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: Lame 3.99.3x (Read 7436 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Lame 3.99.3x

I finished work on 3.99.3x. It can be downloaded from here.
The functional extension is invoked by using -Vx+ instead of -Vx.

The functional extension - technically speaking
a) inaccurately encoded frames are avoided
Compared to 3.98.4, Lame 3.99.3 does a good job at avoiding inaccurate frames. It uses a frame packaging strategy which covers the issue while being efficient in not introducing unused data in the output file. Lame 3.99.3 doesn’t use specific restrictions any more for 320 kbps frames. Moreover the way the sfb21 extremely high frequency range is treated helps on the problem.
3.99.3x goes a bit beyond. -V0+ avoids usually 50 to 75 per cent of the out of data space situations left over from -V0. For more details read the doc file provided by the download link.
b) keeping a minimum audio data bitrate
Lame 3.99.3x controls the audio data bitrate and keeps it above a certain value depending on -V level. The control is more stringent compared to my previous 3.98.4x version. (This is also the case for the data space control for avoiding inaccurate frames.)

The functional extension - properties of -the various -Vx+ levels
The functional extension is working from -V7.5+ to -V0+.
For level -V7.5+ to -V2+ it is assumed that users care much about quality and efficiency. That's why bitrate increase from -Vx to -Vx+ is very moderate in this quality level range. Nearly nothing is done for avoiding inaccurate frames - it's not necessary here. Minimum audio data bitrate requirements are not very demanding.
For the levels above -V2+ up to -V0+ it is assumed that users care very much about quality, but not about efficiency. That's why this quality level range covers the average bitrate range from 200 up to nearly 320 kbps. -V1.5+ makes internal use of -V1, -V1+ uses -V0 internally. Minimum audio data requirements howver are still quite moderate up to -V1+. Above -V1+ up to -V0+ -V0 is used internally, and minimum bitrate requirements are becoming more and more severe.

The functional extension - a listening test
a) -V5+ against -V4.75 (135 kbps both on average for my standard test set of various pop music)
-V4.75 is the better choice. For details see the doc file.
b) -V2+ against -V1.55 (200 kbps both on average)
-V2+ is the better choice. For details see the doc file.
c) -V0+ against -V0 (maximum settings both in their environment, unequal contenders of course)
-V0+ is the better choice. For details see the doc file.
d) Comparing with 3.98.4x
For a comparison I ported the 3.99x mechanism back to 3.98. I also transfered some 3.99 goodies (frame packaging strategy for levels up to -V2, sfb21 behavior).
I did all the tests I did for 3.99.3x for this new 3.98.4x. The results were similar, with the exception that 3.98.4x -V5+ makes sense in the 3.98.4 world.
I compared the 3.98.4x results with the 3.99.3x results, and they were in favor of 3.99: As well as for -V5+ as for -V2+ 3.99.3 is to be preferred. For -V0+ none of the versions is better than the other. For details see the doc file here.
The difference is due to the improvements of 3.99 over 3.98 and has nothing to do with the functional extension deschribed here.
lame3995o -Q1.7 --lowpass 17

Lame 3.99.3x

Reply #1
Interesting.

I have tried V0+ for one album. It gives very high bitrate > average 319 kbps . I suppose this is the way it works.   
~17.5 kHz lowpass is incredibly adequate and the encoder sounds amazingly clean and still detailed. I like it.  Will give a detailed try to it. 

Lame 3.99.3x

Reply #2
Thanks for 3.99.3x. I´m going to start testing tomorrow.

Lame 3.99.3x

Reply #3
@IgorC and Hanseat: Your tests are very welcome.
@the moderator who corrected the topic: thank you very much.
lame3995o -Q1.7 --lowpass 17

Lame 3.99.3x

Reply #4
Fantastic, halb27!

Thanks a lot for your efforts! Now, if I could only convince the XLD team to allow standard LAME binaries, so I could swap myself... ;-)

Lame 3.99.3x

Reply #5
Hey halb27, would you mind making and uploading 3.99.5 with your extension? I guess not until it is finalized and released (soon I think). Thanks, and thanks for making my MP3 encoder of choice.

Lame 3.99.3x

Reply #6
Yes, there will be a 3.99.5x, but as you said, not before the official 3.99.5 will be released.

Moreover, I am thinking about some additions, and 3.99.5x could be an occasion to put them in. In case I will it will probably take a bit longer especially as I am busy with other things at the moment.
lame3995o -Q1.7 --lowpass 17

Lame 3.99.3x

Reply #7
hey halb27, just wondering what the status of 3.99.5x is now that 3.99.5 has been officially released. Please excuse my anxiousness. Thanks!

Lame 3.99.3x

Reply #8
Now that 3.99.5 is out I'm going to work with 3.99.5x within the next days.
First I'll find out whether or not it makes sense to take care of some additions. So please be a bit patient at the moment. As long as you don't have the very special problems 3.99.4 and 3.99.5 solved you can use 3.99.3x. That's exactly what motivates me to investigate additional functionality: without it you won't get any advantage of 3.99.5x over 3.99.3x. The bitrate bloat issue for quiet passages doesn't apply to 3.99x, and the other issues fixed with 3.99.4 and 3.99.5 are not relevant to 'normal' usage of Lame.
lame3995o -Q1.7 --lowpass 17