Which DAC measurements do you like to see next?

[Angsty]

I actually have zero interest in more stand alone DAC measurements that are merely a chip with a decent implementation of the application note around them.

Zero interest because those avoid the true challenges, imo.

I am interested in DACs with preamplifier capabilities, awesome integrated headphone amplification, and configurable sub output(s).

That’s funny to me because my primary DAC for many years was a Bryston BP-25DA preamp. It’s dated regarding inputs now (only coax), but with 44.1kHz/16 bit audio, it sounds just as good as my more contemporary DACs.

My perspective on DACs now is that I buy them for inputs and other features, not for SINAD and linearity.

 

[Tell]

Is it just me that's totally not interested in seeing yet another 100€+ perfectly measured DAC review? I mean we know that DACs are a solved problem with enough cash spent on it, so now I just want to see how good DACs can get when spending less than 100€ , or maybe even less than 10€.
Sure we might not get 120dB SINAD, but say 90dB with a flat frequency response is more than good enough for most application, as long as it has the features one needs, so either a small desktop DAC with balanced outputs or a small dongle for your laptop or phone.
Not that I have any particular in mind, so this is more of a broad request

 

[Basic Channel]

Is it just me that's totally not interested in seeing yet another 100€+ perfectly measured DAC review? I mean we know that DACs are a solved problem with enough cash spent on it, so now I just want to see how good DACs can get when spending less than 100€ , or maybe even less than 10€.
Sure we might not get 120dB SINAD, but say 90dB with a flat frequency response is more than good enough for most application, as long as it has the features one needs, so either a small desktop DAC with balanced outputs or a small dongle for your laptop or phone.
Not that I have any particular in mind, so this is more of a broad request

I would go even further than that and would like to see random laptops built in DAC tested. I would like to know whether generally speaking it'd make any difference to upgrade a DAC from bog standard. Or conversely at what point it'd be so bad that I'd be able to hear it on my setup.

 

[Bleib]

I would go even further than that and would like to see random laptops built in DAC tested. I would like to know whether generally speaking it'd make any difference to upgrade a DAC from bog standard. Or conversely at what point it'd be so bad that I'd be able to hear it on my setup.

To me that would be pointless, there's just too many of them, and they would far more quickly be outdated and unused than say buying a SMSL SU1 which would be OK for decades.

 

[NTK]

I would go even further than that and would like to see random laptops built in DAC tested. I would like to know whether generally speaking it'd make any difference to upgrade a DAC from bog standard. Or conversely at what point it'd be so bad that I'd be able to hear it on my setup.

FWIW, I took a scope capture of the headphone output from my Thinkpad P1 Gen 3. The signal is an impulse as I was trying to look at its impulse response (i.e. the response of its upsampling filter). This was what I got. See that glitch near the peak of the impulse? It was less than 10 μs wide (period of a 100 kHz sine wave). WTF!

No problem with an Apple dongle.

 

[Tell]

To me that would be pointless, there's just too many of them, and they would far more quickly be outdated and unused than say buying a SMSL SU1 which would be OK for decades.

Would still be interesting to see a random selection of them to see how good or bad it might be

 

[somebodyelse]

I guess Archimago's laptop tests were a while back. In my experience they're still hit and miss with a lot of them being noisy.

 

[Basic Channel]

To me that would be pointless, there's just too many of them, and they would far more quickly be outdated and unused than say buying a SMSL SU1 which would be OK for decades.

That's a fair point. And I have thought about buying the cheap good DAC based on the measurements here (i think it's 60 EUR or so) but I stop and wonder whether there is anything wrong with my laptop output.

 

[Head_Unit]

Which DAC measurements do you like to see next?

I took this heading differently as in it made me think about actual tests which we never see any more:
- Noise modulation as mentioned by @mindbomb...I can't remember offhand how this was run. You could pulse a certain tone from low to high amplitude quickly and see if the noise floor pumps although I don't think that's how Stereophile used to do it. Mmmm...is this related to how dCS says their Ring DAC has a more random character of background noise, uncorrelated to the music?
- Staircase linearity-very visual, not sure offhand if it presents wildly different information than a linearity sweep
- The one (on a Denon test disc?) where a very very low frequency had a small high frequency tone riding on top. This "swept" the high frequency tone through the various bits to hear artifacts.

 

[oleg87]

- The one (on a Denon test disc?) where a very very low frequency had a small high frequency tone riding on top. This "swept" the high frequency tone through the various bits to hear artifacts.

Multitone already does this (as well as a bunch of other tones).

 

[dimxbox]

Audalytic AH90 ($600)
Topping DX70 Pro Octo ($600)

 

[nanook]

SMSL DO100 Pro
2x ES9039q2m, otherwise very similar to the D-6s.

6x OPA1612A, this could mean they have implemented the I/V stage as described in the datasheet of the ES9039q2m. This topology would be a bit cleaner regarding the individual outputs of the I/V stage (non-inverting + inverting) because the I/V- stages reference with respect to the common mode signal up to some kHz.
So far just speculation as long as I have not seen a photo of the PCB.

 

[The Me]

SMSL DO100 Pro
2x ES9039q2m, otherwise very similar to the D-6s.

6x OPA1612A, this could mean they have implemented the I/V stage as described in the datasheet of the ES9039q2m. This topology would be a bit cleaner regarding the individual outputs of the I/V stage (non-inverting + inverting) because the I/V- stages reference with respect to the common mode signal up to some kHz.
So far just speculation as long as I have not seen a photo of the PCB.

I'd be interested in this as well. Here are some pictures of the PCB. My unit does not have the 4V/5V toggle switch as was advertised.

 

[Bleib]

My unit does not have the 4V/5V toggle switch as was advertised.

This is where Topping is better than SMSL. A simple setting in the setup menu and voila

 

[nanook]

I'd be interested in this as well. Here are some pictures of the PCB. My unit does not have the 4V/5V toggle switch as was advertised.

Thanks for posting the photos. They are not clear enough to read the values, but the topology rather looks like they have mainly doubled the I/V-stage (light blue) compared to the D-6s (which is not a bad thing; it needs the offset adjustment potentiometers (green) and it likely is the lowest noise implementation).
They seem to sum the outputs of the dual I/Vs behind the passive EMI-filters, where they have some series resistance anyway (dark blue).
The AVCC generation for the DAC chips seems to be similar to the D-6s (orange) but there are two more regulators(?) (yellow question marks), so I can't really guess what they did and at what voltage they operate the analog section of the DAC chips.
Supply for the OpAmps seems to be +/-12V here, eventually they have supplied the muting switches from the regulated 12V too, but this is just guessing.
No idea what the blue film caps and the neighboring OpAmps(?) are used for.

I like the large through hole caps at the output of the SMPS - easier to change in case and they will survive a lot longer.

Anyway, this looks like small, nice improvements compared to the D-6s (in addition to dual ES9039q2m) - nothing audible at all IMHO.

I'm curious, how the topology of the I/V-stage proposed in the datasheet would compare to this implementation (assuming, both use THD-trim coefficients optimised for the individual product), but to me it seems they did stick to the proven topology they used in the D-6s.

 

[The Me]

Thanks for posting the photos. They are not clear enough to read the values, but the topology rather looks like they have mainly doubled the I/V-stage (light blue) compared to the D-6s (which is not a bad thing; it needs the offset adjustment potentiometers (green) and it likely is the lowest noise implementation).
They seem to sum the outputs of the dual I/Vs behind the passive EMI-filters, where they have some series resistance anyway (dark blue).
The AVCC generation for the DAC chips seems to be similar to the D-6s (orange) but there are two more regulators(?) (yellow question marks), so I can't really guess what they did and at what voltage they operate the analog section of the DAC chips.
Supply for the OpAmps seems to be +/-12V here, eventually they have supplied the muting switches from the regulated 12V too, but this is just guessing.
No idea what the blue film caps and the neighboring OpAmps(?) are used for.

I like the large through hole caps at the output of the SMPS - easier to change in case and they will survive a lot longer.

Anyway, this looks like small, nice improvements compared to the D-6s (apart from the dual ES9039q2m - nothing audible at all IMHO.

I'm curious, how the topology of the I/V-stage proposed in the datasheet would compare to this implementation (assuming, both use THD-trim coefficients optimised for the individual product), but to me it seems they did stick to the proven topology they used in the D-6s.

Thanks for the breakdown. I too wonder what benefit the film capacitors add since they are not used in the DO100 or D-6s. I only have a pretty basic understanding of electronics, but I figure they must increase the quality of the power supplied to the DAC section in one way or another.

I would think having dual DAC chips at least provides some kind of technical benefit. Maybe just not an audible one?

Here are some more pics, if they help. My phone isn't very good at taking macro shots. Neither am I.

20240409 085044 HDR

Image 20240409 085044 HDR hosted in Freeimage.host

freeimage.host

20240409 085006 HDR

Image 20240409 085006 HDR hosted in Freeimage.host

freeimage.host

20240409 084937 HDR

Image 20240409 084937 HDR hosted in Freeimage.host

freeimage.host

20240409 084734 HDR

Image 20240409 084734 HDR hosted in Freeimage.host

freeimage.host

20240409 084715 HDR

Image 20240409 084715 HDR hosted in Freeimage.host

freeimage.host

20240409 084658 HDR

Image 20240409 084658 HDR hosted in Freeimage.host

freeimage.host

20240409 084544 HDR

Image 20240409 084544 HDR hosted in Freeimage.host

freeimage.host

20240409 084517 HDR

Image 20240409 084517 HDR hosted in Freeimage.host

freeimage.host

20240409 084506 HDR

Image 20240409 084506 HDR hosted in Freeimage.host

freeimage.host

20240409 084404 HDR

Image 20240409 084404 HDR hosted in Freeimage.host

freeimage.host

20240409 084211 HDR

Image 20240409 084211 HDR hosted in Freeimage.host

freeimage.host

20240409 084145 HDR

Image 20240409 084145 HDR hosted in Freeimage.host

freeimage.host

 

[nanook]

Here are some more pics, if they help. My phone isn't very good at taking macro shots. Neither am I.

Thanks again for the detailed photos. EDIT: They are really good! Of course you can't compare the cell-phone lenses with a good camera - the cell-phone lenses almost always have significant aberration towards the outer field of view.
I have downloaded them and I will take a closer look.

Two things were quite obvious:
- The offset alignment is implemented with the same values for the resistive divider as in the D-6s
- The AVCC_DACx generation uses the same divider ratio as the D-6s, so it seems the analog section (AVCC) is supplied with ca. 3.67V (same as D-6s). This would explain why they reach a 2.7dB improvement in SINAD - 3.0dB would be the simple theory, but the I/V-stages contribute a bit as well.

The SOIC8 besides the blue film caps is the MOSFET switch for the OpAmp- supply (same as D-6s), so maybe the film caps (very low ESR) shall filter the several- 10 kHz ripple from the SMPS?

EDIT: Don't get me wrong. I recently ordered another D-6s and when it got finally shipped a few days later the DO100-Pro appeared. I would have happily paid the 50€ extra to upgrade to the DO100-Pro, but the D-6s was already on its way ....

 

[The Me]

Thanks again for the detailed photos. EDIT: They are really good! Of course you can't compare the cell-phone lenses with a good camera - the cell-phone lenses almost always have significant aberration towards the outer field of view.
I have downloaded them and I will take a closer look.

Two things were quite obvious:
- The offset alignment is implemented with the same values for the resistive divider as in the D-6s
- The AVCC_DACx generation uses the same divider ratio as the D-6s, so it seems the analog section (AVCC) is supplied with ca. 3.67V (same as D-6s). This would explain why they reach a 2.7dB improvement in SINAD - 3.0dB would be the simple theory, but the I/V-stages contribute a bit as well.

The SOIC8 besides the blue film caps is the MOSFET switch for the OpAmp- supply (same as D-6s), so maybe the film caps (very low ESR) shall filter the several- 10 kHz ripple from the SMPS?

EDIT: Don't get me wrong. I recently ordered another D-6s and when it got finally shipped a few days later the DO100-Pro appeared. I would have happily paid the 50€ extra to upgrade to the DO100-Pro, but the D-6s was already on its way ....

You're welcome. I'm glad you were able to get a better look at the board. You have provided some really good information. Thank you.

I ordered a DO100 a few months back even though I figured SMSL would be releasing a dual ES9039Q2M model in the near future to replace it. The DO100 Pro does seem to be a pretty nice upgrade over the DO100, and the two are basically the same price here in Canada. It certainly feels like a better value for what you get. I was also pleased to find out that the remote has much better range than the DO100.

I really feel that I am hearing greater clarity and detail in the high frequencies as well. They sound slightly less recessed and more airy to me than the DO100.

The only "issue" I've had so far is that the ARC input requires the DAC to be switched off and then on again to establish a connection with my TV. It is mentioned on SMSL's website that this is a known issue due to "the diversity of HDMI ARC interface protocols".

 

[iugurbas]

Hello everyone

Here is a photo of my 'Rod Rain Audio DA10' DAC with its new OPAMPS ->

(other changes are also made)

I still haven't taken any measurements on this device because I don't have the instruments to do so.

I have tried many OPAMPS and these are what I find the most 'musical', that is my own opinion: this notion is purely subjective since there is no device capable of measuring the quality of musicality...

Regards.

Hi again, i bought a rod rain da10 a few months ago, its's awesome dac. i love it and time to opamp upgrade
But I need help clarifying some concepts.
which ones for I/V section, which ones Buffer? (i only know 2 opamps for XLR, other one for RCA)

Regards.

 

[pavuol]

I concur with the new Yulong DA10. I asked Yulong about the update from the DA9 and someone there wrote back:

"Da10 is based on new digital signal and clock processing to bring better sound"

AK4497 based with a lot of interesting support goodies. On a short list for upgrade possibilities but I would love to see this thing tested before deciding to drop the expense. I have a huge appreciation for ASR as a Schiit and Audio-GD owner.

http://www.yulongaudio.com/pd.jsp?id=40

Not anymore "recent product" this Yulong D10, but quite cute design, something different to the myrriad of generic DAC boxes..