I was't thinking at the time I guess an ideal sine wave with properly dithered data would not have any observable THD against the noise floor. OTOH the same claim is made for the phono add on which is all analog and for phono even more pointless.
Very good remark.
If working to reduce the distortion of an operational amplifier at those unmeasurable levels is understandable, nobody knows what can be the need for scientific or military purpose, what the hell for Audio ?
Harmonics are parts of the original sound, whatever the instrument. At those levels, it is less than moving the microphone to some millimeters, I'm sure NOBODY can tell the difference (Mask effect) with an other that have 10 time more. And, if any, it do not disfigure the instruments and some would prefer the sound of the more distorted one and find-it more realistic and ageable (tubes ?).
Anyway, the instruments will be reproduced by a speaker, that will add a lot of distortion. So the question is: can-we measure those levels of distortions with a microphone ?
I consider this as an other audiophile B.S. An attempt to make some naive people ready to give 10 000$ for a Rasberri with a good looking box around.
Which does not detract from Bruno's technical merit: hats off for the sports performance.
It could be a reference to compare with a cheap one in order to buy the cheap one once no noticeable difference can be found.
How many people, in this forum, will spend this money for such a DAC, when we can find a very satisfying one for 10 to 50 time less and buy hundreds of records with the saved money ?
Last edited:
We cannot hear the difference btw 0.005% distortion and 0.006% distortion but we can hear the difference between amps that measures 0.05% and 0.005% thd while listening on speakers with 0.5% distortions and that might be because :
1: hearing is logarithmic
2:The distortions generated by amps aren't the same with those generated by speakers and our ear is very sensitive to different patterns of distortion.
3.THD might not be the only distortion we hear when listening to different amplifiers and we don't necessarily hear their specs...
Thanks George, this really compliments the Stereophile measurements. The mountain ridge plot is the most relevant one. They are all pointing the same thing. Under 2kHz, it is visible how the mid woofer starts to beam. Then where the horn loaded tweeter comes in, the drop off stops and a plateau becomes visible. This creates the perceived brightness.
Getting a speaker straight on axis used to be non trivial. For one thing timing differences are hard to compensate for using coils and caps. The Sequerra speaker uses a physical offset to get it right, which is probably the best solution. With DSP correction, this all has become easy. (Active analog is in between allpass filters can do the trick here and aren't that hard to do.) In short, going active allows anyone with half a brain to get the on axis FR straight.
The art is now in getting the power response right, think lumen vs lux. It is disappointing to see that KEF falls somewhat short in this respect for living room use with their LS50.
There are basically two ways to enjoy a speaker like this. The first is to listen near field, the other is to heavily dampen the room so that the off axis output is less reflected back to the listener.
There is some scientific underpinning for this in the Haas effect. Early reflections are filtered out by the brain so that only the primary source is consciously perceived. However, reflections do impact on the perceived timbre.
Last edited:
Not many, I doubt they would be attracted to a forum with DIY in it's name.
Didn't JC say he couldn't afford the stuff he designed?
Good post. One of Linkwitz's reasons for going dipole was the even power response, then the Haas effect, if it can be implemented, means little if any room treatment is necessary
It would be hard to find a real physical process linear enough to benefit from the low THD. Noise reduction and removal of interferors and confounders tends to be far more important. Many of the latest SD DAC data sheets don't even suggest non-audio applicatons.
The part I highlight reflect what I think.
The question about HD is, for me to understand if this added distortion is prejudiciable or not to our listening pleasure.
I'm more in concern with IM, that add parasitic signals to the message that do not belong to the natural harmonics and blur the listening.
And time behavior (micro dynamic) that is not measured with any standing waves. Not measured at all, in fact.
By example, the damping factor difference of two amps will modify the waterfall of the same speaker. Nobody seems really interested by this aspect, always focused on HD numbers.
At least, noise reduction threshold is easy to figure out. If I do not hear any white noise from my listening position at high enough music level during the silences, why to bother ? And I will be sure that, if I cannot hear any noise at this level, I will not hear any distortion, masked by the signal itself at the same level. Right ?
I focus on your " Interferors and confounders". Could-you precise ?
Really? Just because you don’t do it or don’t understand it does not mean that it’s not done.
I think Scott was referring to real engineering problems such as LIGO and VIRGO where you require the best minds in many discplines all working together and not a marketing droid in sight
It is audibly distinguishable from an ultra low distortion HPA, and easily so. It sounds quite good to many people though. The initial impression tends to be, 'Wow, this is good!' If one keeps listening then the intermodulation of things like vocal harmonies can be heard. No one would probably know if they didn't know the vocal harmonies were not recorded that way (which was known in the A/B listening case, since the ultra-low distortion HPA did not exhibit the same effect).
The above having been said, I probably should have ignored the bait, which was to the effect that good amplifiers are audibly indistinguishable. I would agree that good low-distortion-type amplifiers probably are indistinguishable under many or even most listening conditions, but not necessarily under all conditions for at least some listeners.
Oh, please. No need to be insultant.
You should be surprised, reading my CV.
I was at the head of the electro-acoustic department of a big hifi manufacturer at the age of 25, worked with Mr. leon (Elipson), Raymond Cooke (Kef), and gerard Gogny (Gego, one of the inventor of Orthophase) to design studio monitors for the state French radio (ORTF). It is a all life passion and i'm > 70 now.
I am not susceptible, but none of these three friends/masters would ever think or say such a thing about me. Rather, we explored together the means of measuring our listening impressions and explaining the phenomena we experienced without being able to explain them at this time. And today, still, we are far from the mark.
Last edited:
I was talking about non-audio applications mainly. Any signal not related to the input, EMI, RFI, supply incursion, would all qualify.
Not sure I agree that my post was insulting. Intentionally condescending? You bet.
If that’s the bar for insult, it is very low around here.
But the preference could still lie with those that distort in some way, however, wouldn't it depend on the complexity of the material? Many talk of 2nd harmonic adding to the pleasure of simpler music, I intuitively think more complex music would suffer.
Congratulations on an impressive CV, no sarcasm intended. I don’t know what that has to do with time domain measurements of amplifiers, though? Can you explain what you think is missing from the suite of measurements offered by an AP instead of pulling the “micro-dynamics” trope out of a hat?
If everyone you do that to were to take it upon themselves to do it right back to you like you deserve, then the forum would deteriorate into flame wars. Most people around here are civil enough and respectful enough of others to keep that from happening. Now and then intentionally rude people have to be dealt with to keep things civil. I hope the moderators are paying attention.
- Status
- Not open for further replies.