[Delirious Lab]

Something strange and scary happened as I was trying out Foobar today.

The setup: MP3 (192 kps average) => Foobar/Wasapi => Fiio E07K => Onkyo receiver => Polk RTI/A1 speakers with PSW-10 sub.

It sounded good. Too good. At least as good as the CD player, a 6-disc Onkyo from the mid-1990s.

I only had a half-hour of first impressions, but it seemed that the lowly Fiio made wonders out of these low-res rips. Imaging and soundstage were very impressive.

I want to make a more systematic comparison between the external DAC and the CD player, but in the meantime my question is this: Would the latter's DAC be so obsolete after 20 years as to be outperformed by entry-level portable stuff today?

I fear the answer... the old thing only has RCA outs so it's worthless as a transport.

 

[BackToAnalogue]

The answer to that is that they have improved enormously. In fact they have improved enormously in the last 3 to 4 years.

I had been using a cheap Behringer 202 16 bit DAC (but a very good 16-bit DAC chip) until fairly recently in my study. Comparing that to my 10 year old expensive Cyrus CD player, the Behringer sounded slightly better than the Cyrus. The Behringer cost £20 and a PC CD drive another £20, job done, HiFi CD quality for £40.

So then I bought a cheap HiFImeDIY ES2013 DAC for the exorbitant sum of £45. And it sounds incredible, and with the few 24/192 files I have managed to find it is quite breathtaking.

Then I bought a DX90 with twin Sabre ES9018KM DACs. Not yet burned in or tried properly but through phones the sound is awesome with the best recordings. I handed it to a friend a few days ago, got him to plug his ipod phones in and listen to some 2009 remixes of REM's Murmur and Reckoning and some Joy Division all in 24/192. I left him with it for a few minutes and when he handed it back he said he had never heard anything like it before. The sound totally drew him in, and only through cheap phones. And this is someone who has been listening to music seriously for over 40 years.

Then I gave my daughter my old behringer and put it between her laptop and her cheapish Creative speakers. I was not expecting to hear much if any difference because of the basic speakers and iPlayer source, but we both noticed it immediately. The sound was far better.

So the truth of the matter is that we were all had. CDs were first released in 1982, the same year as the Sinclair ZX Spectrum 48K computer which had to be booted from cassette tape recorder. So that is the level technology was at that time. The CD itself was an amazing invention for the time. But a DAC is a specialised type of computer, and the technology to build one which actually sounds reasonable has only been around for maybe 10 years maximum. In the 1980s computer chips were no where near capable of doing the complex job required of a DAC and so they were all a massive compromise.

If your CD player is more than a few years old, buy a new cheap DAC. It will be the cheapest and most effective HiFi upgrade you have ever done.

 

[BackToAnalogue]

I should also have added. Don't bother using your old CD player as a transport like the HiFi magazines tell you to. That is complete nonsense because you will be connected via an SPDIF interface. Hook the DAC up to a PC through USB, preferably Asynchronous, and then you can forget about jitter, or any other rubbish they come up with, forever. Job done. It works 100%, no issues theoretical or otherwise. Simples!
 
Oh and chuck your 20 year old CD player in the bin unless you can sell it on Ebay.

 

[stevekinghung]

Hi , I have very old Denon DCD-2560GL , Cambridge Audio 640C v2, I bought the Denon 20 years ago, but it still sounds the best , even I compared to my DACs (I have MDac, Dacmagic Plus and Matrix mini-i Pro) So maybe I have different opinion about this. I am still using Denon every weekend to enjoy my CDs
 
And also, I have AK100MK2 and Ibasso DX90, DX90 clarity is slightly worse than AK100MK2, but much worse than my IPOD Classic 160G. I have AK100MK2 coz size & able to play hi-res music and DX90 coz of its high gain power output. But overall sound quality, especially balance feeling, I think IPC still the best for me.
 
Mostly music files are ALAC and AIFF , which ripped by iTunes of my CDs. And I listen to pops, classics and jazz basically. 
 
I am sorry that I have totally different feeling of new and old equipment
^^

 

 

[duetta]

Since you asked, I have an old Bel Canto DAC1 that I've owned since the turn of the century, and I had the opportunity last year to compare it to an HRT Microstreamer that I picked on an impulse. 
 
To do as a fair comparison as possible, I connected the Bel Canto through a Trends Audio UD-10 USB to SPDIF converter that I had been given by a audiophile friend, fed FLAC files from a USB hard drive connected to a home PC running Audiophile Linux, using the Toslink connection that Bel Canto claimed the DAC1 preferred.  The other setup involved using the Microstreamer fed same FLAC files from a duplicate USB drive, through a Windows 7 laptop.  To make a long story short, after comparing a couple of different recordings, I concluded that I actually preferred the much older but also much more expensive Bel Canto - Trends Audio combo (which would have a list price around $1,500 back in the day) to the sub $200 HRT Microstreamer.  The Bel Canto sounded fuller, a bit richer, and with a slighter deeper soundstage.   In comparison, at least through my big rig (Apogee Duetta Signatures, Musical Fidelity A3CR preamp, Conrad Johnson MF-200 power), the miniscule HRT unit sounded a bit more threadbare, with a more compressed from front to back soundstage (especially on large scale orchestral repertoire).
 
Again, for $168 it wasn't bad at all, but I did find that this older if much more expensive combo still held its own - which frankly surprised me to no end. Given the passage of the years, I didn't expect that at all.  So the Bel Canto stayed in the big rig for now (until I pick up a legit USB DAC) with the HRT being relegated to laptop / PC duty on work computers. 
 
I'm very curious to hear (as least as far as my tinnitus allows) how much better a relatively affordable modern fuller-sized DAC will sound in comparison to the Bel Canto. 

 

[glitch39]

yes, these various chips have tremendously improved. there is no doubt about it.
 
however, for some reason, the older DACs sound more musical to me. warmer, less analytical, more analogue, etc......

 

[BackToAnalogue]

I was interested in the above post so did a bit of checking on the Bel Canto. It contains a PCM1704 DAC chip which is quite an old design but a very high end chip. It is an R2R design as opposed to the majority of modern chips which are Delta Sigma. And no I have no idea what the difference is yet but plan to find out. R2R designs are much more expensive to manufacture and the PCM1704 seems to be generally regarded as still being the best DAC chip made, except for mega expensive specialist chips. But it has been withdrawn now by TI and a lot of people aren't very happy about it.

So it isn't surprising that it sounds better because 'they don't make 'em like that anymore'. The Delta Sigma is still a very high end DAC and my comments about DACs being much better apply to the market for the 95% of consumers who can't afford supa-DACS. And to anyone using an old CD player except for a few top end models. There are some extremely cheap and very good DACS available now.

If I was lucky enough to own a Bel Canto I would certainly still be using it if it sounds as good as people say. And if you ever want to sell it I would love to hear what a PCM1704 DAC sounds like.

I am very new to this DAC game myself so my listening experience is limited to 2 modern DACs. My comments are from a technical perspective based upon a knowledge of the development of IT capabilities over the last 30 years. A DAC is fundamentally a mathematical engine, and the better the Maths is and the more powerful the engine, the more accurate the approximation of the analogue wave form it produces will be. Now that should also sound better because it is more true, more transparent, less coloured etc , but it may not. Or it may to some people and not others. But overall the quality of DACS should have improved enormously purely due to the much faster data processing capabilities we now have. And the difference between now and 20 years ago is enormous. But I bet there are still a lot of decent mid-range CD players from the mid 1990s still being used, and a £30 DAC would make a very noticeable and cheap upgrade.





I have also developed Tinnitus fairly recently and I have the 'Punk Wars' to thank for that and so I can't listen to music on headphones much now. A message to all you young uns out there who like to have the volume up at 245 on their latest DAP and Super Fi IEMs - learn to enjoy it at 210 - or you will end up with Tinnitus just like us 2 old guys. And then you won't be able to use them at all.

 

[gregjnjarvis]

I should also have added. Don't bother using your old CD player as a transport like the HiFi magazines tell you to. That is complete nonsense because you will be connected via an SPDIF interface. Hook the DAC up to a PC through USB, preferably Asynchronous, and then you can forget about jitter, or any other rubbish they come up with, forever. Job done. It works 100%, no issues theoretical or otherwise.

I.....I don't think this is quite how it works.

 

[BackToAnalogue]

Really, is it not, tell me why?

I'll be a bit more precise. You can forget about any jitter being introduced by the CD drive or the communications or receiver chips. A huge improvement on SPDIF.

There can still be jitter between the timing of the original sample and processor clock in the DAC but it is very small. And I am more than happy to be corrected if I have missed some engineering wrinkle to do with DAC implementation. I tend to think of them as CPUs and I realise that isn't quite right now.

There is also a but. Asynchronous communication requires much faster processing than synchronous. This is because all of the parts of the chain have to spend most of their time in 'idle' otherwise they aren't ready to receive the next packet of data because they don't know when it is going to arrive. If they spend half their time busy processing then delays very quickly start to stack up. So you need at least an order of magnitude increase in processor power and probably more (I am not an expert, I did this stuff 30 years ago). Thanks to the wonders of modern Mathematics we now have very fast, and to my generation of comms wonks unimaginably fast, data communications.

That is why implementing USB is a real pain in the backside for HiFi wonks like Jason at Schiit. Read his chapter on this. It very quickly made me realise that I was being a little hard on them :>)

It is probably also why a few DAC suppliers say that their DACs prefer Synch mode. They don't like to say that is because the design of their kit isn't quite fast enough for Asynch.

 

[sandab]

But a DAC is a specialised type of computer, and the technology to build one which actually sounds reasonable has only been around for maybe 10 years maximum. 

Today DACs are application-specific computers, but not because they couldn't be built before then since DSPs capable of doing everything a DAC does today have been around since at least the early 90s.  An AD SHARC DSP from the mid 90s would have zero difficulty doing any processing a modern DAC does. 
 
Instead, the BIG difference is abandoning AES3 (or the consumer SPDIF equivalent) in favor of USB.   AES3 was designed to be a digital connection to replace analog, and this means minimal propagation delay.  Clocking comes from the source - if the disc rotates a little faster, playback is a little faster; if it's slower, playback slows down.  This implies minimal dejittering, buffering, and processing.  If there's any processing it's in a separate filtering box with a predetermined propagation delay.  Without this isochronous streaming model it wouldn't be applicable to commercial applications as a replacement for analog cabling.  At a concert the speakers need to be lip synced to the performer, same when watching video.  Analog didn't introduce significant propagation delays, so AES3 was spec'ed to stay within similar limits to be a replacement.  Various implementations over the years have violated the streaming model in favor of improved quality, most notably buffering transports that read off the platter into a buffer and reclock the stream rather than relying on the rotation speed of the disk.  When the buffer gets low the drive speed is turned up a notch, when it's high the drive is slowed down.  Enough buffer is used to hold multiple rotations in case a complete rotation or two is required to recover sync.
 
The big recent difference is the adaptation of USB audio, especially asynchronous.  This is flow-controlled, packet switched, and the host simply keeps a buffer on the destination device within low- and high-water levels.  The device actively flow-controls based on this.  Playback and hence clock generation is a completely internal affair out of the sample buffer.  Even with external USB-to-SPDIF converters like the off-ramp it moves the clock generation out of a noisy computer or transport and into a dedicated box that only needs to do one thing really well.  These same attributes that make async USB excellent for hi-fi also makes it poor for studio applications like video mastering since it's impossible to lock the audio to the SMTPE master video clock - there's no telling what the DAC will do.  But it's really just horses for courses, and it's nice to see hi-fi audio get its own horse for a change.

 

[BackToAnalogue]

I agree with all of that and you explained it better than me.

You say DACs haven't really improved because of the superiority of the old type of DAC chip. But am I right that you are talking about an expensive studio product? I think if you look at mid price - affordable to a normal audio/music enthusiast who has a life- type equipment then there has been a big improvement. But I am no expert and I certainly haven't done any testing. I just think surely with the technological leap that has taken place in the last 10 years there must have been a big improvement in DACs. And if there hasn't been then a few engineers at Burr-Brown et al must have been sitting around with their fingers up their backsides.

When I first started reading about SPDIF a couple of months ago I was pretty amazed at the technology that was still being used. Even more amazed when I read in one of Jason's articles that SPDIF cables do not have separate clock pins, clocking is buried in the data. I can't even think how that works, I always thought you had to have clock signals wired in on RS232 for Synchronous file transfer. You can't do it with just twisted pair. Or that was the case a long time ago. It was interesting to find out the reason that it was used. It was also due to speed so I guess they just picked the nearest Synch protocol to hand.

What I don't get though is that everyone involved must have known that using SPDIF was a big compromise which caused jitter and really needed sorting out when the technology became affordable. So when the 16-bit parallel (see note) PC-Bus became commonplace and relatively cheap. Late 90s at a guess. Why did the audio industry not just create a cut down version of that and connect the CD drive to the DAC using a parallel interface? Surely that would work a treat? (At this point a hardware engineer somewhere will realise that I am a software guy and maybe explain why I am wrong?)

Sadly I suspect I know the answer to that one. But hopefully someone else will provide a different reason and my cynicism will go unrewarded.

For those not familiar with the term a parallel interface has 16 (or 32) wires and all of the bits arrive at the same time on the 16 wires. Only available in expensive mini computers in 1982 (the PDP-11s I was working on). But nowadays cheap as chips. Forgive the English pun lol.

 

[shortcuttomonct]

Really, is it not, tell me why?

I guess I meant that as a general rule, you cannot say:

Hook the DAC up to a PC through USB, preferably Asynchronous, and then you can forget about jitter, or any other rubbish they come up with, forever. Job done. It works 100%, no issues theoretical or otherwise.

The ongoing work of DAC manufacturers to increasingly lower jitter to some astonishing rate is evidence that "async USB" is simply not some magic format.  So is the fact that many DACs do seem to sound better when using SPDIF or other input besides USB - whether via a USB converter or otherwise.....
 
So yeah.  Just because it's a newer DAC and using async USB, I'm afraid does not mean it is automatically going to have no jitter or amazing sound quality compared to something else.

 

[BackToAnalogue]

Well it should be better if jitter over SPDIF is a real problem and I have explained why in some detail above and so has Sandab in his post. I have also explained why it is not straightforward. So I agree with you on that. And the DACs that prefer SPDIF do so because they aren't quite fast enough for Asynchronous. I am sure they sound absolutely fine using SPDIF because that is what they were designed for.