[Typhoon859]

TLDR at the bottom (may be much to read but answers if you know I think would be simple)
 
Hi guys.  So MP3Gain claims that many files are presently clipping and will continue clipping even when the dB values are lowered.  Some of it doesn't make sense to me/I'd like some confirmation on my assumptions. 
 
The Alan Parsons Project Greatest Hits is the album I entered the songs of into the program.  I find it to have the perfect volume, listening in the car, headphones - wherever.  I took the average dB value of the track analysis done by MP3Gain which was 96.7733333 so I just decided to use 96.7.  The values range from 95.1-98.8.  That not being of utmost relevance, my query is really about the following.  About 24 of the 30 total tracks which I have there are red with a "Y" under the "clipping" column.  Not really knowing anything for sure, even though the selected tracks aren't all consistently above a certain dB value, that's irrelevant and it's just saying that there are parts in those song that peak (when it's above ___ value relative to the max of a 16-bit 44.1kHz file), right?
 
Well firstly I'd like to mention that if there is clipping occurring, I definitely haven't heard any, regardless of the output.  That being said, Alan Parsons is a brilliant Audio Engineer - world renowned as one of the best, clipping obviously being one of the things he has always been cautious of.  There could be some other technical issues involved here, but regardless, the accuracy of these indications wouldn't seem to be relevant since I honestly hear no problems and therefore wouldn't care if those files are actually "clipping" or not. 
 
The next "clip()" album I'm guessing indicates if the track will clip after the targeted volume would be set.  If that's true, it appears as though for all the tracks to be satisfied, the target volume would need to be 93.1.  I guess that's not pertinent though considering the mentioned above and the fact that I really just want to lower volumes louder than the desired.  So, do I just ignore all this? 
 
The problem then becomes, how do I do this?  In Foobar2000, when converting a file with tags for ReplayGain, it can be set to only apply gain, only reduce, or do both depending on the file.  In MP3Gain (which is what I need to use since my player, the Cowon J3, doesn't read ReplayGain metadata), that option isn't made clear.  All I see of any relevance is an option not to clip when doing Track Gain which in itself I don't quite understand the effect of.  What I would probably wanna do after lowering everything to 96.7 is raise everything below 93.7 to 93.7.  I could technically do both manually by first doing Track Analysis on everything, removing everything 96.7 or below, applying gain to what's left, then doing the opposite for 93.7. 
 
This leaves the inaccuracy of MP3Gain in question.  Firstly, it seems to only lower/raise everything by step values of 1.5.  That on its own is ridiculous.  Now, secondly, it adjusts values +/- .8dB of the targeted volume.  Technically, that would make each track at least .1dB more accurate which I feel isn't even close worth an adjustment of 1.5 for such a small "improvement", especially when it's just taking a positive of a similar volume to the corrected and bringing it to the other side (or the negative and bringing it to the positive).  I'd rather hear it as is.  I have a feeling I will be told this cannot be improved/fixed.  If that's the case, does anyone have any suggestions for a potential way of accomplishing what I'd like to whether it's a different program or an alternate method altogether? 
 
 
For more information on my intentions here with my music, well, it's really just the obvious; it's what everyone wants to accomplish.  The only thing is, raising/lowering everything to the same value is not a solution, especially to a low value like the recommended.  It is said that technically no quality is lost through this process but I would disagree.  Sounds/ambiance in the quieter ranges of songs is essentially lost/lowered to a point where its potential purpose is lost.  Even if that point isn't accurate so to speak, the higher the value of reduction on a song is, it seemed to me the muddier it also began to sound (especially heavy metal, e.g. bands like Disturbed), regardless of any reason I can come up with.  I am sure of this.  Plus, amplifiers/players or whatever need to work harder, meaning, more battery life being wasted and/or (most importantly) more noise getting added pointlessly into the mix, which is not only annoying but once again effects sounds in the quieter ranges since it essentially overlaps them.  Whether any of what was mentioned is directly the fault of the developed method here really doesn't matter since nothing can really be changed.
 
The problem with everything being set to the same dB value is that things that should be quieter end up "to the human ear" ACTUALLY sounding louder than the things that SHOULD sound loud.  For example, Enya/soft Classical pieces should not make you want to lower your average maintained volume of 50% (for instance) while Metallica/DragonForce, or whatever, makes you want to raise it.  That just makes the whole thing backwards, more confusing, and therefore more annoying.  That happens when quiet tracks are made louder and loud tracks are made quieter to supposedly the same level (not exactly sure how that level is measured and found). 
 
 
 
TLDR-
1) Is clipping really relevant unless MP3Gain is indicating that the track is clipping after applied gain (assuming what I understand about the indicated values is correct)?  And technically, how about even then?..
2) Is there a way to apply ONLY reduction where it'd be necessary or vice versa? 
3) Is there a way to get MP3Gain to make more accurate adjustments, or otherwise, some alternate option to this program? 
 
Thank you all very much.  Any help is appreciated, sincerely. 
 
 
-David

 

[Typhoon859]

Way too long or what? >.< 
 
Not everything has to be read/answered.  Any responses to anything would be helpful really.  Does anyone know?.. 
 
Thanks again

 

[Typhoon859]

Is nobody here interested in balancing their music? 
 
If anybody has found a way that works for them, maybe you guys can share it?  If there's a better way/just a way that works well, that could pretty much be an alternate answer to all my inquiries.  Please?  I don't normally do this and feel kind of stupid for essentially begging but I'm just a bit surprised that nobody had anything to say.  I don't know what more to give back than a thanks >.<

 

[Megaohmz]

Regarding your question of "can audio still be clipping even if the volume is turned down". Well I have to say it will still be clipping if you have a source file and it was recorded WITH the clipping. I remaster all of my music before I re-encode to a 320 CBR mp3. I am very careful not to make the music clip, by turning the main volume down and checking the volume meters making sure I am not peaking in the red. I also run a diagnostic on the file to see how much of the file has clipping artifacts, then I fix them. I will normalize the volume to -0.1 dB and this generally fixes clipping problems. There is also a clipping repair tool you can use that works "ok". I use Adobe Audition CS5.5 to remaster my music. This is a very good program to use to ensure your remasters are as clean as possible. It takes some time to learn the in and outs of this program but if you spend a good week watching tutorials and practicing you will be very happy with it.
 
To make it more clear on how a file can still be clipping even if you turn the volume down, well if the file is clipping when you convert it to an mp3 or any other format, you will actually record the clipping. Clipping is when the sine wave looks like a flat square wave at the top and bottom of the peak of the sine wave, instead of being a smooth round oval shape. If you ever listen to a square wave you will notice how "bad" it sounds. This is a very unpleasant sound for people, and results in the "clipping". This is why it in very important to learn how to properly read your dB meters while you are remastering, so you don't make tracks that are forever clipping. It is always better to turn it down a little more while you are applying an EQ, because when you normalize to -0.1 dB it will turn the volume up as far as it will go before it starts to clip, with -0.1 dB to spare (as a buffer).
 
By the way if you properly remaster your track, it will have a completely different "look" on the editor screen, where you can actually see the sine waves. The appearance will have much less of an even volume appearance at the peaks, and some of the sounds will look like (and they are) they are much louder than the other sounds. This is a much more realistic way of recording music. This is why most CDs sound so terrible, the fact that the audio engineers don't know what the heck they are doing, and give all the loud peaks of sound all the same exact peak levels. They also don't consider the way humans perceive sound and don't boost the frequencies that are hard for us to hear, which is generally bass and treble, and truncate the sounds that we hear easily, the 1.5 kHz to 3 kHz freqs.     
 
I hope this helps.
 
 

 

[Typhoon859]

Quote:

Regarding your question of "can audio still be clipping even if the volume is turned down". Well I have to say it will still be clipping if you have a source file and it was recorded WITH the clipping. I remaster all of my music before I re-encode to a 320 CBR mp3. I am very careful not to make the music clip, by turning the main volume down and checking the volume meters making sure I am not peaking in the red. I also run a diagnostic on the file to see how much of the file has clipping artifacts, then I fix them. I will normalize the volume to -0.1 dB and this generally fixes clipping problems. There is also a clipping repair tool you can use that works "ok". I use Adobe Audition CS5.5 to remaster my music. This is a very good program to use to ensure your remasters are as clean as possible. It takes some time to learn the in and outs of this program but if you spend a good week watching tutorials and practicing you will be very happy with it.
 
To make it more clear on how a file can still be clipping even if you turn the volume down, well if the file is clipping when you convert it to an mp3 or any other format, you will actually record the clipping. Clipping is when the sine wave looks like a flat square wave at the top and bottom of the peak of the sine wave, instead of being a smooth round oval shape. If you ever listen to a square wave you will notice how "bad" it sounds. This is a very unpleasant sound for people, and results in the "clipping". This is why it in very important to learn how to properly read your dB meters while you are remastering, so you don't make tracks that are forever clipping. It is always better to turn it down a little more while you are applying an EQ, because when you normalize to -0.1 dB it will turn the volume up as far as it will go before it starts to clip, with -0.1 dB to spare (as a buffer).
 
By the way if you properly remaster your track, it will have a completely different "look" on the editor screen, where you can actually see the sine waves. The appearance will have much less of an even volume appearance at the peaks, and some of the sounds will look like (and they are) they are much louder than the other sounds. This is a much more realistic way of recording music. This is why most CDs sound so terrible, the fact that the audio engineers don't know what the heck they are doing, and give all the loud peaks of sound all the same exact peak levels. They also don't consider the way humans perceive sound and don't boost the frequencies that are hard for us to hear, which is generally bass and treble, and truncate the sounds that we hear easily, the 1.5 kHz to 3 kHz freqs.     
 
I hope this helps.
 
 

Thank you.  I am very interested in these kind of things and am actually going for Engineering in Sound & Acoustics (it involves pretty much all that has to do with sound and technology).  I understand what clipping is and this was some good info.  It was in fact useful for me and I sorta figured things out with MP3Gain.  I assumed that everything was taken into consideration but apparently, there's just many things the program lacks which I had to find ways around for.  In terms of clipping, I have too many songs to go through with an individual approach so what I figured out with my current method (based around MP3Gain) will have to suffice.  My friend and I might make a modification with the necessary added tools/functionality (in its basic form).  It'll of course still just be a form of implementation of ReplayGain. 
 
I'd like to say I know all that you mean by "Remaster" but I don't.  Out of curiosity, what would be the full list of things that you could/would potentially alter in a track you may have?  It seems like you're talking about doing more than just adjusting volume.  The way I see it, if the song was done the way it was, it's how it should be, even if it sucks.  For all you know, maybe the artist wanted it to sound this way.  I look for the best mastering of the song of course but apart from that, I just make sure to convert it properly and balance volumes, sometimes even at the expense of clipping. 
 
Keep in mind, this is obviously not how I'd operate for recording my own music or just in general.  But anyway, thanks again for your reply