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 48kHz in LAME really untuned? (Read 4646 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Is 48kHz in LAME really untuned?

It is usually said on HA that encoding a 48kHz source with LAME is not really recommended as LAME is untuned for sampling rates higher than 44.1kHz. People usually mention that the 44.1kHz modes have been extensively polished, and that you should use it instead.

I don't have a firm grasp of how LAME works, so I don't understand how 44.1kHz could be so different from 48kHz when it comes to tuning. For example, the reference library for Ogg Vorbis uses the same tuning parameters for 44.1kHz and 48kHz, and the developers seemed to be satisfied with the results.

Could there be a way to "reuse" the tuned-44.1kHz parameters in the untuned-48kHz modes? Or is there a lot more that I don't know much about?

Enlighten me, please.

Is 48kHz in LAME really untuned?

Reply #1 –
Well, for a long time people have said that for ~128kbps FhG encoders are superior to Lame.
For a long time people have said that for ~160kbps with Lame abr is better than vbr.
For a long time people have said that for ~128kbps with Lame abr is better than vbr.

Perhaps it is just the same thing with 48kHz encoding...
Honestly I do not know any reliable report regarding the fact that 48kHz encoding would sound bad. (perhaps there are some, in this case I'd like to know)

Is 48kHz in LAME really untuned?

Reply #2 –
I'm pretty sure this originally came from Dibrom when he was first tuning the --alt-presets.
daefeatures.co.uk

Is 48kHz in LAME really untuned?

Reply #3 –
A quick search gave me this:
Quote
No it's not SSRC quality.  For that matter I decided not to use this hq option after all.  The reason is because I've apparently found a few bugs which actually can cause worse quality when using 48khz files.  I don't think this is an issue with resampling, but more an issue with the masking calculations and such in LAME.  I'm not 100% sure what is causing it yet because I haven't had more time to test, but at any rate, it won't be included in the current presets.[a href="index.php?act=findpost&pid=1853"][{POST_SNAPBACK}][/a]

Is 48kHz in LAME really untuned?

Reply #4 –
Quote
A quick search gave me this:

But this is a 3 years old quote

Is 48kHz in LAME really untuned?

Reply #5 –
Yes. I would be inerested in recent ABX results, but unfortunately my hearing isn't that good. (And I'm afraid I don't plan to practice listening for artifacts.)

edit: s/listening to/listening for/g
 

Is 48kHz in LAME really untuned?

Reply #6 –
I was following since years the guides at Doom9 regarding the audio encoding stage of Divx encoding. They produce 48 Khz Lame Encoded mp3 transcoded from AC3 sources. I've encoded one houndred musical films (some 160 ABR, some 128 ABR) and I've never found an annoying artifact present only in the encoded tracks. I can actually hear some artifacts but they are present also in the ac3 or Dolby Sorround sources!
And if really 48 Khz Lame were untuned someone has to tell Doom9 guys to not recommend that type of compression.....