|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Follow Ups:
Kurt,what do you mean by ideal distortion? Is it 12dB/oct? One possible explanation: fourier amplitudes of square wave behave like 1/n where n=#harmonic. If you use ideal distortion than it generates harmonic amplitudes with 1/n^2 behavior. Thus you restore original input with error behaving like 1/n for large n. This might explain why this effect is not good at LF - there you have big error. This also suggests that there will be no restoration effect with more complex signals. Hope this makes some sense.
N-set
The "ideal distortion" is described in Feb. and March issues of AudioXpress magazine, an article "Harmony and Distortion", by Jean Hiraga, written in 1981. This "ideal distortion" was found experimentally in 1930 by Wegel and Lane. Later reexamined by Kuriyakawa and Kameoka of Toshiba Labs in the 1960's. In the article, there's Figure 9 and 10 that "shows" that complex signals pass relatively cleanly to the output in the time domain compared to other series of distortions.The distortions listed are:
2nd: -22 dB
3rd: -25 dB
4th: -30 dB
5th: -37 dB
6th: -46 dB
...This distortion is about 10% THD and yet it is high fidelity, subjectively. There are lots of even and odd order distortion, something that should seem to wreck a square wave, and it doesn't. The 3rd masks the 2nd, the 4th masks the 3rd, the 5th masks the 4th, and so on.
The graph I drew is a calculation of a square wave FFT to the 11th harmonic as an input and then multiplied by the series of this "ideal distortion" for an output.
Can't draw any definitive conclusions, but appearances are interesting nonetheless.
Kurt
Kurt, where can one get those articles? Somehow aviable through the web? What concerns appearances that more odd curve on your plot doesn't look bad either.N-set
http://www.audioxpress.com/Ask for the Feb and March 2002 issues, especially Part 2 in the March issue.
Regarding the curves shown, the more odd order distortion appears good, too, and I had one of a more even order that appeared good, almost like the odd order distortion. But both did not reduce the ripple as much and the "ideal distortion" trace had the lowest amplitude bumps and the least "Gibbs phenomena" pre-ringing amplitude and yet had the fastest risetime by a notch, all at once. The odd and even distortion traces were actually pretty similar. If you look closely, the second point in the "ideal distortion" trace is off the mark, an error in one cell of my Excel calculations, because I failed to restore it properly after some editing. Minor glitch.
Kurt
Hi,
Tnx, did...
Do not understand though what this would mean on less distorted signals or other (complex) waveforms...
Any help to pop in? as I can't figure it out (donno where to start the denestling so I need a nestor so to say).mvh /Pär
It's a pretty complex issue. The whole premise for the investigation comes from that Hiraga article. A 400 Hz sinewave distorted by the "ideal spectrum of distortion" sounds identical to a pure 400 Hz tone, yet it appears well distorted. The 2nd harmonic gets masked by the 3rd and the 3rd by the 4th, etc. all the way out to the limit of hearing.This ringing square wave input is what you usually get from a CD with digital filtering, as another example. The amp can smooth out the ringing through the ideal distortion. In other words, it makes digital sound more natural. It apparently subjectively makes other artifacts sound more natural. You lose the effect as the frequency gets higher where some of the needed higher order distortions aren't audible to properly mask. The effect may not be good for low notes, either, according to jj. So it appears this will help the midrange best, which is something SETs are notable for.
It's peculiar and I'm still trying to figure it out exactly because my results don't fully corroborate all of Hiraga's claims, but I can buy into the psychoacoustic argument that it works.
Kurt
Hi,Thanks a lot for the further explanation (or second different writing of the same thing as in linked post).
To sum it up: I need more distortion coz I still prefere Deccagold/DeccaInt/TD124 to Philips CD104 :-)
For me its back to the concertina OPT stage...
mvh /Pär
This post is made possible by the generous support of people like you and our sponsors: