Clipping... let's dumb it down.

Topic: Clipping... let's dumb it down. (Read 4592 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.

Clipping... let's dumb it down.

2005-12-07 06:38:28

(I've searched, but still can't find the info I'm looking for. I'm going to use some rudimentary terms, and make some statements based on assumption. )

As far as I can see, if an mp3 track has clipping then the damage is already done. Logic tells me that the problem occurred during the encoding process. It's my understanding that clipping takes place when a CD is mastered at a level too high for the preset 'record' level used by the encoding software... hence the track is clipped.

Is there any mp3 ripping software that allows you to regulate the 'volume' level of the encoded track? The best analogy I can use dates back to my days of taping CD's on analog cassettes: A Bruce Springsteen CD might have played louder than a Metallica CD. So when I went to record the CD's onto tape, I'd have to set the record level lower for the Springsteen album. Does this make sense?

Thanks for any advice.

Clipping... let's dumb it down.

Reply #1 – 2005-12-07 06:50:26

You can use wavegain on the files prior to encoding, but I'd suggest you look at mp3gain as it is a completely reversible process.

Clipping... let's dumb it down.

Reply #2 – 2005-12-07 07:30:50

also worth noting, "clipping" is a relatively general term, and can occur at many different levels. some cd's come already with some clipping. clipping can also occur during encoding, as you addressed. also, clipping often can occur during playback.

Clipping... let's dumb it down.

Reply #3 – 2005-12-07 09:00:00

Clipping can even occur within individual 'tracks' (is this the right term? I'm referring to instruments/vocals etc. individually) when mastering.

Clipping... let's dumb it down.

Reply #4 – 2005-12-07 09:45:24

Lots of R&B / Rap tracks are mastered to clip badly, to get maximum bass effect, so if the track is already clipping and an mp3 encoder adds just .001% to the signal then it appears to be going over +32767 or -32768, when in reality it is up there in anycase.

Clipping... let's dumb it down.

Reply #5 – 2005-12-07 10:00:26

Bass effect? Hahahaha no! It sounds like utter shit. Very weak, with no "oomph". Too bad music is recorded like that now.

Edit: Just in case, Spoon, I'm not actually disagreeing with you.

Clipping... let's dumb it down.

Reply #6 – 2005-12-07 10:10:38

Quote

As far as I can see, if an mp3 track has clipping then the damage is already done. Logic tells me that the problem occurred during the encoding process. It's my understanding that clipping takes place when a CD is mastered at a level too high for the preset 'record' level used by the encoding software... hence the track is clipped.
[a href="index.php?act=findpost&pid=348248"][{POST_SNAPBACK}][/a]

No, both of your asumptions are wrong.

In your case, there are *three* phases where clipping may occur.

1. During the recording of the CD:
This cannot be fixed. You basically bought a cd with damaged sound on it. This is not unusual - a significant amount of CDs is recorded this way nowadays.

2. NOT during the ENcoding-phase:
MP3 can store higher volume than the original CD. Thus, even if the signal goes above the CD-maximum, it will not clip. In short - encoding itself almost never causes clipping.

3. During the DEcoding-phase(playback):
Even though the signal in the MP3 may not be clipped, it will clip when its decoded and then played. This is because the volume-limit of your soundcard is the same as with normal CDs. Thus, the signal is decoded from the MP3, then doesn't fit anymore in the available space - and clips. This can be avoided two ways: A. The "proper" way: use replaygain or something similiar to scale down the volume of the signal so that it doesn't clip anymore. B: scale down the signal BEFORE encoding it.

- Lyx

Clipping... let's dumb it down.

Reply #7 – 2005-12-07 17:39:47

Quote

2. NOT during the ENcoding-phase:
...
3. During the DEcoding-phase(playback):
... .

This says the encoding process can increase the signal level above that of the wav source, at least at some peaks, but these will not clip. If true, encoding must create numbers greater than the maximum 16 bit value, and these numbers can be stored in the mp3, no? Is there a reference for this information?

If decoding causes clipping because of those higher values created in the encoding process, how could any program correct that post decoding? Once something is clipped, scaling it down to 1/2, or 1/10, can't fix the clip. It reduces the level, but the information from the > 16 bit mp3 number is already gone. If it were possible to decode it into a floating point file, rather than an integer file, then the higher value in the mp3 file could be created in that floating point wav file, and rescaled without significant damage, but decoding into a 16 bit integer file will have lost the peak information before any attempt can be made to fix the problem.

Is it possible, before encoding, to figure out if the problem will be created so that corrective measures may be taken? Any rescaling of an 16 bit wav file will loose lower level resolution, therefore, while some might not care about that in their quest for evening out the volume levels, 'just replaygain them all, is not an attractive option to me. I have never had any complaint about differences in volume, I like the album just the way it was created.

Clipping... let's dumb it down.

Reply #8 – 2005-12-07 18:28:28

Lyx:

If item #2 in your post is correct, then how is it that I can rip a track from a CD, then have it analyzed by mp3gain and find that it's clipped? No playback involved.

Essentially I notice that a new U2 CD comes in at 98db on mp3gain and an old Van Halen CD comes in at 89db. (U2 clips, Van Halen doesn't.) I attribute this to higher mastering levels these days. Isn't there a ripping program that will allow me to control the level at which I digitize my music, so U2 isn't blasting through the roof?

Thanks to everyone for their input.

Clipping... let's dumb it down.

Reply #9 – 2005-12-07 19:35:26

Quote

Quote
2. NOT during the ENcoding-phase:
...
3. During the DEcoding-phase(playback):
... .

This says the encoding process can increase the signal level above that of the wav source, at least at some peaks, but these will not clip. If true, encoding must create numbers greater than the maximum 16 bit value, and these numbers can be stored in the mp3, no? Is there a reference for this information?

Correct. I am quite certain references about this are abundant, but i'm currently too lazy to look it up. Among other things, you can check it yourself - all you need is a replaygain-capable app which can display the peak values. I own tracks which virtually screw LAME to such an extend, that the peak value will be 160%! Such a value couldn't exist, unless: A. it was introduced during encoding... and B: it could be stored in the ENcoded will without clipping.

Quote

If decoding causes clipping because of those higher values created in the encoding process, how could any program correct that post decoding?

By scaling it down during decoding BEFORE it is put into the target output format.

For specific technical details, someone else will need to help out.

Quote

If item #2 in your post is correct, then how is it that I can rip a track from a CD, then have it analyzed by mp3gain and find that it's clipped? No playback involved.

The "clipped" marker in mp3gain does *not* relate to what is ENcoded in the mp3, but instead what would happen if it would be DEcoded at the given volume.

- Lyx

Clipping... let's dumb it down.

Reply #10 – 2005-12-07 23:27:10

@ aea & AndyH-ha: I suggest you to look at this thread: Wavegain vs. MP3Gain, Why the former might be better...

Notice