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 fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0) (Read 8976 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #50
So, The patched -q3 with noise amp=0 should be same as -q4 (also noise amp=0) in normal versions if i m correct ?

Yeah, I think so. -q3 sets noise_shaping_stop to 1, -q4 sets noise_shaping_stop to 0 - however, that config variable appears to be unused and the other settings are identical.


Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #52
@maikmerten I prefer patched versions, but I think there is no difference between patched q0 and q9.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #53
Did anybody have a listen to the samples I posted at https://hydrogenaud.io/index.php/topic,126120.msg1047065.html#msg1047065 ? It would be useful to know if ABR has the same problem or if I'm imagining things.

I will soon check these.  In my experience abr has similar problem.

 

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #54
MP3 is just so patched, that it:

First: is not MP3 anymore
Second: do not works uniformely and simply anymore

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #55
MP3 is just so patched, that it:

First: is not MP3 anymore
Second: do not works uniformely and simply anymore
What on earth are you talking about?

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #56
@Klymins Thanks for testing. q0 and q9 should still make a (tiny) difference (they're not the same settings, e.g., for searching the best Huffman codes), but these may not be very evident.

@shadowking Looking forward to your findings!

@a.ok.in I cannot follow, can you elaborate?

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #57
By the way, I prefer a FhG MP3Enc encoded CBR 32kbps MP3 over both of these ABR 32kbps examples.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #58
Okay, The winners for me are patched versions.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #59
@shadowking Thanks for your feedback!

@Klymins  I guess FhG MP3Enc chooses other sample rate and lowpass settings. Can you provide information on what FhG chooses?

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #60
@maikmerten
ABR mode is definitely better on patched version. No need to ABX. It is obvious. :)

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #61
@maikmerten I prefer 22050Hz. FhG-encoded CBR 32kbps MP3 is lowpassed by roughly 8kHz in this speed. I always use the lowest quality.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #62
@Klimins It appears you like lowpassed audio.

With 32 kbps mono, FhG appears to choose 22050 Hz sampling rate with a lowpass of ~7800 Hz.
With 32 kbps stereo, FhG appears to choose 16000 Hz sampling rate with a lowpass of ~4900 Hz.

Transferring these settings to LAME, overall results are somewhat comparable, as far as I can tell.

OP's techno/rave sample:

Code: [Select]
lame --resample 16000 --lowpass 4900 -q0 -b 32 rave16.wav rave-lame-patch.mp3

The Moby Dick TTS sample:

Code: [Select]
lame --resample 22500 --lowpass 7800 -q0 -b 32 google-mobydick-male.wav moby-lame-patched.mp3

(LAME encodes with the trivial patch applied)

They still sound *different* IMO, but I guess preferring one over the other is now a matter of personal preference.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #63
@maikmerten Looks like you was not used the lowest quality setting on FhG. Try it, it's cutoff point is a little bit higher at almost 8kHz. Also, I prefer 11025Hz for 32kbps stereo MP3. As my MP3Enc does not encode above 30 seconds and Adobe Flash CS6 does not encode 32kbps stereo, I have to use Fastencc for 32kbps stereo, and it works very well at this sample rate at this bitrate at stereo.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #64
Great... now fastenc... the encoder famous for its pirated copy for Windows ACM codec that had several annoying sound bugs.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #65
Great... now fastenc... the encoder famous for its pirated copy for Windows ACM codec that had several annoying sound bugs.

I didn't experience any sound bugs.

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #66
Browsing through LAME's version history, I determined that the quality regressions (q0 sounding worse than, e.g., q4) were introduced with this change:

https://sourceforge.net/p/lame/svn/6147/


Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #67
There are certain times where the 'Extreme' preset (-V 0 -q 0 --vbr new) gave me artifacts with tracks that have cymbal decays or certain TR-808 sounds. The issues are usually solved by using -f -V 0 --vbr new instead
LAME: -f -V 0 -Y
 Xing: -V150 -X2 -U2 -HF-1 -TX0

Re: LAME fastest encoding(-q 9) gives better quality than the slowest encoding(-q 0)

Reply #68
Although I claim no great wisdom in this matter, my LAME encoder of choice has long been halb27s tuned lame-3.99.5o-Fast-CRC, which you will find at RareWares.