I am in the market for a new DAC, and PS Audio DS fits the bill. It will go into a rack so display is not a requirement but I want upgradability via firmware, great volume control AND multiple inputs. Is the DS Junior exact same DAC inside as the bigger brother? Just a display difference? If not, what else is different?
Also, I need help on HOW to best connect these as I currently use my HT Receiver as a pre-amp. Here are all the equipment I have and would like to know how best to connect them all?
Receiver, 7.1, mostly fed through HDMI for Movies, front channel PRE-OUTs to My Classe Amp
Classe 2100 Amp - fed by Receiver pre-outs
Bluesound Node 2 for streaming music – connected via optical to receiver
Bluray/CD player – connected via RCA to receiver
HOW do I fit a DS DAC into this mix? Ideally, I’d like the DAC to do all of the DAC work for my streaming music, CD player, PLUS if possible, to play music off my iPhone through Airplay or BT fed into receiver?!
Also, a follow-up question. Anyone knows of a DAC that shows Album Art out via HDMI? It would be ideal for folks like us, mixing HT and Audio together. I haven’t been able to find one. I find album art on the display a bit useless, rather output to a monitor for much larger impact and display!
You’ll have a hard time finding a non video based DAC that will send video out the HDMI connection. Noise from video circuitry can be very hard to avoid and so most DACs, even if they support video out, sound better with all of the video circuitry turned off (including the video clocking…)
The DS Jr and the DS Sr use essentially the same audio processing code in the FPGA (all from the same sources, except for how to talk to the display/control processor.)
On the other hand the DS Sr has a 6dB lower noise floor by replicating the output circuitry 4 times. The DS Sr has a transformer based output so it more naturally handles common mode noise on the output lines. The Jr uses a (good quality) balanced video opamp on the output instead, but opamps can’t always compete with transformers. The Jr has a knob for volume control and input selection which I like a lot. The Bridge II is built into the Jr, it has to be purchased separately for the DS Sr.
Thanks for your response and Marry Christmas! I guess you never stop working! So, I have to forget about album art on my display then… worth the compromise. The only other option I could think of, would be a streamer that has HDMI, so I could use the streamer album art, then get the audio out to the DAC… for now, will skip this as requirement.
Now, would either work with my setup described above? My amp has both Balanced and RCA input, could I connect them like below? I also forgot the network audio, I have a ton of music on my PC on the same network, my Bluesound could access it, and I might get Roon, so would DS act as Roon client on the network?
Streamer --> optical --> DS --> Balanced out --> Balanced input on Amp
CD Player --> SPDIF --> DS --> Balanced Out --> Balanced input on Amp
Network --> Network Bridge --> DS --> Balanced out --> Balanced input on Amp
I guess all those inputs could use the Balanced audio out. My receiver pre-amp out could be connect to my Amp RCA input, so essentially when I wanna watch movies, Amp would come on for the front channels from receiver, when I go to music, the DAC would be the pre-amp via Balanced audio. Right? I could then program it all into a Universal remote?
Now, last question, to take advantage of the DAC clock, I MUST connect via USB, right? meaning I need another streamer with that capability or through Roon/network? My assumption is Optial/SPDIF output use the streamer/CD player clock and not your DAC clock? Right?
Altho with most DACs the issues you are talking about can (at times) be a problem, they aren’t in the DS. The DS doesn’t track inputs clocks. In fact it doesn’t even recover the input clocks (e.g. it doesn’t have any PLLs or FLLs connected to any input). It uses pattern matching to convert the incoming bits into samples and throws them in a buffer. The DS’s clock takes them out of the buffer with essentially no jitter from the inputs.
Thanks, so using Toslink or SPDIF input with DS would still benefit from your DAC clock and getting rid of Jitter! That’s good to know! My understanding was for DAC to re-clock the audio it needs an asynchronous USB input!
To avoid asynchronous sample rate conversion the DAC needs to be the clock master for the system, but that’s not always an option. The DS measures the size of the input buffer and controls the DS’s clock based on whether the buffer is filling too fast or emptying too fast. The control loop is digital and doesn’t know anything about any input clocks, these separate the DS’s clock from the vagaries of the incoming clocks so incoming clock jitter is just not an issue.
You do not want to connect a preamp or source to both inputs, XLR and RCA, on your power amp at the same time. You can cause all sorts of problems. You could switch the cables out, but only have one set plugged in at a time.
You would also have to watch what your DS(jr) volume was set on, or risk sending a very loud signal through your speakers if you forgot to turn it down.
If you want to use your HT receiver as your preamp, your audio digital sources would feed into the DS(jr). Your DS(jr) would then go to your HT receiver through one of the audio inputs, XLR if you have one, RCA if you do not.
One option you did not mention is to use a high quality preamp with a Home Theater mode which would allow you to have a separate 2 channel pathway that would bypass your HT receiver.
Ted, a very basic question and intellectual game if it makes sense:
Given the assumption that a record player has 95% common knowledge inherited and 5% are individual/varying design that defines an own sound quality…that a preamp maybe has 80% common knowledge inherited and 20% own design influencing sound.
My feeling is that in terms of a DAC 20% is common knowledge and 80% is very individual approach…or would you say, all of you DAC designers have quite common basics and just vary in ideas to a much lesser degree of the whole package? My feeling is that even if we’d reduce this to DSD converting DAC’s, by the kind of FPGA programming, variation in opposit to common knowledge is much higher than within non digital audio technology.
Thanks. But my Amp has an input switch between RCA and XLR that can be triggered via IR remote module (should be added). So I could program remote to turn on Amp, switch to Balanced, turn on DAC, select Streamer input, etc. The other options of running everything through receiver is an easier way to do this, but I just didn’t want signal path to go through the receiver for my stereo listening, plus, receiver doesn’t have XLR out so rather have better connection from DAC.
An additional pre-amp may work best but then it adds more cost, heat, takes space in a cramped setup of mine (in an apartment).
I look at it another way - if you ignore cost there are plenty of ways to make a great DAC. The engineering comes in when you attempt to make one for a given price or put other constraints on it. I chose DSD just because I knew it could sound good (i.e. I liked SACDs) and I figured I could add things like PCM later if I wanted. (Fortunately the Spartan-3E I chose for my original prototypes was indeed big enough to do the math needed.)
Most DAC Audio Components use one of a handful (double handful?) of DAC chips from a few companies. The interesting tech is buried in those chips. The analog output stages, power supplies, etc. are often typical engineering driven by things like price, marketing pressures, etc. The principles involved in the chip design are pretty well understood by those companies (and people who have had a good engineering background), tho there’s always room for more innovation.
The “weirder DACs” have more overtly technical designs - there are few who would choose the path I chose. I feel the engineering challenges are fun, but most wouldn’t attempt to sell this much tech (with the high component costs) for as little as PS Audio is selling the DS and DS Jr. Whether some other approach could deliver the same quality for significantly less money I don’t know, but I doubt it. But no non-software based approach is going to deliver the essentially free (for the customer) upgradeability or allow such ad-hoc innovation.
Thanks much Ted! I got a few new insights from your answer and it also raised some more questions to understand.
What I don’t fully understand yet is, why it’s so much easier to make a good DAC when no price point is given. I can understand that e.g. you then can put in as many output transformers or as much computing power etc. as you want…so this would mean, HW overkill alone would make a great DAC and each developer knows the main basics necessary for using them to achieve great sound? I thought the secret of good sound lies more behind the knowledge to use the parts, no matter of cost and that few have it. But I think that’s your modesty to not mention that probably only a handful in the world can make more of it than a product that just enhances old digital strengths and single HiFi aspects without lifting sound to new musical levels as you did.
From you answer I see even more how special the SW approach is and how limited a pure DAC chip approach must be, as one can’t influence the effort put in those chips at least not at certain points of interest and importance.
It’s still unbelievable how competitive the DAC already was at its appearance and how much it improved since then. I can only imagine which huge chip update and engineering effort this would mean with conventional chip designs…it would simply be an all new product probably.
Do you know many and which other manufacturers do also use FPGA approaches and does any of them offer sound quality upgrades (not expecting they’d be free)?
If the answer is as I expect, I really don’t understand why other DAC’s are even considered by customers.
Big thank you once more to Paul and you for making this possible at this price point!
I am aware of these FPGA based DACs. I have no idea which can be user updated.
PS Audio
Chord
Tube Technology
Aqua Acoustic Quality (although the ouput of the FPGA is sent to resistor ladders)
Playback Designs
Holo Audio
dCS (for over thirty years)
It’s always easier to throw money at a problem than to fix it correctly. But at a price pone and over well trod land it’s only the quality of engineering that can distinguish products.
Ah, I didn’t take your original question as how many individuals can build a … DAC, but instead how easy it is to build a … DAC. Philips and Sony put a lot of solid engineering into SACD, IMO the DVD-A consortium just made small “improvements” over DADs, etc. Most of the DVD-A innovations were related to digital rights management and watermarking, brought to you by the same folks that did MQA. Note that DVD-A’s changed the bits in the audio to achieve their goals, and SACD used more expensive stampers to build SACDs and left the bits alone (at least to a first approximation.)
I know that very few individuals can design the architecture, design the hardware and write the software for a new anything, but companies can build good teams and do great work much more easily that find a one stop engineer.
There are two separate things I’d say:
Most companies that use FPGAs use them for some smaller part of the DAC: say doing upsampling or other DSP, or misc control functions, or input processing, but they still have separate hardware for, say, the S/PDIF receiver, the DAC proper… I think I’m the only one that uses an FPGA for everything from the decoding of input bitstreams to the final filter.
I know that Playback Designs still sells their original product because it’s software upgradable and that they’ve added significant function as well as sound quality improvements. I don’t know how may DAC manufacturers regularly offer sound quality improvement updates (but that’s my ignorance, not evidence that they don’t exist.)
Thanks again, so that should mean much more upgrade potential by improved programming…
In fact I meant not necessarily how many individuals are able to do all from circuit design to programming (I’m sure you’re quite alone), but even how many have the knowledge not only to combine HW to functional DAC’s which mainly improve on newer parts used (if you will build a DAC kind of “by handbook”), but also have the knowledge of the right points of major improvement in a design. But I think that’s what you meant with engineering to a price point. It corresponds with my impression that so far only DAC’s with money thrown at them were comparable.
I bet you are correct. To the extent I can tease out what other manufacturers are doing it is unclear to me to what extent the FGPA is involved. I mentioned what Aqua does above. dCs appears to similarly limit the FGPA’s role.
A couple of them indicate the units can be updated, but do not state whether they have actually offered an upgrade.
Now that’s something I was totally unaware of. Must have dozed off in class, zzzz… So was Bob Stuart involved in DVD-A, or were other folks involved with both?
Oh yes, MLP (the compression tech for DVD-A’s) is Meridian Lossless Packing. DVD-A was designed by a group of companies (Toshiba, Time Warner, Matsushita Electric, Hitachi, Mitsubishi Electric, Pioneer, Thomson, and JVC developed DVDs, I think most were part of developing DVD-A, but that’s just a fuzzy memory on my part.)
So, I have to forget about album art on my display then… worth the compromise. The only other option I could think of, would be a streamer that has HDMI, so I could use the streamer album art, then get the audio out to the DAC… for now, will skip this as requirement.