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: Why does LAME 3.90.3 sound so good when compared to newer LAME versions? (Read 2416 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Just curious if anyone else has experienced this. Songs encoded with LAME 3.90.3 sound crystal clear and are a joy to listen to, while the same songs encoded with LAME 3.100 or 3.99 don't have this same clarity anymore. Not sure what the correct technical term is. It's barely noticeable, but it's definitely there.

Settings used for LAME 3.90.3:
Code: [Select]
lame.exe --alt-preset standard D:\Project\WAV.wav D:\Project\MP3VBRPreset.mp3

Settings used for newer LAME versions:
Code: [Select]
lame.exe --preset standard D:\Project\WAV.wav D:\Project\MP3VBRPreset.mp3

Has anyone else noticed this?

Re: Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Reply #1 –
Quote from: ClearSky on
Songs encoded with LAME 3.90.3 sound crystal clear and are a joy to listen to, while the same songs encoded with LAME 3.100 or 3.99 don't have this same clarity anymore.
As per the terms of service (#8 specifically) please show an ABX log or similar.
Music: sounds arranged such that they construct feelings.
 

Re: Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Reply #2 –
Quote from: ktf on
Quote from: ClearSky on
Songs encoded with LAME 3.90.3 sound crystal clear and are a joy to listen to, while the same songs encoded with LAME 3.100 or 3.99 don't have this same clarity anymore.
As per the terms of service (#8 specifically) please show an ABX log or similar.
I haven't tried ABX. I will do ABX, and post the log.

Just curious if anyone else has noticed this?

Re: Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Reply #3 –
Ok, I did a quick ABX test, and I couldn't notice the difference.

Here's my ABX log:
Quote
foo_abx 2.1 report
foobar2000 v1.6.16
2023-03-31 08:36:00

File A: Duran Duran - Save a Prayer (2009 Remaster) Lame 3.90.3.mp3
SHA1: 37ec9e8597e63fd6f58893d9baed782d317435cb
File B: Duran Duran - Save a Prayer (2009 Remaster) Lame 3.100.1.mp3
SHA1: ebd8ef8ed38aca35ed7363dfa2713ddae6f8487b

Output:
Default : Primary Sound Driver
Crossfading: NO

08:36:00 : Test started.
08:40:22 : 01/01
08:42:24 : 01/02
08:43:34 : 01/03
08:44:14 : 02/04
08:44:40 : 02/05
08:45:22 : 02/06
08:46:31 : 02/07
08:46:59 : 03/08
08:47:36 : 04/09
08:48:36 : 04/10
08:49:31 : 05/11
08:50:20 : 06/12
08:50:51 : 07/13
08:51:38 : 07/14
08:52:30 : 08/15
08:52:56 : 08/16
08:52:56 : Test finished.

 ----------
Total: 8/16
p-value: 0.5982 (59.82%)

 -- signature --
d88f8b11b6f129f3222b0b2d1242baa4e24e900b

Re: Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Reply #4 –
You can do a few more, but I think results will be the same. You can hear no differences whatsoever.
When in doubt like this one, use ABX testing.
Error 404; signature server not available.

Re: Why does LAME 3.90.3 sound so good when compared to newer LAME versions?

Reply #5 –
Quote from: ClearSky on
Ok, I did a quick ABX test, and I couldn't notice the difference.
Thanks!

So, it seems your own ears have tricked you. This is nothing out of the ordinary, but many people find it hard to believe or hard to admit. It might very well be you expected something to be different, or noticed something you hadn't heard before while listening. On subsequent listening, your ears and brain will 'hear' something because you expect to hear something. AFAIK this is called expectation bias or confirmation bias.

Because it seems no human is exempt from these biases, it is important to use tools to eliminate that bias, like ABX. In the past I made a lot of live recordings of musical ensembles, and often I tried to make mixing decisions blind, by placing two mixes next to each other, switching back and forth quickly a (random) number of times with my eyes covered, and then listening switching between both mixes (still eyes covered), making the decision not knowing which one I was listening to. I felt great working like this, but that feeling might be confirmation bias as well of course  :))
Music: sounds arranged such that they construct feelings.