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: A little question (Read 5150 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

A little question

What is --ltq fil and is it recommender over default? I am using --standard with --ms 2.

A little question

Reply #1
ltq -term comes from the words "Level Threshold in Quiet" analogous with ath -term which comes from "Absolute Threshold of Hearing"

--ltq fil sets more sensitive hearing curve, meaning that more quiet and more higher frequencies will be encoded.

--ms 2 is the default setting (enhanced ms -mode) used also with --standard profile, so you don't have to type it again.

Whether --ltq fil is recommended or not, depends on your hearing, quality demands and bitrate demands.
Juha Laaksonheimo

A little question

Reply #2
If you have the same aural anatomy of Filburt, you should use --ltq fil

A little question

Reply #3
OK. I have 1 more question.
How does --cvd works?

A little question

Reply #4
Quote
Originally posted by B@RLoG
OK. I have 1 more question.
How does --cvd works?
cvd (Clear Voice Detection) is able to detect voice-like signals to give a higher quality with voices or sounds with harmonic spectra. It uses special analysis to detect harmonics with varying base frequency -  "normal" psychoacoustics are not able to detect such signals and will add audible noise to these signals.
Juha Laaksonheimo

A little question

Reply #5
Thanks. And now the final question. Just yesterday i encoded several files and got this:
After encoding comes the message that "6 internal clippings occured" and that i must reencode with --scale 0.978 or --xlevel.
What is it? What should i do in this situation?

A little question

Reply #6
You don't need to do anything, my bet is those clippings are not audible. But if they are or you want to be sure, run replaygain on the files with command "replaygain --auto *.mpc" (mppdec needs to be in path).

A little question

Reply #7
Quote
Originally posted by Case
You don't need to do anything, my bet is those clippings are not audible. But if they are or you want to be sure, run replaygain on the files with command "replaygain --auto *.mpc" (mppdec needs to be in path).


Replaygain won't help you!.
Those warnings tells you that "internal" clipping occured, replaygain can only help when the clipping occur during decoding.

Those clipping is because a limitation in the mpc format (one of the few). Somehow the internal representation can't handle the high input.
This is not the same as the usual clipping that replaygain takes care of. The internal clipping is a "real" clipping, that can't be fixed after encoding - the clipped information is lost.

The only way to fix this is using either one of the methods suggested in the warning. --scale to make the clipping don't occur at all, or --xlevel that tells the encoder to handle the clipping in another way.

I hope I have understood this correctly...

/Andreas

A little question

Reply #8
Apparently ancl is right because nobody has said anything.
I have never encountered this encoder clipping, I couldn't even create such a clip even after heavy compressing and amplification of some test clips. I would like to test this myself, so can someone tell me where to get a clip that clips?

Edit: Ah, finally got it to clip (I amplified udial.wav 48dB). The --xlevel file plays incorrectly on Winamp, but it's still clearly different from original even when decoded with mppdec. But I couldn't tell the difference between original, normal encoded and the one encoded with --scale switch.

A little question

Reply #9
What exactly does this switch ( -xlevel ) effect?

A little question

Reply #10
Quote
Originally posted by Northpack
What exactly does this switch ( -xlevel ) effect?


--xlevel: use alternative filterbank clipping solving strategy


Case: You couldn't hear the difference, well, which --scale value did you use? BTW, on almost every CD from the last 2, 3 years, i get "internal clipping" errors, that currently can only reliably fixed with --scale. A prime example is the album "Brandy - Full Moon" (released 2002): Internal clipping in every song. Some songs need --scale 0.90 to fix it, others "only" --scale 0.94.

But scaling is not as bad as one might think, concerning additional information being disregarded because it's not covered by the ATH anymore.. even when scaling to 90%, the bitrate doesn't drop significantly. Also, a negative --ltq_gain setting should make up even for more extreme scaling (in theory). So, there can be an audible drop in loudness (see here), but not necessarily in quality.

A little question

Reply #11
Quote
Originally posted by CiTay

Case: You couldn't hear the difference, well, which --scale value did you use? 

I used the value encoder suggested to get rid of clips.

Quote
BTW, on almost every CD from the last 2, 3 years, i get "internal clipping" errors, that currently can only reliably fixed with --scale. 

Do you hear these clippings in the MPCs if you don't use scale?

I have mostly done my encodings using EAC, and it doesn't show these warnings. I only have one ripped song where I hear clipping, but I don't have the original anymore so I can't test if it's there too. This track is called Kaamos, it's from album Cult by Apocalyptica. Starting from about 4:20 the track clips horribly.
I hope someone has it and can provide sample so I can test it.

Update: I just endoded Finnish metal album They Will Return from Kalmah (released 2002), peak level at 100% on every track and no internal clippings occured. Apparently my music is not prone to these clipping problems.

A little question

Reply #12
Quote
Originally posted by Case

I used the value encoder suggested to get rid of clips.


Do you hear these clippings in the MPCs if you don't use scale?

I have mostly done my encodings using EAC, and it doesn't show these warnings. I only have one ripped song where I hear clipping, but I don't have the original anymore so I can't test if it's there too. This track is called Kaamos, it's from album Cult by Apocalyptica. Starting from about 4:20 the track clips horribly.
I hope someone has it and can provide sample so I can test it.

Update: I just endoded Finnish metal album They Will Return from Kalmah (released 2002), peak level at 100% on every track and no internal clippings occured. Apparently my music is not prone to these clipping problems.


Well, tests have shown that clipping can well be audible (http://www.ff123.net/norm.html), and that's also my own experience. Keep in mind that the "internal clipping" does not refer to clipping during decoding.

EAC doesn't really show this error, because it closes the MPC console windows immediately after the encoding has finished (given that you didn't set "don't open external compressor window"). The only trace of internal clipping is a little beep of the system speaker, and some extra lines of text below the encoding results (if you can spot it that fast before the window closes).

P.S. I hope this will all be history with SV8.

 

A little question

Reply #13
Quote
Originally posted by CiTay
Well, tests have shown that clipping can well be audible (
[url=http://www.ff123.net/norm.html][a href="http://www.ff123.net/norm.html" target="_blank"][a href="http://www.ff123.net/norm.html" target="_blank"][a href="http://www.ff123.net/norm.html" target="_blank"]http://www.ff123.net/norm.html
[/url][/a][/a][/a])

Too bad they only tested mp3s. In conclusions they tell it's fixable with replaygain, but in mpc's case it's not.

Quote
and that's also my own experience. Keep in mind that the "internal clipping" does not refer to clipping during decoding.

Do the original songs also clip, or are they just really heavily compressed? Could you give me a sample of real music that clips, I'd like to see how this happens. I have never used scale, and if the problem is bad I will need to re-rip and encode all my albums.

Quote
EAC doesn't really show this error, because it closes the MPC console windows immediately after the encoding has finished

This is precisely what I meant in my post  That's why I don't know if my encodes have clipped, but I only hear clipping in one track

Quote
P.S. I hope this will all be history with SV8.


I'm sure it will.