I totally agree; Paul Frindle has a very open minded approach to the whole subject and a lot of experience with unexpected problems in the field.
The dither problem he talks about is a typical example of a real world problem that can arise.
Oh, i thought as a demonstration example it would be sufficient.
I assume that it would be possible to get some informations from Paul Frindle about the specific conditions, if you are really interested.
Wrt your discussion with jneutron, i´m a bit puzzled, because the minimum audible angle translates to ITDs of below ~20µs when tested with loudspeakers in front of the listeners.
And afair this was confirmed with diotic signals in lateralization experiments.
So are you asking for a scientific experiments in which this minimum ITD was confirmed while listening to music in a stereophonic loudspeaker setup?
Gee, what is the lowest level an audio A/D could measure? What voltage would be 1.58 exp 97 times that minimum? What spacing do you need on a connector for that voltage?
Resistors may be good to 180 db range, wire to 190 db, connectors in between, so how do you hook up a 32 bit device? It must have 194 db range. Oh yeah if 0 db is the perception of hearing at 194 db SPL how much of the atmosphere must we module to reproduce the peak level?
This is not true. You are spreading misinformation.
It is true that modulators in multibit delta sigma converters work at low bit and Mhz rates and then by decimation you obtain high wordlength low sampling rate but that not makes true what you say. In the audio band you can have 21 real bits.
Keith Johnson uses his own designs because fifteen years ago (PM1) and ten years ago (PM2) there was nothing of that quality available, and he and his team were smart enough to make such a state of art converter (PM2). Also Dan Lavry in the AD122-96 used discrete parts in a temperature controlled oven.
But at this moment there are a few converters that perform equally well or better than those.
I offered to many people who discredit digital audio to accept the A/B test of a high resolution analog source against level matched to 0.01dB double digital conversion (AD + DA) with a top mastering converter.
The double conversion was indistinguishable from the analog feed. All this talk about fuzzy distortion produced by dither is nonsense.
I also can say that just because you say it so it does not make it true.
You act like you know everything better than anyone else but after reading the last dozen of pages I think that many of your remarks about how digital audio work in theory and practice are wrong.
chrissugar
If you mean "interchannel time delay," then yes, that's exactly what I was asking for. Doesn't have to be music, test signals are fine as well.
Scott,
Oh, you mean the ones that no-one (with minor exceptions) has been making in the last decade or two?
Please show me ONE SINGLE pro-audio system that uses a SAR or FLash ADC and a Multibit DAC at at least 20Bit level manufacturerd since (say) Y2K...
Forgive me for being stubborn about this "reality" thing. I am naturally interested in what CAN be done, however when debating what actually happens I do insist on what actually is being done.
In this world there are many things that can be done, from permanent peace, the abolition of arms as well as distributing the food we already grow so no child needs to go hungry, the funding of educational systems so no child grows up ignorant and the implementation of a structure for society that offers a fair deal to all and sustainable development, I could go on.
All these are very nice ideas, but they are not what is the actual reality.
Well, it's my beef, got it?
Ciao T
Oh, you mean the ones that no-one (with minor exceptions) has been making in the last decade or two?
Please show me ONE SINGLE pro-audio system that uses a SAR or FLash ADC and a Multibit DAC at at least 20Bit level manufacturerd since (say) Y2K...
Forgive me for being stubborn about this "reality" thing. I am naturally interested in what CAN be done, however when debating what actually happens I do insist on what actually is being done.
In this world there are many things that can be done, from permanent peace, the abolition of arms as well as distributing the food we already grow so no child needs to go hungry, the funding of educational systems so no child grows up ignorant and the implementation of a structure for society that offers a fair deal to all and sustainable development, I could go on.
All these are very nice ideas, but they are not what is the actual reality.
Well, it's my beef, got it?
Ciao T
Hi,
Which mastering converter?
Nope, it does not, hence I suggested a few methods of observing said distortion.
If so, please illustrate where I fail to understand. But please make sure you are really certain about it. Do note come quoting folksy tales of magic dither or anything of the like, make sure you really are sure that what you point to is REAL.
The sticking point is that I have choosen to identify as distortion what in fact is A FORM of distortion, contrary to the folksy "magic" explanation that it somehow actually lowers distortion, when in fact it observably increases it, for two out of three samples, even though said "folksy magic" explanation is currently favoured in audio (well, they once favoured weapons salve too).
I do not think we per se disagree about the actual technical facts, where we part company is how we interpret things. I am not the least interrested in 64K samples averaged showing a low noise floor, as this low noisefloor is an illusion, what I am interested in is the fact that two in three or many more individual samples will simply be wrong, or distorted.
So to speak, from where you stand, I fail see the wood for the trees and from I stand, you fail to see the individual trees for the wood. I personally appreciate that things are never as black/white, however it often is necessary to go black/white to get the point across.
I note that still most people here cannot see the trees for the woods, pitty that, cause the devil is always in the detail, in the individual trees so to speak and no hand waving or building straw men will change that.
Ciao T
Which mastering converter?
Nope, it does not, hence I suggested a few methods of observing said distortion.
If so, please illustrate where I fail to understand. But please make sure you are really certain about it. Do note come quoting folksy tales of magic dither or anything of the like, make sure you really are sure that what you point to is REAL.
The sticking point is that I have choosen to identify as distortion what in fact is A FORM of distortion, contrary to the folksy "magic" explanation that it somehow actually lowers distortion, when in fact it observably increases it, for two out of three samples, even though said "folksy magic" explanation is currently favoured in audio (well, they once favoured weapons salve too).
I do not think we per se disagree about the actual technical facts, where we part company is how we interpret things. I am not the least interrested in 64K samples averaged showing a low noise floor, as this low noisefloor is an illusion, what I am interested in is the fact that two in three or many more individual samples will simply be wrong, or distorted.
So to speak, from where you stand, I fail see the wood for the trees and from I stand, you fail to see the individual trees for the wood. I personally appreciate that things are never as black/white, however it often is necessary to go black/white to get the point across.
I note that still most people here cannot see the trees for the woods, pitty that, cause the devil is always in the detail, in the individual trees so to speak and no hand waving or building straw men will change that.
Ciao T
Bob Adams gave me some advice at lunch, before entering into one of these discussions assess if any amount of evidence will change the others mind, if no walk away.
About the multibit converter stuff he just laughed.
Walking away and laughing seems like a good idea.
Last edited:
I note most people here are right on the mark, both from a theoretical and practical point of view. No amount of fuzzy logic or contorted hyperbole you have thrown into this discussion changes that one single lsb as far as I am concerned.
vac
Last edited:
Definitely, the best reasoning would be the captured transient, rather than walking and laughing. Everyone is able to walk and laugh
Or just let's show something like this:
Attachments
Last edited:
It is possible to build a 24 bit converter, but as far as I know no one does. 32 bits is not yet possible.
- Status
- Not open for further replies.