't much better (if at all). V0 is still better than 320k CBR .

Code: [Select]
1L = C:\Test\Fatboy LAME\13 fatboy_30sec V0.wav
2R = C:\Test\Fatboy LAME\13 fatboy_30sec CBR 320.wav
3L = C:\Test\Fatboy LAME\13 fatboy_30sec CBR 320 q4.wav

---------------------------------------
General Comments:

---------------------------------------
1L File: C:\Test\Fatboy LAME\13 fatboy_30sec V0.wav
1L Rating: 4.5
1L Comment: 
---------------------------------------
2R File: C:\Test\Fatboy LAME\13 fatboy_30sec CBR 320.wav
2R Rating: 3.5
2R Comment: 
---------------------------------------
3L File: C:\Test\Fatboy LAME\13 fatboy_30sec CBR 320 q4.wav
3L Rating: 3.6
3L Comment: 
---------------------------------------
ABX Results:

Re: Is LAME 320k CBR any better than V0?

Reply #4 –
Quote from: btc on
I've performed test to check whether it's not issue with  high -q0-3 being inferior -q4.   https://hydrogenaud.io/index.php/topic,125216.25.html
-q4 isn't much better (if at all). V0 is still better than 320k CBR .

Thank you for performing additional test, now it's clear that -q does not matter much, either way V0 is clearly better than the 320k CBR.

Re: Is LAME 320k CBR any better than V0?

Reply #5 –
With that in mind, can we still safely assume the description of 320k CBR in the wiki for Lame's best settings still hold true, not needing any further rephrasing?

Edit: more specifically, its final "disclaimer" regarding the lack of ABX testing heretofore, which, IMHO, seems to not only be debunked by what we've got here, but to have been kinda of turned upside down, if I may say so.

PS: That, or I may just be reading too much into it.  ::)

Quote
[CBR] Very high quality: HiFi, home, or quiet listening, with maximum file size
-b 320
is an alternative to the VBR settings above.
This CBR mode will maximize the MP3's bitrate and overall file size. The extra space may allow for some parts of the audio to be compressed with fewer sacrifices, but to date, no one has produced ABX test results demonstrating that perceived quality is ever better than the highest VBR profiles described above.
• The older, the lossier
• Listen to the music, not the media it's on.

Re: Is LAME 320k CBR any better than V0?

Reply #6 –
Hello everyone!

Bit reservoir... woow, this was unknown to me as a newbie!
When someone said, that VBR can be better than CBR in certain circumstances, I was always surprised.
I thought, "more than full" i.e. 320 kbps in one frame is not possible.
But if I understand it correctly, simply said, if not needed, information can be stored in frames before and after "dynamic parts" and thus the 320 kbps can be "exceeded".
Is this a big thing or are this rather rare circumstances where it really makes a difference?

If I could add a few OT questions:
An MP3 file created with "-b 320" (my so far used setting, because I thought it was the very best) has (according to MediaInfo) the encoding settings "-m j -V 4 ​​-q 3 -lowpass 20.5".
What does the "-V 4" do here? Isn't that actually for VBR?

If I disable the lowpass using "--lowpass -1", Spek shows me more blue peaks above 20 kHz (up to 22 kHz), but still significantly less than the FLAC/WAV file. Is this a limitation of MP3?
Overall, it's better not to remove the lowpass anyway - right?

Is it normal, that Lame doesn't always stick to the switches when the lowpass is set?
For example, "-b 320 --lowpass 10" should actually create a CBR with 320 kbps - but the result is one with 160 kbps!?

Many thanks and regards,
Martin

Re: Is LAME 320k CBR any better than V0?

Reply #7 –
Quote from: m.fessler on
But if I understand it correctly, simply said, if not needed, information can be stored in frames before and after "dynamic parts" and thus the 320 kbps can be "exceeded".
Is this a big thing or are this rather rare circumstances where it really makes a difference?

It's not gonna be a big thing at 320, because that is by and large good enough: artefacts are rarely audible and even then not annoying, even when using CBR.

It is a big thing in principle, and it is going to kick in at low enough bitrates. Imagine the track starts with silence:
CBR at B bits: The beginning of the track must spend B bits per second.
VBR at B bits: There is no value added spending more than a bare minimum (32 bits/s?) on the actual first part of the track, that leaves us bits to spare - let's spend the vacant bits on something we're gonna use later.

(However, VBR will "only on average material" return the average bit rate. That is part of the purpose: not only allocate bits between different parts of a track, but also spend more bits on difficult-to-encode files - hey, that's what a lossless does!
But when you force it to spend near 320 on each file, you cannot exploit that variation between files so much.)