How much better could digital converters get?

I think we need to look at FPGA chips for the innovation. Off the shelf chips offer so much less.
Ultimately, its down to the quality of the source. In my view SQ is falling as consumer expectations are so much less now.
We live in the age where so many think that phones are Hifi. Recordings just don't need to be as good anymore.
Streaming is also much lower quality. Digital sources are also inferior to analogue.
DAC's can only do so much.
The source determines the end product.

 

It's funny, but having grown up in the era when I used to list to music on a GE trimline suitcase record player system like this one:



I have to say that audio playback quality is much, much better now on average for most listeners. Average consumer electronics -- including phones -- are enormously better than the average quality consumer we used have access to in the '60s and '70s in terms of, well, pretty much everything: distortion, noise, frequency extension, frequency balance, ability to do high spls cleanly and dynamics generally, detail. Today's gear is more consistent too. And that's just the baseline average for most people. The high end is higher too in very much the same areas. And streaming lossless FLAC can deliver fabulous sound quality. Some of the best sound I've ever heard in my system is high res digital via Qobuz. I would agree we're not necessarily in a golden age of recording -- so much recording, especially in pop genres -- is done entirely or almost entirely inside the box, and people just slap on the hot plug-ins for the now sound -- but the caliber of recording has long been pretty widely variable and today there are plenty of fine sounding recordings in the genres that I find myself primarily listening to (jazz and classical) coming out of actual studios delivering the recorded sound of music performed in space.

 

That’s the one area I’m keeping my eye on with expectations of improvement in as it progresses as I believe this technology is still very much on the upward portion of the development curve along with streaming technology.

 

Agree.

I grew up in the 70’s and bought my first stereo system in 1983. Todays HiFi is much better with lower distortions. When it comes to extremly high priced HiFi equipment in 2022, I more see it as personal jewelry, not that it sounds better than more affordable equipment.

I think we now are at a turning point in audio. Advanced technology will be more prominent in audio components years to come. You can see that in the reference digital amp from Technics. The Pulse Width Modulation technology is only part of the game, advanced digital technology that «see» the impedance curve of speakers, and amplifier technolgy that measures the pick-up connected to the turntable is another part of it. And more to come.

If I look into the future, I’m very sure that artificial intelligens will turn the industy upside down. It will be scary real. You will have the possibility to have Elvis singing in your living room in «flesh and blood». I’m not sure that I want that…
By that, no one will listen to tube amplifiers and turntables any more, it will sounds like…distorted music.

 

I agree with you; rather my comment was meant to denote that when you do spend <$X> you expect a certain level of performance.

However, as you note, spending <$X> does not mean you will get that level of performance.

That's what Wadia did with their Digimaster algorithm and Exogal tried to continue with.

Sadly the market has for the most part gone with off-the-shelf DAC chips and filters save for companies like dCS and others doing custom ladder DACs.

 

After inflation, how do those prices compare? i.e. do you think you're getting better sound due to technological advancement, or due to spending more money?

 

Now introducing [trumpet fanfare]
The Bob Carver Distortion-Tunable DAC!!!
(actually that's possible by now I'm pretty sure-not certain as it's not my area of electrical engineering)

 

Yep exactly, and Wadia’s contemporary Theta was doing custom filters in DSP chips. My 28 year old Theta is what I was thinking of in my comment earlier. It is 28 years old. I’ve heard some modern DACs and they aren’t better. At what cost does it take to better my 28 year old DAC? Whatever it is, that cost can come down. There is room for the sound quality of low cost DACs to rise.

 

FPGAs are off-the-shelf devices

 

actually, Wadia started by using fast 16-bit programmable DSP devices; later product generations moved to alternative hardware architectures to perform digital signal processing

 

My point was simply that the RME is that good despite it's modest price. I would put it up against any DAC twice it's price , maybe more. It is pretty amazing for the money

 

I have tested the RME ADI-2 DAC and found it terrible. Steril, lifeless, artificial unengaging.

 

As like anything in sound, one person's big, real improvement is inaudible, or of no relevance to another.

My first domestic CD player was the Linn Karik / Numerik and I never got to own the excellent CD12, though a work colleague and good friend did.

The last time I heard the last spec Linn CD12, was through a very well set up, revealing, full range active system. I was genuinely shocked at how the then reference level CD12 was comprehensively outperformed by the then new, Linn Klimax DS.

Moving forward to the present, Linn's latest DSM, with their own designed Organik, DAC is another leap for in performance. This video shows the lengths and investment that some manufacturers are prepared to go to and regardless of people's opinion of higher cost DAC and Linn Products, is fascinating viewing.

If you believe that all DACs are nearly identical, at the least this shows why this is not the case.




How much better will DAC technology be in 10 years time, for less money.....

 

Not my experience at all.....

 

I think it's both. I've always felt that the Op amp section of a DAC is where the magic happens.

 

FPGAs are programmable chips. They are empty of information when they are new. Many DAC manufacturers use FPGAs as part of their DACs. Chord, AGD, etc.

 

FPGAs are configurable chips.

They contain a pre-defined set (often chosen with certain application space in mind) of processing, control, memory, test, and other functional elements; ranging from ie patches of configurable logic and basic processing elements to function-specific hardware engines to programmable instruction-set processors; and all linked by configurable interconnect fabric/busses/etc to allow for creating desired system and architectural functionality, routing and dataflow, and control.

 

I agree with what you say, and to the language you use.

 

We've reached a great place in regards to DAC performance but not so much in component reliability. I love many of the Chinese DACs but I read too many stories about them failing after firmware updates. ASR's forum is full of those. Gustard tests great! One year later - Paperweight.

 

Exactly. FPGAs are off-the-shelf devices that can be reconfigured for various purposes. Many FPGA DACs (PS Audio, Chord, etc.) inherently function as sigma-delta devices but with customizations here and there whether it's PS Audio's Ted Smith backstory with 30bit upsampling then downsampling to DSD128 (eg. the relatively noisy PerfectWave DirectStream and all its firmware variations) or say Chord with their huge digital filter FIR taps (stuff you can play with on a computer if you want).

In the 2020's, I think it's important to realize that companies can market all kinds of rationales for why their converter might sound "best". The way these things work however are very much performing beyond human hearing if one steps back and run some controlled listening trials to reduce bias to assess the sound quality.

Sure, some (like NOS DACs) can sound "better" to some folks, filters could make a difference, etc. However, these are not necessarily "better" technically but more "different". If we're basically aiming to get "high fidelity" digital conversion, objective results tell use we're there already even if manufacturers might disagree or listeners have various opinions which they're entitled to.

 

DSD256 and 384K PCM with 21 bit resolution is close to physical edge. For 22 bits and more cryogenic cooling will be required. And technology for that is already available. But to fully enjoy it you will need to build the anechoic chamber as listening room.

 

Audio Note latest 16-bit D/A converter (whitout streaming possibilities and remote control) has a retail price at $153000. In every D/A converter the heart is the D/A chip itself. The converter can never sound better than the design and specs of the D/A chip, no matter how good the PSU is. In this Audio Note there is an old AD1865 R2R chip. At the time the AD1865 was released, the price was about $3.
Is this Audio Note Dac science or religion?

 

Or does it make sense to have a Dac that is better than the converters used on the recordings one listens to. I don’t believe major studios change converters with every model change. I’m guessing once they find ones they like, they stick with them for many years.

 

I dunno about DACs, but "modern" source material fed to these DACs need lots of improvement. History shows it can be done.

Meanwhile, software for streamers have a long way to go IMHO.

 

I’m more concerned about ADC. How much more information from the tapes can be transferred to digital.

If one takes video transfers as a guide, the difference from film to DVD, HD and now 4K quality are astonishing.