To DAC or not to DAC?

Audio Gear Talk
1

I need to clarify the intention of the subject as its pun may not be understood: the point is not about whether we need a DAC or not, which of course will always be there in any form. The subject alludes to the idea of whether we should bother (the shakespearian conundrum) to improve the DAC we already have, be it the built-in one or an external one.

Background: I enjoy listening to music (mostly vocal Jazz and bossa nova) and not so much to my gear. However I strive for perceivable sound quality and improve it if possible.

My setup: Roon core on iMac, Bluesound node N130, class A/B mono-blocks 43W per channel, passive preamp, Harbeth P3ESR, 44.1kHz/16bit files (Redbook format = CD format).

I listen from “near field”: 2m from speakers which are 2m apart.
Imaging is flabbergasting, vocals and instruments are live and bodied as if they were positioned on a soundstage in front of me. Speakers disappear.

This is as far as I go in appreciating the delivery.

Now, I loved so much the presentation of my P3ESR that lately I bought their new version, the XD (eXtended Definition).

At the very first listening I could appreciate a clear difference (more analytical, less warm, although with same vocal magic). It took me a few hours to decide that I also preferred it over the older, but the point here is that I could immediately hear differences without needing to AB: I only had to decide whether I enjoyed them.

Now we get to the core of this post.

Armed with this late experience of “improved” delivery, I wanted to experience the Dynaudio special 40 which I always read about as capable and neutral and also compare my Node with another streamer.

I booked an audition at local dealer (Special 40, Naim class AB amp, Bluesound node vs Cambridge Audio CXN, Roon source):

  • Special 40 were absolutely decent speakers, but not as engaging as my P3 (according to my taste)
  • I could hear a difference between the Node and the other streamer.

The difference between streamers was unquestionable, again not sure if I liked the CA better but it carried a sense of being more crisp and open. This I could clearly detect at each AB. As DAC is allegedly the big difference, the Node not having a reputation for top notch DAC, I assumed that DAC was the cause of the different renditions. Assumptions are good, but need to be tested for proof or rejection.

With that in mind I decided for a candidate external DAC that I would be willing to invest on, only if it proved to deliver a superior experience. I settled for a Chord Qutest, 1500€, about 3x the Node price! It has excellent reviews and measurements, it’s built rock solid and it is a nice piece of gear to place next to my mono blocks. So I borrowed one for testing from a local dealer.

First thing, I changed the output level of the Qutest from the default 3V down to 2V to match the Node’s output level (which is also the prevailing output level on modern gear). Higher output level would otherwise yield higher loudness which is perceived as better sound while it’s indeed just… louder. (Which made me wonder why Chord would want to set 3V as the default, when already 2V is on the high side, requiring significant attenuation in the pre amp stage already…)

With that done, this time around I could not hear any difference whatsoever between the Node DAC and the “far superior” Qutest DAC. Nothing at all, for as much my confirmation bias wanted to convince me that it was worth buying that cool looking piece of gear.

First conclusion: no audible difference between Chord Qutest DAC and Bluesound Node DAC.

I also tried with an old Apple AirPort Express, thinking that at least there the difference should come out: the Qutest could not even beat the AirPort Express DAC, no difference whatsoever even there.

Then, if it’s not the DAC, it must be the streamer I thought. As I did not have the Cambridge Audio with me, I now did an AB testing between my Node and the Apple AirPort Express, with the analogue output of their respective DACs. I configured both to be exactly the same and linked them in a Roon zone, so I could stream to both of them at the same time and AB in my amp (luckily the AirPort has the same 2V output level of the Node).

Also in his case, absolutely no audible difference in all the AB testing I did!

To summarize my observations using Roon as a bit perfect transport and Redbook format files:

  • a Chord Qutest DAC did not deliver any audible difference over the Bluesound Node internal DAC
  • Bluesound Node did not deliver any audible difference over an old Apple AirPort Express.

My conclusion is that:

  • Different speakers deliver audible difference
  • An excellent external DAC does not improve the performance of an honest internal DAC
  • Different streamers of different price points sound the same (unless coloration is purposely added).

So, where did the audible difference (between streamers) that I had experienced at the dealer come from?

My only 2 reasonable assumptions are:

  • The setup was not correct (e.g. the 2 streamers may have had some different volume levels in Roon / native, different DSP settings…)
  • Cambridge Audio CXN may have a strong acoustic signature before the DAC or in the DAC (hopefully not).

With all this said, I’ll go back to enjoy the music out of my very capable nimble system without pondering further changes.

Back to the post subject, whether “to DAC or not to DAC?”, my recommendation is “not to DAC”, sit down and enjoy the music. You most likely have all what you need for hours of great musical enjoyment.

4 Likes
3

If you read my post, I explained that I compared both internal vs external and then 2 different streamers with their internal. In all situations the outcome was the same: no difference.

1 Like
4

If you can’t hear differences between DACs, then your musical journey will be a lot less expensive than mine. I have tried a lot of DACs starting with the built-in DAC of an iMac to Schiit Audio DACs to PS Audio DACs finally settling with the Holo Audio May KTE DAC.

Most of the time it has taken me a lot of hours to understand if the differences I hear are better or worse…but the May KTE with HQPlayer was a stunning step up from anything else…

You have to decide what sounds better and what you want to spend money on. But, in my mind, DACs are super important…

5

That’s been my experience as well. When two reasonably competent DACs are precisely level-matched (to within +/- 0.1 dB or better), finding differences in a quick A/B test is rather difficult. However, listening to one for a half an hour or so every day for a couple of weeks and then doing the same for the other will often produce changes in listening behavior that can be used to deduce a preference.

For example, with which DAC did you more frequently listen to albums from start to finish? How many times did you put on a favorite track and find yourself listening from there to the end of the album?

3 Likes
6

I only listen through my Node’s DAC, at times full albums, at times songs. The point here is that during my listening session at the dealer, I could immediately sense the difference between the 2 streamers (one of which being the Node), but when I tried to reproduce a similar difference at home in different level matched comparisons (and using a very competent external DAC, much better on paper than what used by the dealer), I always failed. Everything sounded the same. I personally think that if an AB test doesn’t show differences on the spot, then there’s no point in continuing. Memory, psychology, confirmation biases, tastes’ adjustments all affect the final judgment. I liked what I heard throuh the Chord Qutest. If I had stuck with it for a couple of weeks, I would have only remembered that I was liking it and forgotten that it sounded exactly the same as the Node’s.

But again, I just wanted to share what was for me an enlightening experience. Not trying to convince anybody otherwise.

5 Likes
7

There are more productive ways to upgrade your system than by focusing on the DAC or streamer.

I have Harbeth P3ESR-XD speakers as well, which are excellent. The small size and sealed box design of the P3 is challenging for dynamics, and would benefit from more power than the 43W/ch you have now. I upgraded from a 50W/ch amp to a Schiit Vidar (100W/ch @ 8 ohms), and it made an audible improvement in headroom, transients and bass.

Adding a subwoofer fills in the low notes below the P3’s range (they only go down to 75 Hz), and enhances many recordings. Careful setup will make a subwoofer blend seamlessly with the P3.

3 Likes
8

Have you got your qutest connected via USB? It needs a USB connection.
Also it’s recommended by the designer that you use the power supply the chord comes with.
The designer says using linear supplies degrades the sound.
I upgraded to the qutest from aurelic and the difference was startling.
I also recently upgraded from aries to lumin u2 mini and once again the difference is startling.
Not all streamers are born equal. They do make a difference.
Otherwise a cheap and nasty matchbox bridge would sound identical to one costing 200 times more.
Chord use FPGA DAC chips which are regarded as the finest in the industry.
Check your settings in roon especially regarding the upscaling etc. The chord can be touchy about such things.
If you can’t hear a difference between the node and chord then I would suggest that there’s something not right in your set up.
Don’t forget that the qutest is exactly the same DAC as the Hugo 2 without the connectivity of the Hugo 2. SQ is identical. So you’re actually getting a DAC that is knocking up 3k For a fraction of the price.
I don’t mess around or worry about voltage. I don’t do russ Andrews style snake oil power cables and linear transformers. I do do high quality analogue cables though.
Good luck and I hope you find a solution

1 Like
9

Great read and if you’re happy, you’re happy, enjoy your music

2 Likes
10

I love that you went about this systematically and came up with what appears to be logical conclusions. From my personal experience, I started with a Khadas Tone Board. This is a bare bones (no case, only board) DAC based on a ESS chip. I loved the sound but thought less than $100 there must be something better out there. I bought a used Qutest thinking it would be a huge improvement. TBH, I can’t tell the difference. I continue to use it because it’s a better piece of kit, but is it worth 15 times as much as the Khadas? Only your wallet knows.

4 Likes
11

Interesting post. My experience matches yours, including the music preferences. The only thing I’d add to your detailed description is my experience with different CD transports using the same DAC. I had a Pioneer Elite CD Player and traded it for a Rega Apollo. There’s a clear difference in sound. The Rega yields a bolder, fuller sound, less analytical. There’s no question about it. Was it the DAC? No. The Rega sounds exactly the same via its internal DAC (Wolfson) and via coax to the internal DAC (AKM) of the Hegel H120 (my integrated), with Revel Performa loudspeakers. However, it sounded fuller than the Pioneer when both were used as transports to the Hegel internal DAC. Do I think DACs don’t make any difference? Not really. My impressions are only as good as my own (limited) experience and higher end setups may help higher end DACs to live up to their full potential and price. I don’t really know. What seems clear to me is that the relative importance of DACs has been blown out of proportion in recent times, maybe because there’s so many of them in the market and its so easy to come up with another “next generation” dac chip, each one being the best thing since sliced bread. It’s also relatively easier to experiment and buy new DACs, compared to trading amps or loudspeakers (where the real impact is clearly felt).

1 Like
12

Hi Vitor, I agree that transports could have an audible difference (e.g. I heard it at the dealer between Node and Cambridge Audio), but I also think they shouldn’t.

Adding a sound signature can be intentional in an otherwise commoditized market, but this has nothing to do with HI FI.

When in my test I compared Node and AirPort express, meticulously configured to be the same within Roon, I could not here any difference in their presentation, thus suggesting that well done streamers (not only the DAC portion) can (and should) sound the same.

So when I hear that Node and AirPort Express sound exactly the same and Cambridge Audio different, my bet goes with Cambridge Audio being less competent (albeit possibly more pleasant).

Don’t quote me on this specific comparison, as I don’t know how the setup of Node and CA was done at the dealer: maybe my perceived difference was caused by an inaccurate setup of the 2 streamers in Roon (different volume levels, different DSP…)

1 Like
13

Yes. It’s a carry-over from a couple of decades ago, when people were still figuring out how to do this right, and there were significant differences between the sounds of various competing well-regarded DACs. But it is pretty much a settled science these days, chipified as all tech eventually becomes, and there should be little or no difference between streamer/DAC combos.

Indeed!

Perhaps it would be kinder to say, less “high fidelity”.

1 Like
14

I tend to agree that competent/transparent streamers don’t sound so diffferent, as some people claim. I was using the Zen Stream as a Roon End Point, feeding the H120, but when the Hegel became Roon Ready I came to the conclusion that its internal streamer sounded as good as the ZS. Both sound very good to my ears and work flawlessly with Roon. So, I moved the ZS to a Head-fi system in my office, coupled with the Zen Dac V2 and the Sennheiser H600. I’m now enjoying two systems (speakers and headphones), fully integrated by Roon. I stopped looking for gear and wondering if I was getting the most out of what I have. I’m happy and enjoying music with outstanding sound quality. There’s definitely a lot of noise in this business.

1 Like
15

As a 40+ year cost-is-an-object audiophile, I’d like offer a little perspective please, on what I think I’ve learned about the possible importance of a DAC.

First, as obliquely alluded to here, the more resolving your system is, the more a DAC matters. If you have VERY carefully auditioned and selected every piece of your system, AND have selected high value components (value determined by audible performance per dollar spent) AND you have tuned your listening room to audio playback, once you’ve gotten to the I-could-have-bought-a-nice-new-car-with-that-money level, IMO your DAC is of the highest possible importance.

On the flip side, if you’ve invested 5 to 10 times less than that OR you are not carefully auditioning every piece in your system (yes, I’m including every cable, vibration isolation, power supplies, etc.), or your gear is in a suboptimal playback space, then it still matters, but matters a WHOLE lot less.

For myself, my highest investment in my system is my DAC, and the positive difference was instantly audible when I went from a well regarded DAC to one four times the cost, obviously from a reputable manufacturer. It was so much better that it rendered some of my signal improving gear instantly obsolete.

On a final note I will say that if you have not bent over backwards to minimize jitter (lots of ways to tackle that problem) spending big buck on a DAC is just throwing money away, no matter what else you may have done.

As far as streamers are concerned, I cannot comment, not having used a DAC that did not have an Ethernet input, fed from a Roon core for as long as Roon’s been available.

Good luck in your journey. :slight_smile:

1 Like
16

Thanks for your message. I guess you’re confirming my statement (“My impressions are only as good as my own (limited) experience and higher end setups may help higher end DACs to live up to their full potential and price”'). I few days ago I was looking at the website of an European hifi dealer: the cheaper DAC was the Qutest and the most expensive was in the region of 65,000 Euros. I hope these DACs are able to turn dust into gold. I have no reason to doubt your description of the effect of high end DACs, given the right setup and gear, but, has you’ve said, in entry level, mid-level, whatever you call a hifi system costing less than 10,000 USD/Euros, DACs may not matter that much because most tend to be pretty good and resolving these days. However, this is not what so many reviewers and audiophiles believe or want us to believe. Anyway, I don’t want to engage in any sort of controversy about this. It’s just my opinion now, based on my limited experience. Cheers

17

Yes, both confirming and providing a bit of personal experience, without being specific about gear. Just talking about general principles.

I guess the point I’m trying to make, is that DACs can make an enormous difference, but only when everything else is right. And using the chain analogy, our systems are only as good as their weakest link. So when I say everything has to be right, I mean EVERYTHING.That plus helping people understand that a spectacular DAC is not going to make a massive improvement to a modest system. It’ll almost certainly will make it better, but likely not nearly enough to justify the cost. While in a different system, where everything is right, and the rest of the gear (and the room) is up to the quality of that very same spectacular DAC, it will likely make a huge difference vs. lesser DACs.

1 Like
18

The good news here is that there are now lots of “spectacular” DACs, and they don’t cost very much, either! Rejoice!

After all, there’s only one correct waveform to generate for a given digital input, and we’ve seen over the past ten years lots of low-priced DACs that get that right, for only a few hundred dollars. Switching to a different DAC that generates the same waveform won’t change your sound, regardless of the rest of the equipment in the chain.

4 Likes
19

To the best of my understanding, that is quite incorrect, and while a reasonable conclusion to draw, it just does not line up with the facts. In that regard, it’s a lot like the old “bits are bits” arguments regarding Ethernet cables and switches. Sounds logical, but we’re recently learned it just ain’t so.

I’m no hard core techie on this particular matter, but this is my humble understanding…

The process of digitization is by nature lossy. That’s why higher sample rates are preferred, due to a potentially more accurate reproduction of the original waveform. Higher granularity of the sample leads to a more faithful analog reproduction. Thus ALL DACs are making things up, filling in the blanks between the samples, as it were. And there are different methods for doing such, and some DACs do a better job than others.

Then there is the fact that the process of recreating an analog signal from a digital source creates artifacts, things added to the signal that were not there originally. The most common being pre and/or post ringing. There is no “agreed standard” way of dealing with those artifacts, so different DACs deal with such differently, and thus WILL sound different.

On top of that they all have to deal with jitter - jitter in the incoming signal - jitter created in the DAC itself - and jitter induced by external vibrations. I know from personal experience that such jitter (from all three sources) have a profound effect on the quality of the analog signal produced. How a given DAC deals with that jitter will have a significant difference in the musical signal that they produce. As you move up to more expensive DACS they do an increasingly better job of dealing with that jitter, as you spend more, generally speaking.

It is true that the quality of DACs has improved dramatically in recent years. And many fine DACs can be had for not a lot of money. But those same, fine, inexpensive, DACs generally do not do an adequate job of dealing with jitter. As a rough touch point to see how hard any given DAC was designed to combat jitter, look at the quality and isolation of the power supply, and look at the case. Is it heavy, built out of only one or two pieces? Those aren’t proof of dealing with jitter well, but they are strong indicators of such. And you can know that for any DAC without those things, the DAC could sound better if those things were better paid attention to, as - if correctly designed - they can reduce jitter and make any DAC sound better.

That said in most systems - I’m just guessing what that is now - whether or not the DAC does a good job dealing with jitter or artifacts may not even be audible because the level of resolution needed to hear it is just not there in the playback system and room.

So for most people an easily affordable, quality DAC may be all that is needed. And thankfully there is a bumper crop of those these days.:slight_smile:

PS - Apologies for the long post, but it seemed a service to our community to clarify the facts.

2 Likes
20

Who is “we”? I haven’t learned anything like that recently. The more I learn, the more I understand why bits are bits.

Pretty much all ADCs these days digitize at a much higher sample rate than the final rate, usually 128fs or 256fs (i.e. 128x or 256x the CD sample rate) for good reasons. But, regardless of how lossy the digitization is, once the process is done, a DAC can only take the digitization as the truth, and there is still one correct analog waveform that corresponds with the digital waveform.

According to the sampling theorem, there is only one method of “filling the blanks”. Most DACs today, even budget ones, are perfectly capable of doing this with an accuracy that exceeds our hearing capabilities.

Those artifacts, as I said above, are inaudible, including ringing. And there are agreed standards of dealing with it, the two common ones being linear phase filters (which have both pre-and post ringing) and minimum phase filters (with only post-ringing). This doesn’t change the fact that there is one correct way of generating the analog waveform - the linear filters. I am willing to bet most people - if not all - won’t be able to tell the difference between the two in a blind test.

A DAC using an asynchronous interface like USB is not affected by that. Unless of course jitter was introduced at digitization time, in which case it cannot be corrected.

Accurate DAC clocks are a commodity these days. Their jitter is orders of magnitudes smaller than what we can perceive.

With all due respect, that is nonsense.

Power supply and the case are not jitter factors. Unless of course you believe in the nonsensical vibrational jitter.

It’s the human auditory system in our current evolutionary stage that is lacking the said resolution.

I would say to the contrary.

6 Likes
21

Cambridge Audio internally upsample any digital input to 384 before passing to the DAC (always have)

Also with the range of DAC technologies/chips around there is scope for differences.

As far as I am concerned CA are VERY competent