DAC measurements explained?


Hi,

Would some kind person explain what “DAC measurements” mean?

How are they measured? What are they measuring?

Why are they interesting/important to look at?

How do they relate to the sound and what do they tell you about it?

Thanks!
leemaze
Probably a good place to start is in the Stereophile reviews. Lots of DAC's measured there and free to view.

But, just in general, you have 0s and 1s in a CD, or file, or stream. Measuring a DAC measures how closely to an idea reproduction of the music the DAC can come to. Bits in, analog out.

Noise, jitter, linearity and frequency response are the big ones, and their value is arguable. I mean, once you get below a certain point, it's hard to prove things matter.
Can you hear the difference between 0.01% distortion and 0.00001%? most don't think so.
I think the best things that most measurements can tell usu about a DAC is whether the designers and manufacturers paid reasonable attention to detail for the money they are asking. :)

In the case of headphone amp / DAC combinations, you can tell a little about how well they will do with different headphones too.

Besides Stereophile, here is a great review (I mean well written) of a headphone/DAC that goes into some detail.

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-topping-dx3pro...
All of this is essential reading, but the explanation of digital specific measurements begins on page 5:

https://www.stereophile.com/features/112/index.html
Thanks guys!

Any info on how these things are measured?  The actual process involved?

Thanks!
They use an audio analyzer, Audio Percision is a popular one, Stereophile measurements always start off stating what gear they used. And the ones used can be $25,000 to $50,000 and higher, it is not cheap to create a piece of test gear that can be very accurate down to insane thresholds.  
  
Let’s walk through the measurements of the Benchmark DAC3 ameasured by Stereophile (AudioScienceReview also measured it), link to measurements
 
Figure 1: Testing the impluse and seeing what type of filter is used, normal people don’t really need to care about this.  
  
Figure 2: Testing the filter for 44.1kHz (CD and most digital music). Ideally, there should be nothing after 1/2 the sampling frequency (talking the red line), so 22.05kHz. We can see that by ~24kHz, the red line is down to -120dB, so that’s good. The blue line is also showing aliasing and “images”, there are no spikes above -100dB, so that’s most likely not audible, which is a good thing. 
 
Figure 3: Testing frequency response. We see it is super accurate for the Benchmark, almost no alteration. 
 
Figure 4: Testing noise and whatnot. We see it’s so good it’s almost not on the graph (meaning even >$20,000 equipment can barely keep up).  
  
Figure 5: Testing distortion. We see besides the 1kHz tone, and distortion spikes are super low. 
 
Figure 6: Testing square wave. Should look similar to this , and both red and blue should be close in shape and level. 
 
Figure 7: Should look like a sine wave. 
 
Figure 8: Testing distortion for bass notes. 
 
Figure 9: Testing for IMD (Intermodular distortion) in the treble, which is more audible. Again, besides the two main notes, any distortion is very low (disregard the rise all the way to the left). 
 
Figure 10: Testing for jitter. The levels should be low and any distortion spikes (timing errors in regards to how in sync the 1’s and 0’s are) should not be higher than the green line. 
Post removed 
@elizabeth

All modern CD systems have error detection/correction, no real difference between a modern one you buy from Walmart and a $20,000 one in this regard. DACs deal with jitter, they aren’t making up for data error like CD players can.
The data on the CD is physical not 1s and 0s, the laser reading process is analog and the error correction codes doesn’t correct all errors. But, hey, what do you want for nothing?
Post removed 
You’ve learned my lessons well, grasshopper. See, it’s not really complex, after all. 🤗

Beyond jitter, Signal to Noise ratios, THD, IMD, etc, how well any particular brand and or modle therein pulls off the ensemble manipulating everything into anlog is the trick. if there is indeed a trick to be had.


spec sheets are not a road map just a general or rule of thumb guide.


in other words, several dA converters can tout (by their own in house determined stats) or claims their specs are beyond reproach yet the analog aspect is as key as is the digital side of the conversion. 


is it an up sampling, over sampling, or non over sampling DA?


how well put together is the whole thing? what about the power supplies?


then too, dA’s come in various topologies.


does it excell at converting High Res files but is only so so with red Book?


ultimately, it wil all come down to its presentation. is it liquid? Dry? Dark? overly warm? critically resolving? bright? soft on the bass? transparent to a fault? does it do justice IYO to the entire bandwidth? bottom, middle and upper ranges?


lastly, how synergistic is it in your arrangement once it has landed on your rack?


there is no true concensus for which dA will work out best in every outfit.


albeit, popularity around herre tends to explode on those units in the more attainable, less can’taffordium modles. usually those below $5K. often far less and around $2K to 3K.


the D/A waters as I’ve seen them get really murky about the 5K level, if not a tad below it. meaning, at this point these are all competent capable converters. which one’s voice suits the prospective buyer is the real question.


specs are nice things to consider. always. but each and everytime the proof is really in the ‘pudding’.. 


I’ve heard D/As costing enormous sums I’d not own even if I had the dough to throw at it/them.


for free? of course I would take what ever.


do consider more than just numbers on a sheet of paper. then do everything possible to get which ever unit into your own hands and listen for yourself, that measurement is the only one that matters in the end.  


good luck. enjoy.


@geoffkait

What would you say to those that rip CDs using their computer’s drive and those using dedicated transports, yet they use the same checking software and database to verify the rips are identical?
That’s what they all say. The original CD was supposedly perfect, too. Or it might be a clue that CDs aren’t perfect after all.
@geoffkait And yet software was installed off CDs for years and installs worked 99.99999999% of the time. Even one bit being read wrong can render an entire software package unusable and somehow it didn't happen. Voodoo.
astelmaszek68 posts03-02-2019 9:12am@geoffkait And yet software was installed off CDs for years and installs worked 99.99999999% of the time. Even one bit being read wrong can render an entire software package unusable and somehow it didn’t happen. Voodoo.

>>>>>Yes, interesting, isn’t it? Yet, we audiophiles have been improving the sound of CDs for almost as long as they’ve been around.

By the way, another good argument for the perfection of CDs is buffering, like in my Sony Walkman portable CD player that supposedly makes the player immune from error. But guess what? It’s not. Looks like you fell for the Big Scam - “ Perfect Sound Forever”

Guess what? Reed and Solomon were not (rpt not) audiophiles. They were just some old guys hired to create the illusion that CDs worked perfectly.

“It’s what I choose to believe.” - Dr. Elizabeth Shaw in Prometheus
Post removed 
@elizabeth

They aren’t guessing though. The data is in packets, if the data is missing/wrong, they have an algorithm to know exactly what the missing data is, as it’s related to the other data in the packet.

Again, you can literally drill a small hole in a CD and it’s outout would be identical to the same CD pre-drill.

Again, there are checking databases, if even 1 but was wrong, it would show up as not a perfect copy. 
 
No need to talk about issues that don’t exist, like cable risers to reduce static electricity.