As regards directionality I don't want to be bothered by such things. Nor burn in. My criteria is that the sound has to work pretty well from a cold start, and to continue to improve to and/or plateau at an acceptable level - and never degrade, over time. The latter has been been the killer problem, for me, over the years - and whether that should be called Hirata or PE distortion, or something else is neither here nor there, it just has to be sorted out.
Edit: I might just add that the sound has to work no matter what the room is like - no room treatments, fiddling with positioning, sweet spot agonising. Plonk the speakers down wherever, turn it on and convincing playback should happen. If not, then there is something wrong with the system - the plane is not safe to fly ...
Edit: I might just add that the sound has to work no matter what the room is like - no room treatments, fiddling with positioning, sweet spot agonising. Plonk the speakers down wherever, turn it on and convincing playback should happen. If not, then there is something wrong with the system - the plane is not safe to fly ...
Last edited:
Sorry but letters to the editor written by yourself just shows the rest of us of your stubbornness to accept truth in testing for the past 35 years and that is all.
On that article about double blind testing what I got from that was that the relays used had some type of electrical problem if they could be absolutely identified by the test subjects. The silver platted switches showed they did not have this problem and they should have been substituted into the experiment and the experiment redone. That would have been the better test and could have been used to evaluate the original assumptions or to test for the preferred result. Nobody said that a double blind test can't be a failure or incorrectly implemented, experiments are often a failure due to poor setup or improper procedure. This is exactly why most test protocols are tested by multiple researchers to look for these types of systemic errors. it happens all the time in science. Peer review is a necessary evil most of the time before a real result can be proven or dis-proven. The exact procedure must be well documented so it can be duplicated and confirmed. That is what I take away from that discussion.
Sorry ED, I misunderstood what you said (and I appologize for my hasty reponse) and as you know i have posted plenty of original stuff on non-idealities of components. That being said with respect to noise, the real part of the impedance of almost any capacitor that one might come across in practice is very low. I would find it hard to see it contributing anything of significance in a real circuit. Same with the noise of the DA, it is very difficult to make it show up without huge impedance levels.
Bruce Hofer just last month re-phrased one of my old data sheets on input common mode capacitance. I just don't think there are any arguments for these precision instrument level numbers having anything to do with component preference in listening.
Amplifier designs at the 1ppm level exist without any care beyond good engineering practice.
EDIT - BTW I made an oscillator at the -130dB level 1kHz and 7Vrms without any bespoke components. It's there published for all to see.
Last edited:
And I mentioned to them, after that piece, that using inverting op amps in series reduces distortion even more, wonder how long that will take to be invented there.
BTW ED is Erectile Dysfunction!
1PPM is achieved with good engineering of all issues from PC layout to component selection and design. But I still want 160 dB. One major manufacturer has long designed based on the theory that if you cant' hear below 1% THD there is no reason to design better. Their products do not find acceptance by the golden ear gang. (For actually good reasons, you can hear 15th HD at -100 or more.)
As it is easy to avoid capacitor problems, if you know about them, I see that as a non issue for the skilled. But I can still show you lots of commercial products that have those kinds of problems.
Many many years ago, in the first Disco in town, they had problems. I met with the equipment designers and they were baffled why their products didn't work. They followed the schematics in the National data sheets exactly. (These days data sheets show the power bypass capacitors, then they only mentioned the issue!)
They also used a 120 volt to 125 volt step up transformer for the lighting system to allow for the voltage drop in their power switching circuit. Ever see a transformer rated for 125 volts 30 amps?
Ed
Well the Wavac and any number of SET amps get the golden ear crowd jumping.
I found it, forgot it was in 1994 20yr. ago. It was built ordinary breadboard style. A sensible settling response yielded -130dB, really wild experimenting with almost no control loop gain looked better but not for everyone (many minutes to settle).
EDN -- 11.10.94 Oscillator keeps THD below 1 ppm | EDN
As is obvious to all, it's having the right level of precision in the right places - and that precision or quality may be extremely necessary in the areas related to interference. From my experience things that people jump up and down in excitement about are only marginally relevant, whereas the issues that most only pay lip service to are highly critical.
The 1ppm amplifiers are nigh well useless if they blithely, effectively channel a good chunk of the garbage coming in via the mains straight to the speakers ...
Do you mean a change in level vs frequency in the transformer that is level related?
If it appears as pure noise, yes. If it disturbs the operation of the amplifier, creating IMD products that are audio input signal dependent, in various ways, then not so easy. It's certainly not a published test, in any form as far as I'm aware - at a conventional level of measuring I would like to see spectrum analysis of the handling of a multi-tone audio input, while varying levels of different possible interference frequencies are superimposed on a relatively pure mains feed.
I have searched and have asked for pointers to credible information regarding this...repeatedly
He most likely would to some degree - but how is he measuring it?
I only ask these questions because my ears have repeatedly 'measured' the impact of these things not being done to a sufficient level - that 1ppm figure is nonsense in the face of the obvious, audible distortion resulting from insufficient engineering ...
Edit: some people may query why I listen to music at loud levels - and the answer is that all the inadequacies of the electronics bellow and scream at you - the usual get-out-of-jail-free card is that it's a lousy recording, or the speakers aren't good enough - well, IME, it's always been the electronics at fault, every time ...
Last edited:
Hence as has been stated before, the power supply is as important as the rest of the circuitry, especially in audio where basically you are designing a variable power supply to drive a motor. So you need clean power in...the same with a lot of circuitry, if not all.
Look around for EMC testing and engineering, all this is covered, as is filtering the mains.
- Status
- Not open for further replies.