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 V0 Not Transparent? (Read 7894 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Lame V0 Not Transparent?

I compared Xenakis' "Noomena" in FLAC vs. LAME v0 mp3, and these were the results:

foo_abx 1.3.4 report
foobar2000 v1.3.1
2014/06/17 21:45:20

File A: F:\Users\Brandon\Music\Xenakis orchestral works\cd1\04-Noomena.mp3
File B: F:\Users\Brandon\Music\Xenakis orchestral works\cd1\04-Noomena.flac

21:45:20 : Test started.
21:47:43 : 01/01  50.0%
21:50:07 : 02/02  25.0%
22:16:18 : 03/03  12.5%
22:25:54 : 04/04  6.3%
22:32:46 : 05/05  3.1%
22:34:15 : 06/06  1.6%
22:37:06 : 07/07  0.8%
22:39:04 : 07/08  3.5%
22:40:15 : 08/09  2.0%
20:07:16 : 09/10  1.1%
20:08:39 : Test finished.

----------
Total: 9/10 (1.1%)

I'm actually a bit surprised.  I'm using HD800 headphones with an O2/ODAC combo.  Should I consider a DAC1?

Lame V0 Not Transparent?

Reply #1 –
Any chance you can upload the problematic segment from both files?

Lame V0 Not Transparent?

Reply #5 –
No one (that I'm aware of and whom I respect) claims that MP3 at its maximum setting is absolutely transparent, no matter what (regardless of listener, equipment, content etc.). So, yes, it's possible that MP3 might not be good enough in your case. Try using different codecs, if possible (e.g. Musepack at "braindead" setting, or Vorbis at some ridiculously high preset). Or just use FLAC exclusively.

I'm not sure but isn't the first column of the ABX log the real time? Then how could you start at 21:45:20 and finish at 20:08:39? Did you do the 10th attempt on the next day?

Lame V0 Not Transparent?

Reply #7 –
Quote from: Martel on
No one (that I'm aware of and whom I respect) claims that MP3 at its maximum setting is absolutely transparent, no matter what (regardless of listener, equipment, content etc.). So, yes, it's possible that MP3 might not be good enough in your case. Try using different codecs, if possible (e.g. Musepack at "braindead" setting, or Vorbis at some ridiculously high preset). Or just use FLAC exclusively.

I'm not sure but isn't the first column of the ABX log the real time? Then how could you start at 21:45:20 and finish at 20:08:39? Did you do the 10th attempt on the next day?


Yes.

Lame V0 Not Transparent?

Reply #8 –
I wouldn't worry about the legality of 12 seconds for encoder analysis. de minimis certainly applies, even if fair use doesn't.

In the original above 16 KHz there's a lot of noise (hiss, dither, and a ~20 KHz tone):



LAME is obliterating all the background noise, yet does preserve the meat of the signal. It thinks the noisy parts above 16 KHz are worth preserving when there's not much else going on:



The MP3 does sound slightly duller to me, though that could be due to what's been done below 16 KHz as well.

I'm wondering if dropping the lowpass slightly or using -Y would result in transparency.

Lame V0 Not Transparent?

Reply #9 –
Quote from: mjb2006 on
I wouldn't worry about the legality of 12 seconds for encoder analysis. de minimis certainly applies, even if fair use doesn't.

In the original above 16 KHz there's a lot of noise (hiss, dither, and a ~20 KHz tone):



LAME is obliterating all the background noise, yet does preserve the meat of the signal. It thinks the noisy parts above 16 KHz are worth preserving when there's not much else going on:



The MP3 does sound slightly duller to me, though that could be due to what's been done below 16 KHz as well.

I'm wondering if dropping the lowpass slightly or using -Y would result in transparency.


Is the original on the right?  I thought the FLAC sounded cleaner.

Lame V0 Not Transparent?

Reply #10 –
Quote
I'm actually a bit surprised. I'm using HD800 headphones with an O2/ODAC combo. Should I consider a DAC1?
A different DAC isn't going to make it any easier or harder to hear compression artifacts.



Lame V0 Not Transparent?

Reply #11 –
Quote from: DVDdoug on
Quote
I'm actually a bit surprised. I'm using HD800 headphones with an O2/ODAC combo. Should I consider a DAC1?
A different DAC isn't going to make it any easier or harder to hear compression artifacts.


What about cryo-treated USB cables?

Lame V0 Not Transparent?

Reply #13 –
Quote from: theriverlethe on
Is the original on the right?

The original is shown in the first image, and the MP3 is the second. The top half of each image is the left channel, the bottom half is the right channel.

I'm curious to what you ABX results you get with something like -V0 --lowpass 18500.

Lame V0 Not Transparent?

Reply #15 –
Quote from: theriverlethe on
I'm actually a bit surprised.  I'm using HD800 headphones with an O2/ODAC combo.  Should I consider a DAC1?
Don't be. I've ABXed 320kbps MP3 on $20 earbuds through integrated sound. MP3 isn't transparent in some places. No one serious about archival quality uses lossy any more.
 

Lame V0 Not Transparent?

Reply #16 –
I was able to ABX your files (9/11) in the 6.5...8.3 sec. range.
I was not able to ABX the lame3100m -V1 --insane-factor 0.75 result.
lame3995o -Q1.7 --lowpass 17