Charles,
Could you please point out where I misquoted you (and maybe even answer my questions)?
Could you please point out where I misquoted you (and maybe even answer my questions)?
you also claim its not using oversampling... how does that work with your 32bit volume control then? or you are talking legally about the dac chip not being oversampling, but its OK for the volume control to use oversampling?
so what about these people that are buying it to use with in ear monitors, ive seen quite a few of them, do you tell them that they will be operating at 6-8 bits at low volume? ive mentioned that quite harmlessly in the threads previously trying to let people make informed decisions, yet people always took me as a detractor or didnt believe me, when its just simple fact with the way the dac is set up. they cannot have been very well informed of these pitfalls before purchase.
yet with all of this you claim superiority and charge more than the most expensive players for performance significantly lower than a version 1 ipod. using digital volume with a 16bit dac, not inspired. it really would have been cool object, I would have thought it cool if you werent making all these claims of superiority, even in the same sentence as saying its operating at 4 bits
so what about these people that are buying it to use with in ear monitors, ive seen quite a few of them, do you tell them that they will be operating at 6-8 bits at low volume? ive mentioned that quite harmlessly in the threads previously trying to let people make informed decisions, yet people always took me as a detractor or didnt believe me, when its just simple fact with the way the dac is set up. they cannot have been very well informed of these pitfalls before purchase.
yet with all of this you claim superiority and charge more than the most expensive players for performance significantly lower than a version 1 ipod. using digital volume with a 16bit dac, not inspired. it really would have been cool object, I would have thought it cool if you werent making all these claims of superiority, even in the same sentence as saying its operating at 4 bits
Last edited:
The way I read this part of the posting by Charles:
it seems the volume control is just a scaling operation on the 16-bit data, but with the calculation done in 32 bit. So just what everybody else would call a 16-bit volume control.
One pointless post an absurd thread doth not make. Please keep this on-topic Julf - cross examination of Charles on aspects of sigma-delta theory most certainly is OT. Start another thread if you really want to engage in a willy-wanging contest.
Why would a volume control use oversampling if the DAC doesn't?
it was a question, when someone calls something a 32bit volume control (which being cortex is possibly running on a non audio xtal frequency), I would normally associate that with spinning it out to 32bits for higher precision and then spitting it out into 16bits, but if its still just running at 16bits fullscale but just allocating 32bits for the calculation, then thats another thing. my apologies.... its a minor point
but I agree and i'm happy to just leave it, its clear what the strategy is, to baffle and confuse those not able to read into the carefully worded detail. no amount of argument will change that.
Last edited:
Ah, so him jumping in out of the blue to make claims is OK, me questioning them is OT. I see. Sure, roger willco.
Me thinks the emperor will be very cold this winter.
Ha, my experience is mainly pratical, I spen all day every day designing PCB's, and have done for over 27 years. No I am not soley responsible for EMC because we work in teams, all providing their expertease in thier own area, but when it comes to EMC problems we all work at them as a team, as any explanation or idea can help solve what are usually a pain of a problem. I study Henry Otts stuff as well as others, Dr Eric Bogatin, Dr Howard Johnson, Keith Armstrong and many other sources of information to do with the complexity of getting a signal (be it analogue or digital) from source to desitination, and to stop said signals causing problems to other signals or broadcasting as EMI. That is the reason why I do understand the comlexities of laying out PCB's for mixed signal, high speed, analogue, power electronics and SMPS (I specialise in laying out high reliability SMPS's for critical systems), because I have done it for many years.
Hopefully the links I have posted will explain why an unbroken contigous ground plane is so important to digital signals (and also to analogue), as you have to consider both the signal and its return path as one entity, failure to do so will compramise your signal, and I do consider that very important especially where mixed signal design is concerned. Digital only is quite tolerant as you are in the digital domain so as long as your switching points are OK the data will get through. Mixed signal and grounds are more critical as your analogue signal is referenced to that ground (arbitry 0Vs), eben moderate digital switching can cause differentials in the ground, especially with a high impedance ground (star ground anyone), this can effect the low level analogue levels and in the case of audio loss of fidelity, in other analogue based systems it can be catastrophic. This is not theory, this is from practical knowledge and real life problems I have seen. So comments that I make are to give what little information and help I can regarding the subject matter, I am not as hot on the maths and theory of basic electronics as I use to be (when I left colledge, amny moons ago) as said I have concentrated on PCB designs since I got volunteered to become the PCB designer at my first job (my theory these days is as good as my dyslexic typing) but when it comes to PCB layout I do now my stuff.
Again with multiple devices clock distribution is critical, and for somthing as complex as this we would use simulation to both check the layout and determine any and what sort of termination to use.
Have fun ( I think you are)
Marc
Hopefully the links I have posted will explain why an unbroken contigous ground plane is so important to digital signals (and also to analogue), as you have to consider both the signal and its return path as one entity, failure to do so will compramise your signal, and I do consider that very important especially where mixed signal design is concerned. Digital only is quite tolerant as you are in the digital domain so as long as your switching points are OK the data will get through. Mixed signal and grounds are more critical as your analogue signal is referenced to that ground (arbitry 0Vs), eben moderate digital switching can cause differentials in the ground, especially with a high impedance ground (star ground anyone
), this can effect the low level analogue levels and in the case of audio loss of fidelity, in other analogue based systems it can be catastrophic. This is not theory, this is from practical knowledge and real life problems I have seen. So comments that I make are to give what little information and help I can regarding the subject matter, I am not as hot on the maths and theory of basic electronics as I use to be (when I left colledge, amny moons ago) as said I have concentrated on PCB designs since I got volunteered to become the PCB designer at my first job (my theory these days is as good as my dyslexic typing) but when it comes to PCB layout I do now my stuff.Again with multiple devices clock distribution is critical, and for somthing as complex as this we would use simulation to both check the layout and determine any and what sort of termination to use.
Have fun ( I think you are)
Marc
You really are getting the hang of those strawmen now
Not my meaning, what I meant was :Do you know how delta-sigma actually works? Have you ever heard of pulse density modulation?
That's willy-wanging.
Clearer now?
but I thought it was R2R?
though it could of course be some unknown industrial R2R 16bit dac... I suppose? hard to tell with so little detail
What was R2R? Charles thinks his DAC's R2R but he's mistaken on that because Philips use a neat little trick called 'Continuous Calibration' to set the weights of individual current sources. This isn't compatible with a resistor-based ladder architecture.
The industrial ladder DAC isn't unknown any more, though I think Cees still scratches off the markings. Its voltage out, and Charles' is current-out, as is mine.
The industrial ladder DAC isn't unknown any more, though I think Cees still scratches off the markings. Its voltage out, and Charles' is current-out, as is mine.
Last edited:
There is no oversampling taking place in the Tera-Player.
Every digital volume control decreases volume by calculation and the lower the volume, the fewer the bits that contain the remaining signal. Full resolution is obtained at full volume.
The digital volume control in the Tera-Player works and sounds very good, else I would have chosen different means for volume control.
It is not my fault that sigma-delta DACs don't sound good, I would be happy if they would. They may be considered a good idea that just turned out bad. And now a whole industry with bad ears is creating stuff for people with bad ears. Please take this as my very subjective experience of the matter.
IMO, when quoting the true resolution of a sigma-delta DAC, always the resolution of the output stage should be quoted, ie. 1 bit, 2 bit or maybe up to 6 bit in case of the ESS Sabre DAC (which are damn many bits for a sigma-delta DAC).
Speaking of 24-bits or even 32-bits with sigma-delta DACs some other term should be used, here are some suggestions: Noise-o-lution, Fake-o-lution, Dream-o-lution, Mareting-o-lution, Non-Timbre-o-lution, Headache-o-lution, Digititis-o-lution, Artific-o-lution, not-suited-for-music-o-lution ... something like that ...
Charles
Every digital volume control decreases volume by calculation and the lower the volume, the fewer the bits that contain the remaining signal. Full resolution is obtained at full volume.
The digital volume control in the Tera-Player works and sounds very good, else I would have chosen different means for volume control.
It is not my fault that sigma-delta DACs don't sound good, I would be happy if they would. They may be considered a good idea that just turned out bad. And now a whole industry with bad ears is creating stuff for people with bad ears. Please take this as my very subjective experience of the matter.
IMO, when quoting the true resolution of a sigma-delta DAC, always the resolution of the output stage should be quoted, ie. 1 bit, 2 bit or maybe up to 6 bit in case of the ESS Sabre DAC (which are damn many bits for a sigma-delta DAC).
Speaking of 24-bits or even 32-bits with sigma-delta DACs some other term should be used, here are some suggestions: Noise-o-lution, Fake-o-lution, Dream-o-lution, Mareting-o-lution, Non-Timbre-o-lution, Headache-o-lution, Digititis-o-lution, Artific-o-lution, not-suited-for-music-o-lution ... something like that ...
Charles
You really are getting the hang of those strawmen now
Do you know how delta-sigma actually works? Have you ever heard of pulse density modulation?
That's willy-wanging.
Clearer now?
Not really. If you look at the latest posting by Charles, the one that states "IMO, when quoting the true resolution of a sigma-delta DAC, always the resolution of the output stage should be quoted, ie. 1 bit, 2 bit or maybe up to 6 bit in case of the ESS Sabre DAC (which are damn many bits for a sigma-delta DAC)", do you still think my (serious) question is not justified?
And what is the full resolution of the volume control in the Tera-Player?
But do you agree that independent of the output stage of a delta-sigma DAC being 1, 2, 6 or 27 bits, even a 1-bit delta-sgma DAC can achieve a resolution that is more than 16 bits (resolution here defined as "with an ability to resolve amplitude differences smaller than 1/2^16 of full scale)?
Yes, really.
Irrelevant. If you want to explain something to Charles that he's misunderstood, that wouldn't be OT. But willy-wanging remains willy-wanging.
No, not really.
Irrelevant, according to you.
It is willy-wanging if you say it is willy-wanging. But I feel it is totally valid and called for to ask, in all seriousness, based on his statements, if Charles actually understands how delta-sigma and pulse-density modulation works.
You might disagree, but this is a public forum. If you have to resort to pejorative terms like "willy-wanging", that is your choice.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.