Well, I'm experimenting with manual-drawed samples, but the automatic DC offset removal doesn't work well as I think. My request is an manual DC offset window, something like Alt+Click on the button, based on percentage, and if it not too much, the detected (by auto) percentage or value, I don't known how it is implemented.
What would manual DC offset correction that be useful for? OpenMPT can only detect and remove one kind of DC offset (practically a 0hz offset), and the detection on that is always correct if the sample is not broken in another way. DC offset removal does not really work with chip samples and is not intended to be used with them.
Quote from: "Jojo"What would manual DC offset correction that be useful for?
Best for sample and song optimization. Check this (http://www.mediafire.com/download.php?4um3phab3e7y86d) module, render at max preamp, and see the waveform in Audacity or the sample editor, has near 10 percent up. You probably think that is unuseful but you could save dynamic range without clipping <<this is my objetive ::)
QuoteBest for sample and song optimization.
I don't quite understand what you want. Automatic DC offset removal is meant to do exactly that. Comparing your module file without DC offset removed, I get a high overall offset of course (17%), but when I apply DC offset removal on all samples, it is just as low as 0.02% (rendered with the same settings, no normalization). I don't think it can get much better with manual DC offset removal, especially not with chip samples.
Well, is a bit hard to explain, but could you help me by providing an audio editor with manual DC offset? I have to leave my house, I have a commitment :\
Dunno. I just know editors which automatically fix the DC offset, and they were never wrong so far. I'm still not quite sure why you would want manual control over this anyway.
BTW, another technique to remove the offset is to apply a high-pass filter at about 20hz, that will also take care of periodic FM DC offset.
Quote from: "Jojo"BTW, another technique to remove the offset is to apply a high-pass filter at about 20hz, that will also take care of periodic FM DC offset.
Yes, this works fine, and plus remove dc offset & normalize does a very well job. Since I have resolved, the problem was caused by a sample. So... i don't know about keep this thread open, some other tests removes the dc offset perfectly ::)
Well, I don't see a point in keeping it open if you found a solution. :)