I had a bit of a revelatory a/b/c comparison tonight. You would think that there might be some linear progression in DAC technology. I can guarantee you that’s not the case. Up for comparison tonight were the Naim CDS3, the Apogee Element 24, and the Audioquest Dragonfly Red. Apogee is notorious for having some of the best clocking available. However, the Apogee had a very similar sonic characteristic to the Dragonfly, which runs at 1/10th the price, but with somewhat more dynamics. The jump from the Element to the Naim was absolutely no contest. The Naim had at lest 2 octaves more audible treble extension (not brightness), and as such, imparted that much more texture to lower-frequency instruments. The Naim uses Burr-Brown DACs that were introduced over 20 years ago, but it uses them in such a solid and detailed way as to blow everything else away. It requires ridiculously stable and strong power supplies, physically isolates all of the boards on spring suspensions, and then uses at least one DAC chip per channel. Then they also write their own custom firmware. So that’s why a 17 year-old high-end CD player can sound better than one of the best DACs currently available.
it is a very complex issue and opinions are really very different.
While others swear by the food, others say something else and that goes through the whole bank so it doesn’t matter if there is Akm, Cs, Burr brown what else there is. I think that each chip is potential and the focus is more on the amplifier . There is more you can do wrong than on the Dac. From bad components to bad circuits. I tested something for myself. A Cambridge Dac magic 100, Fostex Hp A3 and A4BL. Both listened to Fostex directly and later via line out on a headphone amplifier and the differences were sometimes very big.
In the end it was so that the Cambridge was the best with modifications, then the Fostex Hp A3 and then the Hp A4 BL. The HpA4BL sounded the least dynamic, was not so punchy and did not play so openly and transparently enough, and was musical not necessarily either.
I have no doubts that there are hardware that are 20-30 years old and can still keep up today and are sometimes better.
If you can pair it even better with current hardware. My Matrix Audio Ipro mini2s with headphones doesn’t sound as nice as on my system from the 90s. What suddenly opens up is just as enjoyable as if you have a Dac / Amp The great harmonizes with headphones. Many manufacturers also specifically refuse to do just one thing well and not consider other aspects. And stamp it later as house sound. Which is also an issue. If you look through here, many buy Dac / amp that is paired with a headphone amp.
And why because the Dac is better but the integrated amp is not.
Yeah DAC’s are hard to hear quality upgrades. I often wonder how many ways can you convert digital audio to analog? I year ago i would have thought what we use on our computers where more than good enough.
But ive experienced the Schiit modi and the SMSL SU8. The Schiit somehow added a lot of treble to the sound but the SU8 was a clear improvement as it seemed to not add anything. and it sounded better. I know this isnt the right word but it sounded classier.
When i tried the xlr Grace DAC it sounded better than the Modi also but it added a warmth to the sound and did something to the soundstage i did not like. you could hear it in certain recordings, like Crash Test Dummies Mmmm song, in the decay of cymbol crash.
I’m of the opinion that when you find a good neutral DAC for a decent price, maybe around 200$, your done. But maybe not. People on the forum keep extolling the virtues of R2R dacs. gets me wondering if i should invest in one. I’m told he Bifrost is really good.
Mon once bought a DAC that was over 10k! So i dont know. maybe its harder to hear improvements in sound quality in DAC’s but its still there.
I think it comes down to whether or not somebody :
- Has the hearing or trained ear to notice the difference.
- Has heard all DAC types of implementations and chips.
- Actually has the proper headphones to notice the difference.
- Actually has an amp that doesn’t disguise the difference cough cough…NOT THX…cough cough.
- Actually has the proper quality tracks to notice the difference.
- Has no agenda for confirmation bias (whether measurement or subjective listening).
- Has the ability to quickly switch and properly A/B/…
- If there is more than just rca input, is testing all input types.
- Has a proper and reliable source of power.
- If a DAC has oversampling and non-oversampling formats, is testing both.
There is probably more things I could list…but the rudimentary way in which many test their gear tends to make these subtle differences even harder to notice.
Are you thinking of Antelope perhaps? I personally wouldn’t say this is the case but ignoring that, your element interface doesn’t include this best clocking, it’s a budget interface, so I wouldn’t expect it to compete with a higher end interface with higher end clocking, nor would I say it represents a higher end apogee product either
What is one of the best dacs out there?
Just a bit confused as you are comparing a CD player that still goes for around the 3-5k mark used (some don’t even include the xps2) with a budget audio interface and a relatively budget portable dac/amp, so I sure would hope the naim still holds up lol
I got all that. except 2. 
What is 10 BTW?
Delta Sigma DACs oversample…they digitize an analog signal with a very low resolution at a very high sampling rate.
R2R resistor ladder DACs directly convert by each resistor adding it’s own weighted contribution to the output. Sometimes they oversample part of the signal, while other times they don’t. It depends on the design, but some R2R DACs have the option of both, like the Metrum and Parasound DACs if I remember correctly.
I’m interested to hear the impressions of the new SMSL m400 that is coming in May.
https://www.audiosciencereview.com/forum/index.php?attachments/qq图片20200415161157-png.58809/
Well. thats sound bad on the r2r dacs
Actually, it sounds bad on DS dacs since they cant even properly convert the signal without having to oversample, which makes things less accurate and more synthetic sounding.
I’ve heard both excellent oversampling r2r and nos r2r, really just depends on the design
I never understood the r2r dac. How can resisters do anything but resist current? if it was a transistor it would make more sense
I would agree with you , but add that a lot of budget DS dacs do suffer from cost cutting measures that result in an artificial sounding DAC. The only reason DS dacs became popular in the 1st place is that they are much cheaper to make.
oh god! no! not another one! forget i asked. i didnt ask anything!
Considering how the first DACs are R-2R and how the Sigma Delta is just… well it’s 1 bit just MUCH faster. Truth is there are at least about 3 transistors per output bit. So by the logic of “transitor=better”… well the R-2R has more transistors.
TL;DR: ELI5 answer: R-2R DACs use more transistors so they’re morer betterer.
Thanks for explaining that dago. i’m much smarter now
Only time will tell.
I wouldnt count on it