I'd generally expect the decoupling caps to handle most of that, but I could certainly imagine an inductor 'flexing' in response to load changes. The coils in the old CRT TVs used put out a loud whistle at 15kHz - quite a racket to a young person's ears as I recall.
Come to think of it, I think I've also heard noises emanating purely from a transistor when switching heavy loads at audio frequencies - presumably contraction and expansion of some gubbins within it.
Not exactly deafening, but real noises that might even be synchronised in some way to the audio we're listening to. In a world where people imagine they're hearing quantisation noise measured at -110dB, maybe the acoustic noise from the PSU is even louder.
CoperTop,
to pay the price we do for the electonics we play with today somthing has to be compromised, and that is usualy the PSU, they are probably the most costly part of the system (if), so they provide the biggest cost savings. Hence we suffer, and argue endlessly about flawed electronics, such as battery supplies
The noise prototype supplies and their related equipment make can be very audible, many years ago a guy writing the drivers for a prottype floppy disc drive and PSU could get it to play tunes! And nasty SMPS's sound like tinitus.
to pay the price we do for the electonics we play with today somthing has to be compromised, and that is usualy the PSU, they are probably the most costly part of the system (if), so they provide the biggest cost savings. Hence we suffer, and argue endlessly about flawed electronics, such as battery supplies
The noise prototype supplies and their related equipment make can be very audible, many years ago a guy writing the drivers for a prottype floppy disc drive and PSU could get it to play tunes! And nasty SMPS's sound like tinitus.CoperTop,
to pay the price we do for the electonics we play with today somthing has to be compromised, and that is usualy the PSU, they are probably the most costly part of the system (if), so they provide the biggest cost savings. Hence we suffer, and argue endlessly about flawed electronics, such as battery supplies
It can be even more "fascinating" than that.
Years ago (in the end of eighties) I had to write custom BIOS parts for CP/M OS. Specifically. floppy drive IO parts. It was to be done in assembly.
I wrote the code and it worked. Kind of. That is, from time we observed seek errors.
Well, I though there was an error in my code. I very carefully examined it and couldn't find anything. I had two very smart colleagues, and I told them: "Guys, I tried my best, but I really can't find the problem. Maybe you'll have a look ?".
They did have a look and found nothing.
What was even more intriguing that inserting diagnostic print statements cured the problem. We all got really perplexed.
Then we took pulse counter and made it count pulses going to the FD head positioning stepping motor. Surprisingly, the number of pulses was correct, even when the OS/drive were reporting seek errors. I.e. there was no bug in my code.
Then one of the two HW guys took a big (like beer can) electrolytic capacitor and connected it to the PSU output feeding the stepping motor.
The problem disappeared.
I.e. the stepping motor was consuming too much current, there were short dips in the PSU voltage and the voltage was on occasions insufficient (during the dip) for the stepping motor to do its job, i.e. to reposition the head, hence seek errors.
And that explain the diagnostic print statement curing the problem - terminals were slow at the time, so the print statements were produced between the commands to the stepping motor, i.e. they increased time between the pulses, so the PSU could "regain" the "lost" voltage. PSU is a control system aiming to keep output voltage constant, buy its time constant is well above zero. The PSU was slow to respond in tthis case.
...
During my career as analog engineer I quickly learned that investment in power supplies is the first serious investment to be made. One of analog power supply I designed showed no ripple - just the noise.
...
Again, if compressed lossless format is likely to cause sound quality deterioration, the HW is simply crap, and it's the HW which should be fixed/replaced.
OTOH, hard core clueless audiophiles are a bonanza for sales/marketing folks - they sell them so much overpriced and useless stuff ...
ABX, DBT, ANOVA, T-test... whatever
You invoke statistical weakness, yet suffer from the very maladies of mis-interpretation you dismiss.
Statistical tests aren't only valid when you get the results you want. Sometimes the null IS the correct result..... even if you don't agree
I'll go away now...
You invoke statistical weakness, yet suffer from the very maladies of mis-interpretation you dismiss.
Statistical tests aren't only valid when you get the results you want. Sometimes the null IS the correct result..... even if you don't agree
I'll go away now...
When I get tired of my crappy PC lossless HW, I go put on a vinyl LP.
jeff
Just remember, vinyl records quite often had the hole not exactly in the center; 1mm displaced hole means 1mm/150mm = 1/150 = 0.67% linear speed variation, which is much worse than modern days jitter.
vinyl
right... snap crackle pop hiss rumble warp bass feedback skipping isolation mats and all... ahh the days of diskcleaner, record washers... obsessing over anti skate, counterweights... when did I ever have time to listen??....how I miss them
When I get tired of my crappy PC lossless HW, I go put on a vinyl LP.
jeff
right... snap crackle pop hiss rumble warp bass feedback skipping isolation mats and all... ahh the days of diskcleaner, record washers... obsessing over anti skate, counterweights... when did I ever have time to listen??....how I miss them
right... snap crackle pop hiss rumble warp bass feedback skipping isolation mats and all... ahh the days of diskcleaner, record washers... obsessing over anti skate, counterweights... when did I ever have time to listen??....how I miss them
If you have enough money and passion, buy an optical vinyl reader.
By the way, it could have been designed in the vinyl records glory days, but probably big business was interested in vinyl records wearing out, so people would buy new ones ...
ABX is a statistically weak test and cannot -- statistically -- prove that 2 DUT are the same. That is a fact, not an opinion. It is also very easy to confound the results -- either way, that is also fact.
dave
just wrong
doesn't claim to... that's the part you just don't get... anyone saying otherwise hasn't a clue
bad science is bad science... no argument there
doesn't claim to... that's the part you just don't get... anyone saying otherwise hasn't a clue
bad science is bad science... no argument there
vinyl
I've probably forgotten more about vinyl than you ever knew... Luddites rejoice
Your HW was crap. Sound familiar?
jeff
I've probably forgotten more about vinyl than you ever knew... Luddites rejoice
In what way is ABX "statistically weak" compared to any other controlled sensory evaluation?
I paraphrase a working statitician (retired now) for the Australian government pp it was his job to advise on the design of tests as well as run aome.
He throws around alphas & betas and ns. I used to know how all those were related, but it was so long since i got my Honurs stats degree that i am no longer a (theoretical) expert.
I think i can even dig out some SY quotes that support my points.
dave
He throws around alphas & betas and ns. I used to know how all those were related, but it was so long since i got my Honurs stats degree that i am no longer a (theoretical) expert.
I think i can even dig out some SY quotes that support my points.
dave
¡¡¿¿??!! Currents and Voltages. What else could it be?
Bit? What is a bit? There is in reality no bit. The zeros and ones do not exist in the real world. Bits, zeros or ones are only symbols with which we named certain forms in which modulate the flow of electrons (Electronics: Science that studies the electron). In other words, a form of modulate the power supply... as stated planet10.
The only thing that drives a DAC are electrons, as in any other electronic circuit. And in the real physical world can not build systems without loss (entropy). So I do not think there are 'lossless' formats, only formats with varying losses (more or less), and it is logical that each sounds somewhat different depending on how affecting the modulation of the electrons=power supply.
What is clear is that the measures in Audio are not able to correctly represent the reality, and that more research is needed in this area... but it is only a matter of time.
[One of Murphy's Laws: 'Do not confuse symbols with reality, do not eat the menu']
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.