High Resolution Analog
I finished watching the 18-minute rollout video of PS Audio’s DirectStream DAC hosted by Paul McGowan yesterday. It was well produced. There were multiple cameras, the lighting was reasonable and the audio good easy to hear. But the message was nothing but spin.
My favorite part was when they played a DVD of an interview with Gus Skinas from the Boulder Super Audio Center. Gus has been a very strong advocate for the DSD format and for the SONY Sonoma recording system since the beginning when SONY gave him an entire studio and set him up to handle DSD work. I know Gus and have a great deal of respect for his work. It was Gus that recommended iTrax be included in the SONY HRA rollout last fall.
Figure 1 – Gus Skinas raving about the sound of PCM and the PS Audio “subtitle”.
During the interview Gus keeps praising the amazing sound of some recordings that were produced using high-resolution PCM. And the good folks at PS Audio cleverly put subtitles over the top of him stating, “he was actually listening to DSD”. Every time Gus said anything positive about PCM, the subtitles would show up again to make sure that you follow the company line. Gus is a very smart guy…I don’t think he would have a problem knowing when to say DSD rather than PCM. And if they really wanted Gus to give a testimonial…they should have shot it again. After all they live and work in the same city.
In watching Paul talk about the merits of their new DirectStream DAC, it seems pretty obvious that his unbridled enthusiasm support for the DSD format comes from his unfounded and incorrect belief that it is really “high resolution analog”. It’s not one of those evil old digital formats. He called it “high-resolution analog”. I’d never heard this wording before…I wrote it down and decided to think about what it means.
I guess we need to go back and think about analog for a minute. I’m not talking about analog recording but just the contrast between analog and digital. I usually define analog as a “continuously variable” representation of something. Analog signals are “non-quantized variations in frequency and amplitude” according to Wikipedia. It could be the motion of pendulum, the movement of an old school speedometer or the changing pressure in air molecules that we perceive as sound. In its electronic form, analog audio is the alternating current (the moving electrons) traversing back and forth in our cables and through our circuits.
If you look up resolution on Wikipedia the first line states that it is a “measure of digital audio quality”. In the digital world, resolution is the amount of discrete pieces of information after a continuous signal has been sampled or “quantized”.
So how can something exist in both worlds at the same time? It can’t. Saying something is “high-resolution analog” means nothing. It’s just another marketing phrase meant to get audiophiles to buy into their spin machine. The messaging on the DirectStream DAC is full of similar hokum. Just because they say CDs sounded “less musical than analog”, doesn’t make it true. It’s a matter of personal taste and nothing more.
Music is analog, microphones are analog and if you send the outputs of professional mikes through a mixing console to a set of amplifiers and speakers, the output will be analog. If at any stage you slice the analog signal into 44,100 or 2,8224,000 slices (CD sample rate or DSD 64), calling it “analog” is just incorrect. Paul and people at PS Audio know better and they should quit trying to elevate a format that the inventors of that very format called “a mistake” into the next big thing in high-end audio.
I have no problem with people enjoying the music they love in whatever format works for them. You can purchase all of the analog tapes, SACDs, DSD downloads, CDs, HD-Audio in PCM or vinyl LPs you want…but subjecting every input type to DSD is pretty restrictive.
Other companies building high-end DACs are doing BOTH high-resolution PCM and DSD conversion in their native formats (like my Benchmark DAC2). That seems to me to be a much better approach. I would have no interest at all in an over expensive “my-way-or-the-highway” piece of equipment.
Contrary to PS Audio and Paul’s insistence, DSD is NOT a better format in any way. It’s NOT going to deliver “hidden musical details” from your CDs or high-resolution downloads. And DSD and the DirectStream DAC are not the “answer”.
The last paragraph from the product’s webpage is very telling:
“There are millions of PCM based recordings that will be in our libraries for years to come, but with the introduction of DirectStream, you no longer need listen to them with a PCM based processor.”
Of course, they will sound as they possibly can if you listen to them the way they were intended…in PCM. For those that love the sound of analog tape, you could take the output of your CD player and record it on a piece of analog tape and immediately play it back from the repro head…thus getting that “special” analog tape sound. Come on!
I don’t need or want the PS Audio “DSD police” forcing me to “hear it their way”. I’ll pass on the DirectStream DAC and stick with equipment that gives me a choice. And I’ll save a bunch of money, too.
This one requires more attention, Mark. You are the two people whose opinion I value the most. If the two of you disagree on the science or on the results of what Paul’s new DAC is doing I will be irrevocably confused.
This is Paul McGowan making these claims about a serious piece of equiptment, not a snake oil salesman selling a tweak. When it comes to power, Paul IS the final word. His PSAudio Power conditioner is the best one that has ever been made. His transport and PerfectWave DAC are far better than anything else I have ever heard at any price. When Paul McGowan speaks, he cannot be dismissed this simply. I would urge you to call him and listen to this thing yourself.
Paul needs to get that prototype DAC to your lab ASAP, and you need to give it a serious examination followed by discussion with Paul BEFORE the Chicago show next month. What you have to say about this device will be very important to a lot of people.
When it comes to you and Paul, mistakes in a video or with terminology in marketing material cannot prevent a serious study of this device. There are a lot of people like me who really want to know whether Paul has invented something special or not.
Let me not fail to say please, Mark.
Donald…thanks for the comment. I’m with you. I have owned a few pieces from PS Audio and have a great deal of respect for his technical expertise and his drive to build the best possible equipment. And I must admit, it would be fun to check one of my recording through his new box and the original. Would the “DSDized” version have the ultrasonic noise the is unavoidable with the format or would it be clean above 20 kHz like the source?
There is a strong motivation on PS Audio’s part to move in a unique and innovative direction…the DirectStream DAC is both of those things. And if he had done a great job of converting HD PCM to analog like the previous converters AND added the option to direct the signals through the DSD whiz bang…then everybody wins. But I don’t like being told that there is only one way to do things.
The reality I don’t need to listen to the new unit. The sound that I get out of my current DACs is fabulous, less expensive and imparts no “color” to the sound. That’s what I have and want I want. In Chicago, I’ll play lots of really special recordings and people can listen for themselves.
I suppose the proof will be in the pudding. If the new unit outperforms the PWD II by $2000 worth, then whatever misconceived engineering is used will have to be accepted. I understand your points about imprecise terminology or incorrect explanations but the final product is what matters most to the manufacturer and to the consumer. Having said that, you still owe it to your readers to help us by pointing out errors that we are not equipped to recognize. We can use this information to help us make better choices.
Alan…we’re going to get back to personal taste again. If you or any audio listener wants what DSD does to the sound of a pristine recording, then no one can argue with you. I prefer to have the sound of pristine, which was my intent when I made the recordings that I release. You have the option to adjust the EQ or filter my sound anyway that brings you maximum sonic pleasure. PS Audio has taken a very aggressive stand in favor of a technology that I find seriously flawed for the reasons that I’ve written about numerous times in these posts.
I don’t believe there’s any processor or filter that could improve on what I regard as finished. That’s my preference…plain and simple.
I do hope to be able to hear the DirectStream DAC at some point. Paul has been gracious enough to invite me to do just that in Boulder at some point. I have no doubt that it will sound terrific…but I don’t believe it will reveal “subtle details” that I’ve missed when listening through my Benchmark DAC2 or the Bryston SP3.
Excellent analysis by Dr. AIX Mark Waldrep on the mis-information put out by manufacturers.
Indeed, one of the reasons I stopped buying Classical LPs in the 1980s and switched to buying Classical CDs instead was those annoying little “Digitally Recorded” stickers that the record companies used to put on the LP sleeves of their new releases. The suggestion that upgrading the original recording technology would improve on a technically compromised LP sound was, to put it mildly, reptile lubricant.
Gary…think of it as marketing spin. We all do it. I don’t think Paul was pushing out “misinformation” but rather spinning a marketing message that would resonate with high potential customers.
agreed. PSA is pretty good at piling on the hyperbole, so it’s not surprising that Paul’s video is full of fluff.
I understand that marketing messages have to be over the top and I don’t think PS Audio is any more guilty than any other company. But there should be a “mythbuster” or “reality checker” out there to steer things back on track.
Dr. AIX,
I surely have to agree with you 100%. To me and I believe to many others it seems like a paradox to say that you don’t have to listen to the 1000’s of PCM recordings they have in their original format, but that you can enjoy the same(sic) recordings as DSD files after a conversion. Sounds to me like the folks that want to watch a HD digital filmed movie as a 35 mm print, to have that special film look.
How can something get better by degrading the quality?
That’s exactly the point that I was trying to make with my mention of the recording and reproduction of a clean track through an analog tape machine. It absolutely will produce a change in the sound. Whether you consider that a degradation or improvement is up to the listener. As far as delivering something that wasn’t there in the original…it won’t. Will it give you something that you feel is “better”…maybe. But please let me decide.
I do find myself shaking my head at the level of, to be polite, “marketing speak” in some of PS Audio’s literature. For example, this for the Powerwave AC3 cable:
“Inside the AC3 is one OFC hollow conductor for the treble region, a large OFC rectangular conductor for the midrange and multiple gauges of OFC bundled together for the bass. Equipment powered with an AC3 enjoy a significant improvement in sound quality over any stock power cable.”
Leaving aside the argument whether conductor geometry has a significant effect at audio frequencies, I can see the sense in designing a cable to carry the full audio range. Except that this is an AC power cable, with one job: To carry 50 / 60 Hz signal. Higher frequencies are undesirable and unwelcome. Doing a better job of carrying noise and interference into one’s equipment will potentially cause it to sound worse, not better.
(I do note that while in this case the science may be questionable, the price appears reasonable given the apparent build quality and the construction complexity.)
You really don’t want to get me started on power cords. I’ve wired up a lot of studios in my career and the quality of the cable behind the duplex outlet mounted to the wall is the limited factor. You don’t get “better” sound in the last 6 feet…and $500! If people purchase this nonsense then they will love the “telephone tweak”…argh!
Thanks Mark, appreciate you watching the video and taking the time to comment. I am a great admirer of your work. You often come to our forums and post and I hope you don’t mind me doing the same as several folks have asked.
Just to be clear, I absolutely do believe (as you know) that DSD is a superior format to PCM. And yes, I believe that if Sony and Philips had introduced it rather than PCM way back when, we’d have made much more progress towards recreating music properly than we have.
All that said PCM, as it has evolved, is a great format (finally) and even a lowly CD can sound incredible on a good system. There are many good systems. I think the crux of my argument is that using a PCM based DAC, one that is most likely built on an off the shelf processor, can wind up masking information such as subtle details of room cues and overtones. There are multiple examples of this easily identified, including our own PerfectWave DAC – one of the better ones out there.
When you move away from this complex architecture you get to hear what’s been missing. I would invite anyone so interested to come by PS Audio in Boulder for a demo if you’re interested. Like the folks at CAS, like Gus, this is something easily demonstrated.
Your terrific recording are on PCM. That’s fine and I haven’t anything but praise for them. And yes, played back on a proper DSD based DAC like the new DirectStream, you’ll hear all that’s on those recordings. 🙂
Bits are bits, but how those bits are handled and processed makes all the difference in the world.
I would encourage you to give it a listen.
And to Gary Reber, please send me an email explaining what you mean by misinformation. I’d be curious to know what you believe is wrong with the information presented.
Paul, thanks for taking the time to write…you are, of course, very welcome to my RealHD-Audio.com site. Your daily posts and even the theme of this site were the inspiration for my own efforts at blogging. I enjoy reading your posts on occasion and as you know I am a PS Audio owner.
I’d love to listen to my recordings through your latest hardware. Obviously, I do believe that I’m hearing everything on those recording already but I’m always open to new things. My arguments against DSD are well founded…I’ve spoken to a lot of smart people, done a lot of analysis and listening and participated in a “shoot out” in Whitefish Lake about a year ago with a live recording done on analog tape, Sonoma (Gus’s platform) and HD PCM. My friend Peter McGrath and the others present like both the DSD 64 and the PCM…we didn’t like the tape. It was my subsequent analysis of the files that causes me some dismay. There is all of the high frequency noise…intellectually and sonically, I think it’s a bad thing.
My sister and brother-in-law live in Boulder, so I do get there often enough. I’ll definitely make a plan to contact you and stop by the next time I’m in town.
I don’t know what Gary Reber was referring to. I felt that your marketing spin was a little over the top…and I’m confused about the “high resolution analog” terminology but I’ll leave it to Gary to respond.
There is a ‘simpler is better’ philosophy issue here, isn’t there? Namely, that the DA conversion in normal PCM has an inherent complexity that ‘cannot possibly’ be sonically flawless, and that DSD having a simpler DA circuit is irreducibly advantaged.
If the first half of that sentence is wrong, and it is, then the second half is also wrong.
The only advantages of DAC simplicity are price and reliability. McGowan of course doesn’t want to go there.
There is elegance to simple things…but that doesn’t necessarily mean that complex things are inherently incapable of quality. DSD is potentially simpler in the chip count, the kinds of filtering and the avoidance of decimation, but the world of production is NOT! Virtually all recordings are made using very complex system in both analog and digital. The end result is what matters…but it’s nice to know what’s behind the final output.
Well, Paul is exaggerating, but he isn’t totally wrong.
PCM works on samples and the resulting 0’s and 1’s, are computer code. The numbers have to be reconstructed with a decimation filter to produce sound at all. DSD (PWD) comes from a totally different process, the result actually “looks” like a waveform, and it can be played through analog equipment and music will result. Not good sound, but music. There is no decimation/reconstruction involved. Essentially, all that is needed to get good sound is a fliter to remove some high frequency noise.
Since a PWD signal can be played back directly and produce sound, it is not entirely incorrect to regard it as analog. That’s where Paul is coming from. His “high resoultion” claim comes from the high rate at which DSD signal is produced, which at least 64 times the rate of Redbook.
In the end, I agree that all that matters is how his unit sounds. If it is a really good sounding DAC, then he will have some basis for his claims.
Sorry, meant PWM signal, made much the same way as DSD. You can see what Paul has written about it here:
http://www.pstracks.com/pauls-posts/analog-coding/9174/
http://www.pstracks.com/pauls-posts/9181/9181/
Danny, I’m not with you when you say that PWD is in any way analog…at least as I define analog. It might be able to output something through an amplifier and speaker, but that doesn’t mean that the source was not sliced and stored as “discrete” pieces of information. Even if those pieces of information are relative (1-bit) or absolute (multibit).
It’s actually a grey area, at least for PWM. If a measurable parameter of the output signal varies as a close analogue of the air pressure at the microphone’s diaphragm, then it cannot be denied that the output is analog (it is an analogue). In the case of PWM, that parameter is ‘duration of voltage’.
As Bruno Putzys puts in when discussing the PWM stage of a Class D amplifier, “Although superficially the voltage waveform looks like a “digital signal,” the current waveform does not. The current waveform is in fact a pair of sinusoids with a high-frequency triangle superimposed on it. It stands to reason that the myth of the “digital amplifier” would have never taken off if it had been as common to show current waveforms as it is to view voltage waveforms.”
On the other hand, there is no reason at all to believe that there is any superiority to be associated with presenting binary information in an pattern that is analogous of the input air pressure wave. I am only discussing whether the label ‘analog’ is reasonable for PWM, not whether that is a better thing.
I’m not ready to concede on this point. Just because you can plug the output of a digital format to an amplifier doesn’t magically make it analog. I get your point and agree that there is an “analogue” of the input voltage fluctuations in the PWM signal. But there is a lot of “processing” in the digital domain that shifts the noise or filters the HF noise.
“I usually define analog as a “continuously variable” representation of something. Analog signals are “non-quantized variations in frequency and amplitude” according to Wikipedia.”
Something is digital if it has discrete values: it need not necessarily be binary in form. If the subatomic world is quantised, then everything at a larger scale is also quantised. This makes the entire universe digital in nature! When people use the term “analogue”, they are referring to phenomena that have a resolution and dynamic range that only ‘appears’ continuous.
Gordon, I guess you could view the world as being completely digital…but I think that’s stretching things a bit. I agree that quantizing something doesn’t have to result in the storage of the sample values as binary information.
I’m contrarian by nature. 🙂 I’ve never liked my professors distinction between the two. I see analogue as a bulk handling of some quantity. It is a reflection of the limitations of our technology, and also limitations in measurement placed upon us by quantum mechanics. Instead of precisely controlling electrons through a transistor, a typical analogue transistor circuit ‘corals’ large quantities through, like a farmer herding sheep into a pen.
The link in your latest email worked. Most convenient, thanks!
Gordon…I think by now you might know that I’m not one to go with the flow either. There is merit to your skepticism regarding analog vs. digital. But I fear that we’re going beyond what the traditional meanings are with regards to audio and music recording by getting into the quantum mechanics level.
BTW A link in the email, that takes the reader directly to the post on the realhd-audio website, would be most helpful. It would make commenting much easier!
I did try that…but it hung up the emailing system. I’ll get it to work.
Well, I am an engineer but not in the areas that you and Paul are. At the risk of being overly simplistic, PS Audio believes that the inherent error associated with converting PCM data to DSD data is overcome by the superior performance of their DSD DAC and analog circuitry (I think Paul does that very well). That said, it is really just a matter of how the listener perceives quality. All of the other stuff is just marketing hype which has certainly gotten our attention!
Blaine…Paul and the PS Audio team are not trying to overcome the “inherent error associated with converting PCM data to DSD data”. Their DirectStream DAC (which they label on the back of the box as a “processor”…I noticed this morning) takes the PCM input at any sampling rate and “upconverts” it as many as 4 times until they can output a very high sample rate DSD version. I don’t see that as advantageous when I can convert back to analog from my 96 kHz/24-bit PCM data without any of those upconversions and processing. Maybe this is simpler. I get exactly what I recorded back through my DACs.
You’re right. In the end it’s what’s you prefer and I know what I like.
Unless you’re using a ladder DAC, that 24-bit 96 kHz PCM data is being converted to a different digital representation, before the conversion to analog. The ESS Sabre used in the Benchmark DAC 2 has a very complex algorithm.
http://www.stereophile.com/ces2008/011008ess/
Having studied physics and electrical engineering, I do agree with Gordon’s statement that the entire universe is digital (discrete) in nature, but it’s not necessary to go to the quantum mechanics level to see that hearing is digital in nature. Your ears send discrete digital streams of data with each stream representing a specific frequency and the digital data in the stream representing intensity information to your brain, where that digital data is then analyzed and finally interpreted as language, music, etc. The Nyquist Theorem works, not just for PCM and DSD, but for your ears and brain.