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: Weird tremelo effect -q 0 (Read 6332 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Weird tremelo effect -q 0

I was encoding some recordings for a friend and i used the following code to get a strange tremelo effect

Code: [Select]
lame --abr 128 -q 0 --lowpass 19 file.wav file.mp3


after noticing the weird sound, i changed the code from -q 0 to -q 1 and it raised the bitrate 20kbps

Does anyone get this sound on strong sounds like hitting an acoustic string with a pick?

Link to files

http://rapidshare.de/files/23816547/lame_e..._files.rar.html

Weird tremelo effect -q 0

Reply #1
I was encoding some recordings for a friend and i used the following code to get a strange tremelo effect

Code: [Select]
lame --abr 128 -q 0 --lowpass 19 file.wav file.mp3


after noticing the weird sound, i changed the code from -q 0 to -q 1 and it raised the bitrate 20kbps

Does anyone get this sound on strong sounds like hitting an acoustic string with a pick?

Link to files

http://rapidshare.de/files/23816547/lame_e..._files.rar.html

try --preset 128, is the tremolo "effect" still present?

Weird tremelo effect -q 0

Reply #2
Correct me if I'm wrong but I'd say this is not a good idea when encoding at lower bitrates:

--lowpass 19

From what I recall, this was one of the biggest reasons why Blade sounded so terrible at lower bitrates, even at 192.

Edit: spelling

Weird tremelo effect -q 0

Reply #3
Or instead of trying custom command lines that sound lousy, just use -b128, or -V5 if you wanna go with VBR.

Weird tremelo effect -q 0

Reply #4
True, but -V5 isn't really 128k. Its a quality mode around 140k - bitrate isn't as predictable, though it can average around 128k in some cases. When we say that 128k quality has increased lately , we must take these things into account (for LAME mp3 anyway). CBR128 of yesteryear and VBR quality mode of today are different beasts.

Weird tremelo effect -q 0

Reply #5
-V4 gives me average of 140, personally. -V5 is less than that on average.

Weird tremelo effect -q 0

Reply #6
I was encoding some recordings for a friend and i used the following code to get a strange tremelo effect

Code: [Select]
lame --abr 128 -q 0 --lowpass 19 file.wav file.mp3


after noticing the weird sound, i changed the code from -q 0 to -q 1 and it raised the bitrate 20kbps


The original poster said that lame 3.98a4 --abr 128 -q 0 --lowpass 19 causes a artifact whereas --abr 128 -q 1 --lowpass 19 increases the bitrate 20 kbps but (If I understand correctly) it fixes the problem.



Weird tremelo effect -q 0

Reply #9
Just click the free- button.

Weird tremelo effect -q 0

Reply #10
For the LAME MP3 encoder at least, having the lowpass set high doesnt seem to lower quality, i only lower it when i can audibly hear artifacts, otherwise it sounds much more crisp.  But I was specifically speaking about the -q switch, -q 0 is supposed to be more efficient and it made a large decrease in the bitrate but also added in its own tremelo artifact.  The reason i posted this is because i dont think its supposed to be doing that at least =\.


Weird tremelo effect -q 0

Reply #12
Could you tell me where exactly in your files (1'55'') are located these tremolos? I didn't noticed anything wrong with both encodings.

N.B.1 / uploading original samples (i.e. not compressed one) is recommanded: for comparison first, and then for testing additionnal switch or encoders.

N.B.2 / It may be worth mentionning that your files are monophonic. It might explain why the target bitrate isn't respected at all with -q0.
Wavpack Hybrid -c4hx6

Weird tremelo effect -q 0

Reply #13
Yes I know the file is mono, i want the quality to match about 128kbps cbr on one channel or 256kbps stereo.  If you listen to strong notes, there seems to be some garbling or something. Here are just two small sample mp3's I encoded with the original wav file.

This time I took off the lowpass switch to let LAME figure it out itself, but it raised it anyway to about 20k.

http://rapidshare.de/files/23910274/problem_samples.rar.html

If you switch between each sample, the one with the -q 0 switch has a strange artifact on solid frequencies.

Edit: the point is that -q 0 should produce the highest quality  compared to -q 1 and up, but it seems to be doing too much and creating artifacts otherwise.

Edit 2: I used '--abr 128 -q x file.wav file.mp3',  im not really a fan of using presets, they seem to get me lower quality per bitrate.

Weird tremelo effect -q 0

Reply #14
I confirm the existence of "tremolos". It reminds me the kind of artefacts audible with harpsichord encoded with MP3.

Code: [Select]
foo_abx 1.3 report
foobar2000 v0.9.2
2006/06/23 20:29:42

File A: C:\essai\lame_encode_problem_files.rar|random_stuff_small.mp3
File B: C:\essai\lame_encode_problem_files.rar|random_stuff_small2.mp3

20:29:42 : Test started.
20:30:32 : 01/01  50.0%
20:30:35 : 02/02  25.0%
20:30:39 : 03/03  12.5%
20:30:44 : 04/04  6.3%
20:30:48 : 05/05  3.1%
20:30:52 : 06/06  1.6%
20:31:06 : 07/07  0.8%
20:31:11 : 08/08  0.4%
20:31:15 : 09/09  0.2%
20:31:18 : 10/10  0.1%
20:31:21 : 11/11  0.0%
20:31:25 : 12/12  0.0%
20:31:26 : Test finished.

 ----------
Total: 12/12 (0.0%)
tested range: 0:25.000 - 0:27.000

Quote
Edit 2: I used '--abr 128 -q x file.wav file.mp3',  im not really a fan of using presets, they seem to get me lower quality per bitrate.
It would be surprising... People are not constantly working and tuning presets just to make the quality lower. There are faster way to achieve the same purpose
Wavpack Hybrid -c4hx6

Weird tremelo effect -q 0

Reply #15
Yes I know the file is mono, i want the quality to match about 128kbps cbr on one channel or 256kbps stereo.

...now you tell us.

Nevermind what I said earlier about setting the lowpass to 19kHz.  At 128kbps per channel it doesn't exactly apply.

 

Weird tremelo effect -q 0

Reply #16
Also, you can't really directly compare mono vs stereo in terms of bitrate. ie, you can't say that 256kbps stereo will be the same as 128kbps mono, thanks to join stereo (and it's also why saying "64kbps per channel" is nonsense, the bitrate isn't split evenly in JS.)

Weird tremelo effect -q 0

Reply #17
Also, you can't really directly compare mono vs stereo in terms of bitrate. ie, you can't say that 256kbps stereo will be the same as 128kbps mono, thanks to join stereo (and it's also why saying "64kbps per channel" is nonsense, the bitrate isn't split evenly in JS.)

Right, although I don't even think you can really talk about splitting bits per channel in JS since M/S isn't the same thing as L/R.

And it's really only fair to talk about bits/channel when comparing mono and dual channel mode since the bitrate doesn't have to be split evenly in stereo mode.

But the OP's point is a curious one.  Why would -q 0 deliver a poorer-sounding file than -q 1?

Weird tremelo effect -q 0

Reply #18
Well, iirc, -q0 would cause artifacts in older versions of LAME (and as such was not recommended). Perhaps it's still present in 3.98a4.