Childish personal attack and inability to form a reasonable cite noted.
What RNM never seems to be able to do is provide reliable evidence that what he is obsessing over today is actually audible. He seems to pick and choose, psychoacoustics be damned.
Right now he's giving a pass to audible distortion that is < 20 KHz and only a couple of dozen dB down. Why not give the ultrasonic stuff the benefit of comparable amounts of doubt?
Agenda's anybody? ;-)
Ok, I'll phrase it another way. You do appear to have turned up after a hiatus (possibly thrown off another forum) with what seems to be a single minded intention. Looking at your web history you do seem to follow a distinct pattern which is being repeated here. And it's a free forum so no one is stopping you.
Richard is pursuing an interesting line of reasoning regarding if incoming hash that exceeds the closed loop bandwith of an amplifier can cause issues. It may go no where but in the mean time a discussion on the different ways of dealing with ultrasonic and RF common mode and differential mode hash is both harmless and interesting and there is a slim chance that a correlation may be found between the fact that people who dislike filtering their inputs also seem to like ridiculous slew rates. Mean time no one is hurt and we all go home happy.
Richard is pursuing an interesting line of reasoning regarding if incoming hash that exceeds the closed loop bandwith of an amplifier can cause issues. It may go no where but in the mean time a discussion on the different ways of dealing with ultrasonic and RF common mode and differential mode hash is both harmless and interesting and there is a slim chance that a correlation may be found between the fact that people who dislike filtering their inputs also seem to like ridiculous slew rates. Mean time no one is hurt and we all go home happy.
You do seem to be irresistibly attracted to personal attacks, Bill.
Can you make a reply to me without including one?
Its almost like a... single minded intention. ;-)
The digital filter can have infinite stop band attenuation. It still has to go through an analog filter at some point.
The digital signals are not sine waves so it is not 5.6 volts RMS. However there are many if them as opposed to the two analog outputs.
The amount of EMI floating around inside a CD player is enormous. Easily picked up by most rf or EMI probes. Just the leads to the output jacks will pick up some.
Then there is the issue of setting up a straw man to knock down. I never stated the voltage level of the digital signals only the peak analog voltage. As the must be some headroom and dynamic range the output level for the analog may be treated as a .2 VRMS source.
The digital signals are not sine waves so it is not 5.6 volts RMS. However there are many if them as opposed to the two analog outputs.
The amount of EMI floating around inside a CD player is enormous. Easily picked up by most rf or EMI probes. Just the leads to the output jacks will pick up some.
Then there is the issue of setting up a straw man to knock down. I never stated the voltage level of the digital signals only the peak analog voltage. As the must be some headroom and dynamic range the output level for the analog may be treated as a .2 VRMS source.
Last edited:
The other reason and possibly one of the main ones is because everything is on one board and so in close proximity so the shielding is required for internal EMC/signal integrity control.. Similar layout technique's and shielding used where-ever you get a proper RF interface on a board. Other stand alone components are (or should be) shielded either by the enclosure or some other means, whether they are or not depends, a lot of DIY stuff even in metal boxes is not shielded correctly, its all down to slots and seams.
The analog filter is at a far higher frequency, well beyond the audible range, and far above the sampling frequency. In many NOS DACs it is omitted.
They are standard logic levels - typically either 3.6 or 5 volts nominal.
The relevant question is do they constitute an audible problem, and so far the evidence seems to be very hand-wavy.
Enormous is a dimensionless word. Enormous relevant to what?
If the CD is of modern hypercompressed music then it is sitting there banging against FS and its crest factor may be as little as 10 dB, which belies the 0.2 volt number provided.
Any Noise Annoys An Oyster....
I especially note/enjoy the practical and theoretical inputs from Ed Simon....kudos.
Kudos also to Mr Marsh for highlighting this subject......we have all heard Digital Source/Amplifier combos that sound right.....or not right.
IMO, non audio signal related noise is a/the major elephant in the room.
IMO/IME such noise can be a billion dB's down but it can still have a subjective influence....spectrum and dynamic spectrum of such noise is the key.
The nature of this noise/spectral behaviour can be altered/controlled at will
.
Dan.
Yes, I agree with you Bill, such trolling gets boring, really boring.
Yes, this is a properly interesting line of discussion, and present here are inputs from those well practiced in the art of containing/controlling/excluding the collateral affects/effects of RF/EMI.
I especially note/enjoy the practical and theoretical inputs from Ed Simon....kudos.
Kudos also to Mr Marsh for highlighting this subject......we have all heard Digital Source/Amplifier combos that sound right.....or not right.
IMO, non audio signal related noise is a/the major elephant in the room.
IMO/IME such noise can be a billion dB's down but it can still have a subjective influence....spectrum and dynamic spectrum of such noise is the key.
The nature of this noise/spectral behaviour can be altered/controlled at will
.Dan.
How does it work then, if it is "a billion dB's down", how does it turn to be audible?
Controlling EMI (EMC engineering) is well studied and documented, Henry Ott, Ralph Morrison even Keith Armstrong at EMC UK, dealt with every day in normal electronics so I would presume the same methods would work with Audio electronics... yet there always seems some magic when we come to audio and also often a rejection of accepted means of dealing with the subject (ferrite beads come to mind) or avoiding proven layout techniques such as ground planes for low level analogue...
Hi Pavel, of course I don't mean "a billion", but I do mean what is generally regarded as inaudible.
Static harmonic distortions are reported to be inaudible at levels of 60/80dB below fundamental....all good.
Intermodulation and non signal related products are distinctly more audible than this to the trained (and to the non trained, interestingly) ear.....is there a threshold level assigned to this ?......I think not.
I contend that very low level noise and especially it's spectral nature (internally generated and externally generated) and the consequent products are what differentiates good gear from bad gear.
Dan.
Last edited:
Only for you, Dan.
(Heloo Heloo, are you the big madonna with the fallen boobies ?)
Haha, Jacco you need to get out some more, go see a decent big live show....in the past month I have worked a few....Fleetwood Mac, AC/DC, Ed Sheeran, Sam Smith and a bunch more.
Live show 'perceived and subjective' kinda nails it.
.Regards, Dan.
Here's why Mr. Headroom you personally attack me for being a troll. My opinions are different from yours and you haven't yet learned how to deal with reasoned and substantiated disagreement.
I'm sure that some internet troll some place has said "Static harmonic distortions are reported to be inaudible at levels of 60/80dB below fundamental."
Of course that is far from anything like the complete story.
I invite one and all to try to google a formal definition for "static harmonic distortion". That will keep you chasing your tail for a while. It is not a term with a generally agreed upon meaning. So we start out with a statement that is based on terms that lack formal definition, and thus can mean whatever the person writing them wants them to mean. Thus the door is opened for claims that anybody who thinks differently has cognitive problems.
Audibility is a lot more complex than that. As close as the formal literature comes to setting a number for unconditional inaudibility seems to be around 100 dB. Another related number is the dynamic range of the human ear - IOW the smallest signal (noise, distortion, whatever) that can be heard under any condition while some loud sound is being listened to which is about 60 dB. Then there is masking which can be only a few dB, depending.
I didn't promise a rose garden, but if you know enough you can predict a lot.
I just wrote a discussion of the audibility of IM here:
http://www.diyaudio.com/forums/lounge/64538-resistor-sound-quality-184.html#post4550908
So I won't repeat it. The spurious responses created by nonlinear distortion includes those created by both THD and IM, and their audiblity depends on the signal that stimulates them, and the equipment nonlinarity that is stimulated, so they can and should be treated as being very similar.
Their threshold of audibility varies with their frequency and the signals around them that can mask them. It is knowable.
Noise can be analyzed into its components and the audibility of its components is generally far less than that for spurious responses caused by the music, because of their (noises) random nature.
Last edited:
...exactly as predicted!
One problem is the fact that perceived and subjective are words with more than one common definition, so mentioning them without further explanation is kinda short on meaning.
- Status
- Not open for further replies.