Likewise can we have a break from entirely unqualified (what rough magnitude of effect are we talking?) and thus wildly speculative suggestions? It's fine to run up the complexity ladder, but it needs to be warranted as you say (or at least first shown potentially important, theoretically or, 1,000x better, empirically). Otherwise we end up with side discussions that lose track of the original point.
(And Mark, this is not meant to be inflammatory, it takes all parties and a bit of burden of proof. Sure, this can be managed better.)
There are purely Fourier techniques for solving circuit problems with non-linearities. Harmonic balance is part of the big simulation packages and takes an ordinary SPICE deck as input. Look up Krylov subspace to make your head spin.
Last edited:
I fear that the keyword there is "steady state".
Although I appreciated to be mentioned too 🙂
Some people found issues with it even in "steady state" and went ballistic on numerical:
https://computation.llnl.gov/casc/sc2001_fliers/NSDE/NSDE01.html
Although I appreciated to be mentioned too 🙂
Some people found issues with it even in "steady state" and went ballistic on numerical:
https://computation.llnl.gov/casc/sc2001_fliers/NSDE/NSDE01.html
Nothing can move faster in metal than the speed of sound in that particular medium; when confronted with a faster than c transient it just get compressed. Metal has rise time limitations so I expect a whole slew of PIM and TIM to occur in this experiment.
It's fair to solve for the response to 10 min of music sampled at 96KHz using these techniques. 57.6 million point FFT's should be no problem for LLNL. It's a mistake to limit steady state to a few sine waves that are assumed to exist for all time or some such. Led Zeppelin playing "Stairway to Heaven" over and over is also a steady state signal. We used this to simulate ADSL symbols into real top level circuits.
Our in-house SPICE was one of the first to include temperature as it's own state variable so I figure the thermal "memory" is covered here too.
Last edited:
But eventually you'll end up with a family of curves at various volume levels, without much information on how the system transitions from/to each other.
You got it pretty good re music selection, the DAC folks are into Rebecca's "Spanish Harlem".
And you don't want to get anywhere near the guys involved with "speech intelligibility". You will end up with "hearing voices" at all times... ouch....
You got it pretty good re music selection, the DAC folks are into Rebecca's "Spanish Harlem".
And you don't want to get anywhere near the guys involved with "speech intelligibility". You will end up with "hearing voices" at all times... ouch....
Last edited:
I'm not sure why to be critical of what Scott said. What they did was difficult and laudable.
It occurs to me this might be a good time to mention Fundamental Attribution Error. https://en.wikipedia.org/wiki/Fundamental_attribution_error It's a type of cognitive illusion, similar to an optical illusion, except people usually recognize it for what it is when they experience an optical illusion. Not so for FAE. It just pops into conscious awareness as obvious truth; it is perceived just as though it is reality when it is in truth an illusion. Why? Probably some survival benefit to having it operate quickly and automatically. It's the tendency to infer people's intentions from their actions as arising from attributes of personal character, as opposed to the circumstances someone may find themselves in. FAE combined with the tendency to focus on any hint of other people's faults and lack of character, combined again with the tendency to see one's self in the most favorable light, tends to lead to problematic conclusions. Happens all the time with people. Very easy for it to happen here. Have to watch out for it.
What made me think of mentioning this now? I read that Scott and his colleagues did something pretty impressive and that it did not seem to be recognized as such.
I don't know for sure, but please let me hypothesize and then somebody can correct me if and where I am mistaken. What Scott and fellows did was perhaps in some way imperfect. The question then arises, if so, why would they choose to do it the way they did? FAE and other cognitive biases could suggest reasons such as: laziness, unwillingness and disinterest in doing the best possible job, stupidity, greed, and so on. However, those are all about aspects of personal character, and are biased towards assumptions that one may be dealing with some bad type of person.
On the other hand, it may not be so obvious to recognize that circumstances usually play a much greater role in why people do what they do than aspects of character. One possible circumstance I have seen before is that although we have extremely powerful computers today, we have problems that still remain impractical to solve on them. They are computationally intractable with current technology. As a result, sometimes people have to use clever workarounds to get the best possible results practical from the computational power they have available. I don't know if that's why Scott and fellows did what they the way they did or not. We would have to ask him why. But better to ask, if there are concerns that better, more accurate ways were possible, before allowing one's self to believe the possible illusion that they could and should have done better.
Last edited:
I am getting a premium vinyl of 'Rebecca's Spanish Harlem' and I want to compare it to the SACD version that I already have. This is an exceptional piece of music.
Cymbals are interesting.
After a single hit multiple frequencies can be heard to beat and interact, and these interactions change during the long decay period.
Multiple hits can cause the cymbal to go into 'chaos' mode causing all manner of sounds/noise, and distinctly disagreeable.
Interestingly, cymbal behaviour can be profoundly modified by placing a small amount of a particular mixture at the centre of the cymbal and this 'chaos' behaviour prevented.
This effect is not due to change in mechanical loading/damping of the cymbal at the mounting centre.
Dan.
Dusty Springfield has a replacement? Certain audiophile circles enjoy the chanteuse torture, I have my secret weapon.
Last edited:
Let me guess - you substituted the lab CD with one where "The Look Of Love" was replaced with Vanilla Fudge's version of it?
😀
You appear to be saying that a mechanical effect (reduction of nonlinear behaviour) does not have a mechanical cause. What does cause it? Little fairies? Electromagnetism? Quantum tunneling?Max Headroom said:
Nice post , thank you.
In pure mathematics there are some internal restrictions and no external (just like in philosophy).
In applied maths there are some more internal restrictions and a few external.
In physics there are some even more internal restrictions and a good amount of external.
In engineering there are all the above internal restrictions and a lot of external restrictions. It’s about implementation of something real in a real world.
There are members who have the desire to kill their time by proposing wild guesses and abstract suggestions.
They either fail to realize that in engineering there is a reality that someone has to face or they are deliberately trolling.
In this thread, members have already acted multiple times in a way to bring things back into a straight line.
Even in a hobby playground as this virtual place called forum, we people have to self control our relationship with reality as it is a technical forum.
There are rules to be adhered to, in this case Rule #2, Note #1
http://www.diyaudio.com/forums/site-announcements/167561-diyaudio-rules.html
Repeated violation leads to binning and or banning.
Not taking any remedial action for the sake of unrestricted activity of the members (“it’s a hobby after all”) leads to thread poisoning.
Such a poisoned thread will eventually be closed.
George
Well back to what is wrong with opamps.
When I purchased my first prime opamps from a local distributor the were Fairchild uA709c's. The price had dropped from $45.00 each to just $9.00. Or in today's money adjusted for inflation about $30.00 each.
Since a popular opamp like the AD797 costs way less than that but is still one of the more expensive units, how can it be any good? 🙂
Then comes the problem with this thread. It is about opamps in general. There are far more opamps that are really not good for audio such as the micro power, comparators etc. than ones that are outstanding for audio. So if I want to compare a well designed discrete circuit to many opamps the descrete most likely will work better for audio. Duh.
Now opamps designed for audio use are really quite good at it. As the market for such products is small the manufacturers prefer to produce general purpose devices. Perhaps the AD797 is a good example. One of the best for audio but I suspect that is only a small part of the sales volume.
When I purchased my first prime opamps from a local distributor the were Fairchild uA709c's. The price had dropped from $45.00 each to just $9.00. Or in today's money adjusted for inflation about $30.00 each.
Since a popular opamp like the AD797 costs way less than that but is still one of the more expensive units, how can it be any good? 🙂
Then comes the problem with this thread. It is about opamps in general. There are far more opamps that are really not good for audio such as the micro power, comparators etc. than ones that are outstanding for audio. So if I want to compare a well designed discrete circuit to many opamps the descrete most likely will work better for audio. Duh.
Now opamps designed for audio use are really quite good at it. As the market for such products is small the manufacturers prefer to produce general purpose devices. Perhaps the AD797 is a good example. One of the best for audio but I suspect that is only a small part of the sales volume.
I think it attests to the universality and all-round 'goodness' of opamps that the same one you call one of the best for audio (and I agree) was also considered the best for LIGO. We live in abundant times, as far as our hobby (sorry George ;-) is concerned.
Jan
Jan
Narrowing the selection window by resorting to spec data sheets? 😉
I would really like to read an insider’s view on this.
I’ve made a wish, who knows?
George
- Status
- Not open for further replies.