Review and Measurements of Massdrop x Grace Design SDAC

[bunkbail]

Use an EQ to drop the FR above 12kHz by 0.25 dB and see if that's audible in blind testing. I'd bet it's not.

Whatever you're betting, I'll double it. The 0.15dB roll-off is so insignificant.

 

[flipflop]

Use an EQ to drop the FR above 12kHz by 0.25 dB and see if that's audible in blind testing. I'd bet it's not.

Whatever you're betting, I'll double it. The 0.15dB roll-off is so insignificant.

The roll-off starts halfway between -0.5 and -0.75 dB and ends halfway between -1 and -1.25 dB, so it's 0.5 dB as mindbomb correctly stated.
What I can and can't hear is irrelevant. Research shows that changes in SPL below 0.2 dB can be audible.

 

[Mike S]

The roll-off starts halfway been -0.5 and -0.75 dB and ends halfway between -1 and -1.25 dB, so it's 0.5 dB as mindbomb correctly stated.
What I can and can't hear is irrelevant. Research shows that changes in SPL below 0.2 dB can be audible.

The piece of data that is missing from the equation is the bell shaped curve of human hearing which peaks at 1kHz or so. So yes, .2 dB is quite audible for those accustomed to listening for it (i.e. people who mix and master and I suppose audiophiles) when the change is in the "hot range" of human hearing. But a 12kHz shelf -.15dB ? No way imo.

 

[amirisourguy]

The roll-off starts halfway been -0.5 and -0.75 dB and ends halfway between -1 and -1.25 dB, so it's 0.5 dB as mindbomb correctly stated.
What I can and can't hear is irrelevant. Research shows that changes in SPL below 0.2 dB can be audible.

0.5dB at 20kHz, a frequency most people probably lost hearing of since long ago. it's disappointing of JDS especially since it doesn't correlate to their own measurements (-0.12dB at 20kHz) but I'd say it's not much of an audible concern if you're past 15.

fwiw when I measured my OL DAC I got -0.20db at 20kHz so it's a bit better but the early-starting roll-off was there too.

 

[Mike S]

LOL, I can't even remember when I could hear 20kHz...but I tested recently and can still detect 17kHz...regardless, music happens between about 50Hz and 8kHz (many would argue even lower), everything above about 10kHz imo is detail and "sound stage" (i.e. necessary and desireable, but not discernable notes). IMO, a very tiny HF rolloff would not concern me in a DAC if it delivers the goods in the lows and mids. But my perspective is one who mixes and masters.

 

[PuX]

wow, did not expect it to be so good. I wonder why they haven't made a second drop yet.

 

[Arnold Krueger]

Use an EQ to drop the FR above 12kHz by 0.25 dB and see if that's audible in blind testing. I'd bet it's not.

Below is the ABX level matching criteria, which requires one to characterize that approximately 0.5 dB drop at 20 KHz as being some fraction of an octave. If you pick 1 octave, the dip at 20 KHz would have to be like 3 dB to be outside this specification which is way conservative. So your speculation would seem to be well informed. Another relevant authority would be "The Modification of Timbre by Resonances: Perception and Measurement" by Floyd Toole J. AudioEng.Soc..Vol. 36.No.3. 1988 March TOOLE FLOYD E., which I find to be even harder to apply.

 

[Jimster480]

Well the real takeaway here is $79 SDAC > AudioGD NFB-27.38?

Nice job beating a product that is over $1500

 

[amirm]

according to the jds labs website, the ol dac is 2 Vrms. are the specs wrong?

OK, I measured this again this with 1 Khz tone and they are both in spec:

OL DAC: 2.002 L, 2.003 R
SDAC: 2.160 L, 2.168 R

I must have been looking at levels with a different spectrum than 1 kHz 0 dBFS.

I corrected the original post.

 

[Jimster480]

https://www.head-fi.org/threads/massdrop-x-grace-design-sdac-a-budget-beast.859273/

^ Look at this nonsense. Calling ASR bias and saying that they paid Amir to test the DAC favorably....

The guys comments make no sense.

 

[bhigh]

wow, did not expect it to be so good. I wonder why they haven't made a second drop yet.

Probably capacity. They had stated that another drop would happen in early 2018 but it's yet to appear.

The SDAC has shown up in a few other collaboration products like the O2+SDAC, the CTH+SDAC and CTX+SDAC.

I'm very happy with the O2+SDAC combo and ATH-MSR7GM that I have on my desk at work.

 

[mindbomb]

OK, I measured this again this with 1 Khz tone and they are both in spec:

OL DAC: 2.002 L, 2.003 R
SDAC: 2.160 L, 2.168 R

I must have been looking at levels with a different spectrum than 1 kHz 0 dBFS.

I corrected the original post.

Okay. I think the thd 1khz results were also at a few decibels below 0 dbfs, because they actually look too good on both products.

 

[amirm]

I think the thd 1khz results were also at a few decibels below 0 dbfs, because they actually look too good on both products.

Too good from what point of view? If you mean noise level, then that is a feature of my measurement gear which is able to push the noise floor way down. The peaks are the peaks though.

 

[mindbomb]

Too good from what point of view? If you mean noise level, then that is a feature of my measurement gear which is able to push the noise floor way down. The peaks are the peaks though.

Yea, I was talking about the peaks. All the harmonics less than -100db seems too good at this price point. And in the ol dac review, they were higher.

 

[amirm]

I ran a couple of quick tests against Topping D10.

First, 1 kHz residual noise and distortion:

As we saw with OL DAC, the Topping also has significantly lower amplitude second and third harmonic. On the other hand, the higher order harmonics > 14 kHz are either not there or better in SDAC. Their levels are ridiculously low though (around -130 dB) so I say overall, the SDEC loses here again due to those low order peaks.

Next I ran J-test for jitter and noise:

The noise floor is equal in both products. The Topping however has those little green jitter peaks on each side of our main 12 kHz tone which SDAC does not. Once again the levels are astonishingly low at -130 dB or so. Still the nod goes to Grace SDAC.

Overall, this is a mixed outcome. So you can judge for yourself. Me? I really like the cute display on the Topping D10 and its much more sturdy build. So that would be my choice.

I then used my Stax SRM-007t to AB switch between the two DACs on some of my reference tracks. Without level matching, the SDAC soundeded better with more detail and impact. Level matching in Roon took just one tick of the volume control. Once there, I cannot detect any differences.

 

[mindbomb]

Okay, I guess really low THD is not that uncommon at this price point. But the ol dac had much worse performance in the ol dac review than in this review on the 1khz THD test.

 

[amirm]

But the ol dac had much worse performance in the ol dac review than in this review on the 1khz THD test.

That is a different test methodology than this one. In this test a steep notch filter is used to get rid of the 1 kHz tone. That filter likely also impacts some of the spectrum around it. That said, when I have time I will see if I can better characterize the differential.

 

[jseaber]

0.5dB at 20kHz, a frequency most people probably lost hearing of since long ago. it's disappointing of JDS especially since it doesn't correlate to their own measurements (-0.12dB at 20kHz) but I'd say it's not much of an audible concern if you're past 15.

fwiw when I measured my OL DAC I got -0.20db at 20kHz so it's a bit better but the early-starting roll-off was there too.

FYI - This is another matter of different test conditions. amir tested at 48kHz. The specification you cited was measured at 96kHz. Screenshots and details are published here. Setting sampling rate to 88.2k, 96k, or 192k (optical) achieves FR < -0.15dB through 20kHz.

The optional DIP switch filter selections for the AK4490 indeed alter high freq response.

 

[Jimster480]

@amirm You have "SDEC" instead of SDAC 2 times in your previous post.

"The noise floor is equal in both products. The Topping however has those little green jitter peaks on each side of our main 12 kHz tone which SDAC does not. Once again the levels are astonishingly low at -130 dB or so. Still the nod goes to Grace SDEC." <- Right at the end

"As we saw with OL DAC, the Topping also has significantly lower amplitude second and third harmonic. On the other hand, the higher order harmonics > 14 kHz are either not there or better in SDAC. Their levels are ridiculously low though (around -130 dB) so I say overall, the SDEC loses here again due to those low order peaks." <- About 10 words from the end


The tests show that is basically a toss-up between the two? I mean even from linearity tests I think the two are on par!

 

[trl]

https://www.head-fi.org/threads/massdrop-x-grace-design-sdac-a-budget-beast.859273/

^ Look at this nonsense. Calling ASR bias and saying that they paid Amir to test the DAC favorably....

The guys comments make no sense.

I believe you are referring to https://www.head-fi.org/threads/mas...ac-a-budget-beast.859273/page-5#post-14129402 exact post. However, the guy did apologise at the end of the page. It's common mistake some people do when don't read the entire article at least twice and have no measurements to compare with.

P.S.: Now I'm starting to wonder if/when ASR and Amir will get banned on Head-Fi. They already banned words like nwavguy, sbaf and many more. This looks so...communist to me.