:: > :: > This is a little bit problematic, because the number of distorted
:: > :: > bands does not tell you the weight of distortions you will get.
:: > :: > What do you think sounds uglier:
:: > :: > 20 distorted bands, each 0.1 dB
:: > :: > or
:: > :: > 1 distorted band by 2 dB
:: > :: > ???
:: > ::
:: > :: I would say that 20 distorted bands at 0.1dB is preferable, and assume
:: > :: that this is the choice of lame's algorithms. Is this assumption wrong ?
:: > ::
:: > Yes and no.
:: >
:: > That depends on the distortion of the undistorted bands. Also undistorted
:: > bands have an distortion, it is negative.
::
:: You are saying : "has a distortion" != "is distorted" ?
::
distorted bands: threshold < noise
undistorted bands: 0 <= noise <= threshold
So calculate: distortion[dB] = 20 dB * lg (noise/threshold)
distorted bands: distortion > 0 dB
undistorted bands: distortion < 0 dB
zero distortion bands: distortion = -oo dB (very unlikely that this happens)
Noise and threshold are voltages (no power, otherwise multiply by 10 dB).
The distortion = max ( x1, ..., xn) model implies an infinite sharp
transition of the distortion reception. But reality is far beyond from this
model.
:: > First you need a table of the probability to hear distortions:
:: >
:: > distortion probability error noise voltage ratio
:: > [d=dB] [p=%] [i]
:: > -10 50.0 0.0000 0.316 : 1
:: > -3 50.8 0.0004 0.708 : 1
:: > -2.5 52.0 0.0025 0.750 : 1
:: > -2 53.2 0.0064 0.794 : 1
:: > -1.5 55.6 0.020 0.841 : 1
:: > -1 57.9 0.040 0.891 : 1
:: > -0.5 63.3 0.116 0.944 : 1
:: > -0.25 66.3 0.176 0.972 : 1
:: > 0 70.2 0.281 1.000 : 1
::
:: isn't 0 dB distortion == no distortion ?
:: How can it be heared ?
::
No: distortion == threshold.
Decibel (dB) is a logarithmic unit. Zero distortion means -oo dB (oo stands
for infinite), tenth voltage -20 dB, half voltage -6 dB, equal voltage +/-0 dB,
double voltage +6 dB.
Do you know Graham Bell?
:: > +0.1 72.2 0.336 1.011 : 1
:: > +0.2 74.2 0.422 1.023 : 1
:: > +0.3 76.1 0.504 1.035 : 1
:: > +0.4 78.0 0.593 1.047 : 1
:: > +0.5 80.0 0.706 1.059 : 1
:: > +1 87.1 1.28 1.122 : 1
:: > +1.5 93.7 2.34 1.189 : 1
:: > +2 97.2 3.65 1.259 : 1
:: > +2.5 99.0 5.43 1.334 : 1
:: > +3 99.8 8.4 1.412 : 1
:: > +3.5 99.9 9.5 1.496 : 1
::
Search also for "masking_lower" in "lame.c".
--
Mit freundlichen Gr��en
Frank Klemm
eMail | [EMAIL PROTECTED] home: [EMAIL PROTECTED]
phone | +49 (3641) 64-2721 home: +49 (3641) 390545
sMail | R.-Breitscheid-Str. 43, 07747 Jena, Germany
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
