Hmm, a change brought about by rejection of belief and use of the scientific method. Not the point you wanted to make I suspect...
what I wanna tell is already posted in early reply.
first of all, digital circuit is affected by power supply, (clock always show as like-sine wave, not perfect square wave,that makes trigger voltage level more important than just pure digital circuit) and latency always exists in digital circuit, then comes the jitter, jitter does not interfere with output sound quality directly, but jitter converts to phase noise that makes sense. jitter could be stopped by FIFO/buffers, but still it could affect post circuits, that matters.
because the limitation of human's ears not known, still there is no completed model to setup a full covered environment for Audio playback system, nor standard for Audio playback system, that comes this argument.
I don't wanna waste my time explain this cause there are really some recording engineer could tell the difference between what I describe early in this subject.
please challenge this, that's not my report:
Audio Networking: A potpourri of computer network audio findings, including updates on the Naim UnitiServe, gigabit switches, CAT 5/6 cable and ripping issues. Review By Martin Colloms
I wonder how we ever get a DDR memory interface to work, and if you are curious it is the most complex part of a PC for timing budgets, length matching and layout, each one should be simulated, do these regular, still dont understand what you are trying to say.
Will digest the link, as Martin Colloms is the auther I am in for a treat
I wonder how we ever get a DDR memory interface to work, and if you are curious it is the most complex part of a PC for timing budgets, length matching and layout, each one should be simulated, do these regular, still dont understand what you are trying to say.
Will digest the link, as Martin Colloms is the auther I am in for a treat
try read more for operating system and MMU system in CPU before you could fully understand how jitter comes.
length matching is to minimize the difference cause by data/address bus length. I'm not a PCB-layout engineer, but no documents require exact length matching for DDR bus, just like +/-2mm something according bus speed.
digita will have tolerance to errors, not like analog. for databus like they are saying 1.5G, uses 128bit, its bus speed is like 1.5G / (128bit /32bit), should aroud 300-400M, that's the trick in PC industries, in fact, when you are copying data continuously, it could approach what the advertisement annouced, for databus @ 400M, 1/10 jitter could be acceptable, like 1/4G=25ps, converted to voltage ripple should less then 1/10, that's why operating voltage is about TARGET_OPERATING_VOLTAGE +/- 10%, that's for tolerance consideration between different chips, for single chip system, it works might be +/- 20% or more, agree?
better power supply, lesser jitter, that's he result in digital circuit, agree ?
Last edited:
Not realy, PC's work and getting hung up over jitter in the subsystems is pointless, it is engineered to work within the system defined jitter levels, thats why we use simulation.
No such thing as raw battery power for a PC, it is converted through SMPS,s and LDO to the required supply requirements for the various devices and the requirements of their interfaces. Also as well as supply noise, you have switching noise (SSN) wich is probably greater, again digital is tolerant to this noise and again during the development of a PC these things are checked by simulation and testing. Power supplies are engineered to provide the correct voltages within tolerances required, it not majic is engineering and all but the most basic toy PC's are up to the job and work reliably day in day out, as does all digital circuitry, with a good reliability record, it is rare for kit to fail these days if well made.
The DDR reference was to illustrate how we can engineer high speed interfaces that are tolerant of jitter, it is an integral part of any PC and critical to the PC running.
As to the power delivery stystem, that is also simulated these days and a lot of work is done to ensure the integrity of the power supply so that the PC, or whatever equipement works. This all relies on believing bits are bits and trusting in data integrity dosn't it, which of course guru's like Martin Coloks are telling us different, because the golden hearing dont need network anylisers to tell us whats happening to the digital signal as there is this magic noise that can be transmitted through buffers and even recorded with the data that we dont know how to measure yet.
As to the rest no doubt you are going to promote specialist digital cables that enhance the sound, as well as a audiophile music server, etc etc.
No such thing as raw battery power for a PC, it is converted through SMPS,s and LDO to the required supply requirements for the various devices and the requirements of their interfaces. Also as well as supply noise, you have switching noise (SSN) wich is probably greater, again digital is tolerant to this noise and again during the development of a PC these things are checked by simulation and testing. Power supplies are engineered to provide the correct voltages within tolerances required, it not majic is engineering and all but the most basic toy PC's are up to the job and work reliably day in day out, as does all digital circuitry, with a good reliability record, it is rare for kit to fail these days if well made.
The DDR reference was to illustrate how we can engineer high speed interfaces that are tolerant of jitter, it is an integral part of any PC and critical to the PC running.
As to the power delivery stystem, that is also simulated these days and a lot of work is done to ensure the integrity of the power supply so that the PC, or whatever equipement works. This all relies on believing bits are bits and trusting in data integrity dosn't it, which of course guru's like Martin Coloks are telling us different, because the golden hearing dont need network anylisers to tell us whats happening to the digital signal as there is this magic noise that can be transmitted through buffers and even recorded with the data that we dont know how to measure yet.
As to the rest no doubt you are going to promote specialist digital cables that enhance the sound, as well as a audiophile music server, etc etc.
Last edited:
you really have no idea what you are talking about do you? who is feeding you all this crap? just isolate and reclock at last moment before DAC, thats all there is to it, it really DOES render all this talk meaningless jibber (if it wasnt already)
still we await this earth-shattering discovery of the half-bit...sounds like?
clock can look however you like Sine, 'square' whatever, if the decision points are able to be distinguished and are regular, it matters not and the jitter of PC is meaningless, it has no effect AT ALL on a properly designed DAC and interface, which is where these jobs should be done to begin with.
jitter has been shown reduced from 2ns to 2ps and is the same 2ps as if fed by low jitter source. all this fretting and creation of a new audiophile server industry for people with no knowledge of what mysteries go on inside the box will die before it gets going in earnest I hope.
Last edited:
OK, so that's a long way of saying, no, you don't have any data and can't cite any, and that you're unaware of anyone who has done an "ears only" test.
I found this statement in the literature for the Rega DAC interesting:
"Great care has been taken to remove noise generated by the PC and other input sources. During development this was identified as a major drawback with many DAC's on the market today."
They don't elucidate where this is coming from. Rega is a respected old school company. I have their turntable and their tonearm with a Decca cartridge running fine on it which is perhaps the most difficult suspension in a cartridge, but I digress. It's just that they have engineers who know analog and aren't snake oil people, I hope.
Maybe they're just talking about USB, which is kinda crappy anyway. Still, it's so simple to filter out or block low level electricity from a digital line from ever getting to the analog side in a DAC and so simple to correct timing irregularities, so simple that effort toward better sound would be more fruitful if directed at the speaker crossover.
I don't think that most self-deluded audiophiles are stupid or craven. They're wrong. No sense impugning their character. The placebo effect is so powerful that any medical study without it is generally unpublishable. Boils appear on hypnotized people after being told they put their hand on a hot stove. Eastern thought goes so far as to say that all of perceived reality is Maya, an illusion. Christian Science believes that all is mind, there is nothing else. It is impossible to overestimate the impact of belief on perception. Dwell on this phrase: Scary Movie. Objectively impossible. There is nothing to fear about going to see a movie as long as it isn't a Dark Knight premiere or your fellow movie goers are Chechens. Your imagination is stimulated by the sight and sound of the movie, and the emotional centers of the brain can't discount imagination. It thinks it's real.
People who must know better and continue to sell gold CDs by malarkey aren't so sympathetic.
Mere observation would tell you the Sun revolves around the Earth and that the heart is the seat of the emotions. Ben Franklin thought the brain was the body's air conditioner. Wrong. Not stupid, but still wrong.
"Great care has been taken to remove noise generated by the PC and other input sources. During development this was identified as a major drawback with many DAC's on the market today."
They don't elucidate where this is coming from. Rega is a respected old school company. I have their turntable and their tonearm with a Decca cartridge running fine on it which is perhaps the most difficult suspension in a cartridge, but I digress. It's just that they have engineers who know analog and aren't snake oil people, I hope.
Maybe they're just talking about USB, which is kinda crappy anyway. Still, it's so simple to filter out or block low level electricity from a digital line from ever getting to the analog side in a DAC and so simple to correct timing irregularities, so simple that effort toward better sound would be more fruitful if directed at the speaker crossover.
I don't think that most self-deluded audiophiles are stupid or craven. They're wrong. No sense impugning their character. The placebo effect is so powerful that any medical study without it is generally unpublishable. Boils appear on hypnotized people after being told they put their hand on a hot stove. Eastern thought goes so far as to say that all of perceived reality is Maya, an illusion. Christian Science believes that all is mind, there is nothing else. It is impossible to overestimate the impact of belief on perception. Dwell on this phrase: Scary Movie. Objectively impossible. There is nothing to fear about going to see a movie as long as it isn't a Dark Knight premiere or your fellow movie goers are Chechens. Your imagination is stimulated by the sight and sound of the movie, and the emotional centers of the brain can't discount imagination. It thinks it's real.
People who must know better and continue to sell gold CDs by malarkey aren't so sympathetic.
Mere observation would tell you the Sun revolves around the Earth and that the heart is the seat of the emotions. Ben Franklin thought the brain was the body's air conditioner. Wrong. Not stupid, but still wrong.
Last edited:
I don't think bit perfect system has no jitter neither.
I don't think cable will affect playback quality while they have the same capacitance and resistance.
Jitter is only a concern at the real time playback stage...the rest of the digital system is designed to allow for jitter. Bit perfect is how data is transferred around PC's The internet etc etc etc, all digita transfer systems will have jitter, but it is part of the equation when desining and building a sytem, eye diagrams are one way for checking dtat throughput in regard to jitter, it is a non problem because the solution is engineered in, so the user dosn't have to worry about it within the PC, at the final stage of the chain the DAC, then Iif it is a ceoncern there are numerous well engineered solutions, Iancanadas is one that springs to mind.
jitter is one of the thing that matters. FIFO could make the sound quality better, but, not eliminte jitter and phase noise.
correct grounding makes sense, you knew it well cause that's your major. Ian's solution is one of the solutions that could work, but our hardware structure could do better(not the product I show in this topic), that's another way for digital isolation and capacitance re-clock without PLL. cause we don't like the sound using PLL clock, not clear enough.
---let me tell you something that you could not agree, in the very beginning, I didn't neither.
when DIYers telling me that modification of digital power supply's capacitor makes sound different, I was at your side, thinking tha's not realy years ago, but, when why they show me the truth, I was shocked, and comparing with prism sound ADA8-XR, our machine sounds like a low end MP3 comparing with digital clear CD player, that makes us try to setup the model.
we found out that, 300uV rms power supply makes sound not clear as 100uV RMS power supply. that's modificatio to digital circuit, not analog !
for ES9018 DAC, you must make sure that the reference voltage noise under 5uV RMS or the performance THD+n could not reach -120dB ...
Last edited:
The master clock should be where the DAC is, like it has been done
with AES/EBU for ages.
Seems like you have no idea...
Wow you sure know how to screw up a post quote....
I dont bother replying in detail to nonsense; besides you seem to have some sort of anti-logic filter to have gotten this far. no PLL, no problem... fretting about pointless or non-existent issues and 'solutions' that still leave the most important point of the chain exposed, is not something I want, or need to 'learn' about. the best solution to the dacs problems, also solves all the problems before that.
A large isolated buffer (actually isolation after the buffer is most important IMO), clean power, synchronous output on impedance controlled µBNC headers. it works incredibly well, it deletes any amount of jitter anyone is likely to have, the only jitter that matters is the DAC master clock jitter and that of the flipflop. thats it, no amount of mumbo jumbo will touch that level of performance.
Ian was not the first to do such a thing, just the first to offer it to the DIY community in such a clean way, i've been using it for ~18months now (actually i've lost count, more than 12), I was a beta tester, the u.fl connectors are the only thing i'm responsible for, Ian is a machine!
on that note, not sure if anyone should give you any more help to understand, since you are gathering information for a product that 'solves all this' right?
I dont bother replying in detail to nonsense; besides you seem to have some sort of anti-logic filter to have gotten this far. no PLL, no problem... fretting about pointless or non-existent issues and 'solutions' that still leave the most important point of the chain exposed, is not something I want, or need to 'learn' about. the best solution to the dacs problems, also solves all the problems before that.
haha, so you still havent understood what weve been talking about? really?
see above, actually they did a pretty good job, I use their chip, but I prefer synchronous clocking with it. for a Dac chip by itself its about as close to a full solution as we could hope for, but the full solution doesnt exist in one chip, ESS (in async mode) is most effected by EMC on ground and jitter on BCK (not always adressed on USB->i2s convertors very well), besides what specifically are you talking about, you have some study?
A large isolated buffer (actually isolation after the buffer is most important IMO), clean power, synchronous output on impedance controlled µBNC headers. it works incredibly well, it deletes any amount of jitter anyone is likely to have, the only jitter that matters is the DAC master clock jitter and that of the flipflop. thats it, no amount of mumbo jumbo will touch that level of performance.
Ian was not the first to do such a thing, just the first to offer it to the DIY community in such a clean way, i've been using it for ~18months now (actually i've lost count, more than 12), I was a beta tester, the u.fl connectors are the only thing i'm responsible for, Ian is a machine!
on that note, not sure if anyone should give you any more help to understand, since you are gathering information for a product that 'solves all this' right?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.