Jitter below 1 nanosecond is inaudible and not a problem with any of today's CD players/DACs. Regardless of what the "golden ear" crowd claims! And asynchronous DACs (with their re-clocking of the input signal) eliminate jitter (uncertainty of the bit-timing) entirely! Let the rock throwing begin!
DAC's : The missing feature: Signal quality
One thing I wish DAC's would provide is some idea of how much jitter and noise a particular input provides. This is something which I think with a little work could be gleaned from the input circuits.
I want something that tells me "woah, that's a really dirty signal coming in, but i"ll do the best I can with it."
One common source of noise is ground loops. Another may be high jitter from a source like Apple TV. This would also help us evaluate the benefits (if any) of various signal cleaners and reclockers.
Best,
E
I want something that tells me "woah, that's a really dirty signal coming in, but i"ll do the best I can with it."
One common source of noise is ground loops. Another may be high jitter from a source like Apple TV. This would also help us evaluate the benefits (if any) of various signal cleaners and reclockers.
Best,
E
41 responses Add your response
Even the first-generation CD players had jitter levels below 1 nanosecond! What they didn't have (in the first year or two) was full 16-bit capability. Typically 14-15 bits, due to DAC-chip (ladder-type) architecture. This only reduced noise-floor and dynamic range by about 10db! Still far better than any tape/vinyl playback! L-R separation for stereo imaging was unaffected - again far better than any tape/vinyl system! Ditto for harmonic/IM distortion! |
The reason this is not common as it’s not really needed. It’s cool for curiosity or scientific purposes (instead of having to buy an analyzation tool/software), but no real benefit for consumers. As doing a J-Test (extreme jitter, more than you’d likely ever have), even cheap DACs can reduce down to at least -100dBFS, better than the noise floor of a CD, and since almost no content is gonna have sounds meant to be heard below say 40dB, even a cheap DAC would have no jitter when playing 40dB to 140dB. |
Erik, if I correctly understand your desires for signal quality detection at DAC inputs, then I suspect it would be an extremely costly DAC having that attribute. It would likely be on par of an external reclocking device cost. I’m not sure you want to spend that kind of $$$ for a feature that is nothing more than an idiot light. |
The reason this is not common as it’s not really needed. No one has any idea if this is needed or not because end users never get to measure it themselves. The argument here is really that unless it's blatantly obvious to the ear we should not care. To those people I ask: Do you even buy high end equipment? I've read a number of reviews, and experienced enough variance in performance of sources to know that this is not the case. Having some sort of Green/yellow/red indicator about signal quality and noise I think would be helpful. For instance, there are a number of reclockers and signal cleaners on the market. Wouldn't it be nice if our DACs could indicate directly if a particular source might benefit from it? Like, green on CD player, green on streamer, but woah, that AppleTV signal is a mess. So pick your place. Bits are bits and unless the signal drops out I don't care, or I care about signal quality and would like to know if I'm missing something important. Best, E |
If your system has jitter (only thing to take note of for digital connections; linearity, THD, crosstalk, IMD, SNR, etc. are all near identical), then that results in a rise in the noise floor, which would sound like static, it would be heavily masked by music, but if you played say a 10kHz tone on your digital device, and you heard some background static, then you have audible jitter. We are talking random jitter here (not periodic), and if you want an example of what poor random jitter sounds like as a reference point: http://www.sereneaudio.com/blog/what-does-jitter-sound-like Take note of how 2ns of random jitter raises the noise floor to ~ -80dBFS, and keep in mind how I said even cheap DACs can reduce it to -100dBFS or lower (+20dB is the same as having 100x more wattage, so in this case 1/100 the wattage as it’s the opposite direction). |
Ugh. Julian, wouldn’t it be easier to simply provide folks here with a link to your pontifications? smh Hell, I’ll do it for you to spare the suspense for everyone concerned: https://forum.audiogon.com/discussions/why-does-my-dac-sound-so-much-better-after-upgrading-digital-... |
| Post removed |
Look at it another way. We spend a lot of tiem money and effort looking for things to fix. Wouldn't it be good to have a dac that turns red if a particular source is no longer performing well? I mean, if we hear something change, we often look for cables, power conditioners, or even room acoustics (which are always important!) I'd love to have a little light saying "Man, it is really hard to lock onto this device now." Best, E |
@celander Exactly. To answer the question about what we would use externally, jitter, noise and distortion measurements. I don't think we need to have all that test gear in place though. For instance, there's circuitry which is making choices about adjusting the DAC to the incoming signal. Knowing some basic statistics about that. Like, what is the incoming long term variance? How often are there outliers? The complicated view techs might see could be reduced in my mind. I know at least one DAC even claims to compensate for impedance mismatches, so this must be measurable. Best, E |
Erik, I presume one could adjust the sensitivity of the LED indicator (or use a multi-colored LED) that reports distinct levels of jitter. Seems like one would need some sort of jitter level sensor on the output digital stream just prior to final analog conversion. That sensor could be calibrated to feed a signal to the LED to indicate jitter level. Not convinced that the DAC jitter reduction would not overwhelm whatever jitter level you’re trying to detect in the input cable. But presumably one could put a similar sensor on the input side of the DAC input circuit. |
Erik: I always like that you are thoughtful about our hobby and I conceptually like your idea. If a DAC sounded exceptional to me, and it had this feature, and it could be demonstrated to make a difference by allowing the user to rationally react to it, then I would consider it a value add that might affect my purchasing decision. But, it probably wouldn't be a major reason to buy a DAC (for me) and might not survive a company's internal cost benefit analysis, if I am DAC manufacturers' target audience (I'm a 50 something male passionate about audio that owns or has owned more than 6 DAC or CD players in the last 5 years.) As a former R&D guy in an unrelated field, I've seen forwarding thinking ideas die many times because a company didn't think there was an increase in profitability attached to the concept. Regards Al |
Jitter detection and then committing to mathematical transforms of measurement is a relatively complicated affair. Last time I did that, it required the use of an AP SYS2722 measurement system, in a fit of clock designing... $20k or thereabouts. https://www.ap.com/analyzers-accessories/2700-series/ |
Isn't Schiit's implementation of the "buy better gear" LED something like what you're asking for? It's on Gumby and Yggy, not sure about other DACs from them. I believe it has something to do with not being able to reliably lock onto the clock from the source to let AdaptiClock do its thing. To wit, I have an Yggy "B" bought recently and from my CD player via AES/EBU, it has never lit up. But from my PC, via 3M coax digital, it has lit up once that I recall, on a Redbook FLAC album. Doesn't surprise me at all, the sound card in the PC was just a cheapy to get a coax S/PDIF out which unfortunately isn't all that common and optical of that size run is going to be sketchy at best at 24/96 or above. But anywho, I think that's at least a rough approximation of what you're thinking about. It doesn't display any sort of "this is how far off you are" info, just a light on the front of the unit. |
Depending on your budget Ifi Audio has a Usb -3 purifier$129 a steal, that removes 40 dB of noise , plus regenerates a totally new signal parts qualityexcellent , they also have productcalled the defender buy 1-2 removes crap associated with computer noise $59 computer switching tons of High frequencygsrbage both come with full money back satisfaction , also there AC purifiers at $99 I have 3 i have great isolation dedicated lines, Copper audio outlets good line conditioning and these AC purifiers are cumulative. Plugin line conditioner , and AC outlets . Many people usethem Abbington music - UK is the mother company. they have a headquarters in theUS. No risk ,Results or full money back . I am a Amazon prime member shipping is -0 even on returns something to consider if buying . Go to IFi audio cite best bang for the $$ out there IMO. |
audioman - The issue there is the same. I’ve read technical reviews on iFi purifiers that show them to be a really mixed bag. They don’t do enough to control power supply noise, and end up just moving noise from one area to another. See if you can find it on google. Lots of charts and measurements in it. The problem for us consumers is if there is an audible change, we don't really know which is better. What if the purifier, or digital cable is actually awful, but we buy it because it sounds different? I'd rather have some way, built into my DAC, that tells me. Like signal strength meters or tuning centering meters on old FM tuners. Best, Erik |
I just returned a very highly regarded high-end low noise DAC. It has some of the best specs in the industry, lowest distortion, lowest jitter, super quiet dark background. It produced super clean, open, two dimensional "sound". Not "MUSIC". While the digital designers were seeking absolute perfection, they also managed to completely eliminate the three dimensional and sweet sound of music right out of the equation. Could not wait to get back to my older distorted imperfect DAC. Once I did, real music returned. Seems perfect digital specs on paper don’t always make for good music and maybe some of the good stuff gets filtered out of the "signal quality" to achieve sound perfection, not musicality. |
Having some sort of Green/yellow/red indicator about signal quality and noise I think would be helpful. For instance, there are a number of reclockers and signal cleaners on the market. Wouldn't it be nice if our DACs could indicate directly if a particular source might benefit from it? That would be really nice. Unfortuntately, the industry does not even have any metrics for jitter versus audible correlation yet. I'm working on this actually. Equipment that can actually measure direct jitter accurately is very expensive, on the order of $100K. Steve N. Empirical Audio |
I just returned a very highly regarded high-end low noise DAC. It has some of the best specs in the industry, lowest distortion, lowest jitter, super quiet dark background. It produced super clean, open, two dimensional "sound". Not "MUSIC". Based on 22 years of modding other companies DACs and designing my own, I have concluded that certain things compromise SQ in digital audio. Here is my prioritized list: 1) Jitter 2) Digital Filtering 3) Poor power regulation and decoupling in both digital and analog sections 4) Compression from too many stages or poorly designed OP-Amp stages or both 5) Poor ground-plane design 6) Long traces on circuit boards, over 1" 5) Insufficient drive from the output stages Steve N. Empirical Audio |
I have 2 iFi SPDIF iPurifiers in my home theater system to reduce jitter. Makes big difference for movies. For 2-channel, my Synchro-Mesh reduces jitter more. Steve N. Empirical Audio |
There is no way this was a jitter detector. This was a classic S/PDIF receiver error detector. Even my DAC has this. Most logic families used in DACs are too slow to make these type of determinations. We are talking about gates with propagation delays of 1nsec trying to detect a jitter of 10psec. Not going to happen. Steve N. Empirical Audio |
erik - Purifier?? I don't know what product this is. Do you mean the Short-Block? There is no incompatibility with XMOS-based USB interfaces. Sometimes they need +5V from the USB cable depending on the design. The SB cuts this voltage. It doesn't matter because it out of production now. I would highly recommend this product for USB: https://sotm-usa.com/collections/sotm-ultra/products/copy-of-tx-usbultra-regenerator-1 This is the ONLY thing I have tried that makes my USB almost as good as my Ethernet interfaces. Steve N. Empirical Audio |
audioengr3,129 posts11-24-2018 3:20pm Steve, nice list in this post. With respect to the EA SM reclocker, I can see where your PS upgrade fits on the list. Where does the OTL upgrade fit, under jitter? |
The OTL upgrade simply removes the output isolation transformer. If the DAC has a transformer on the input, there is no reason to have two in series. The pulse transformer is really not necessary, but helps with ground-loop noise in some systems where the DAC does not have an input transformer. Both the standard and OTL versions of SM reclocker have very low jitter, but the OTL is slightly lower. If you want the lowest jitter, then OTL is recommended. Plugging the AC power for the DAC and the SM into the same outlet usually minimizes ground-loop noise, so the OTL works fine. Steve N. Empirical Audio |
