Bottlehead Forum
Other Gear => Digital => Topic started by: kgoss on August 23, 2013, 03:31:34 PM
-
I just saw this interesting article http://www.audiostream.com/content/qa-john-swenson-part-1-what-digital (http://www.audiostream.com/content/qa-john-swenson-part-1-what-digital) and was wondering if this is a Q&A session with the same John Swenson who is working on the Bottlehead DAC project.
Ken
-
Is John involved with the Bottlehead DAC? Now THAT is interesting!
-
Great read (the little I could understand...).
John
-
Found it kind of interesting - but it only addresses the "what is digital" part. That's the simple part of the system (its all '0' or '1', how they're defined & some of the timing considerations (jitter)). What's more fun is looking at what changes the infinitely varying analog voltages into patterns of 0's and 1's (and retains the information) - as well as what does the inverse at the other end. Even more fun is when you get into perceptual coding (doing lossy compression - but only losing what people can't hear).
Things like jitter and noise are well known and can successfully be designed around (people have been doing this in instrumentation systems for ages - and many of these have orders of magnitude more sensitivity than audio systems). Sometimes I wonder if some of the effects people experience when rolling cables in digital interconnects are simply because of inadequate designs in the equipment due to cost constraints (or sometimes ignorance)...
BTW: My current occupation is in digital television - deep into the plumbing of the system and next generation DTV technology. I'm finding tube audio to be quite a refreshing change from that when I get home.
Rich
-
Yes its me, this is just the first installment. This all started with some threads on other forums with some people saying they can hear things and other people say it's theoretically impossible, so this series is about going into some details with real implementation issues as to what actually happens in real hardware.
The over all attempt is to at least get out there what can actually happen, whether it is audible or not is a different issue.
Remember this is just the first installment, laying the groundwork for what is to come.
John S.
-
John: I'm looking forward to reading the upcoming articles. Its nice to see something based on real engineering, rather than guesses and snake oil.
Rich
-
Thanks for the confirmation John. I'm looking forward to reading the upcoming articles as well as the Bottlehead DAC when it is released!
Ken
-
I figured the Doc et. al. would "consult" with well known minds for a digital kit. I'm impressed....
-
I thought you guys already knew this?!
We met John at the last VSAC in Vancouver WA. Some time later, Doc kept saying to me that he wanted to do a DAC. I kept saying I know tubes, I even know digital signal processing and information theory - but you need a digital audio person for this. Doc finally put two and two together ...
Maybe Doc will tell his version of the story - it will have happened on another planet of course. Memory is like that.
-
Actually it was Colonel Mustard in the Drawing Room with the Candlestick.
John has been putting up great posts about this stuff on various forums for many years now and I knew he was a brother in tubes. So I always read his posts when I found them, and I was particularly intrigued by a thread that was on a rather obscure digital forum in which the topic was using the PCM1704 DAC chip - at the time that chip that was included on most digital guys' "best DAC chips" list. John held his own handily among some well known guys in the DAC business and proposed some really outside the box type ideas there. We began discussing them together not long after that. Certainly VSAC 2008 was one of the times and quite possibly the time we actually said "let's work together". PJ hasn't lost his memory yet.
-
First, I knew who John was to start with.
Yes its me, this is just the first installment. This all started with some threads on other forums with some people saying they can hear things and other people say it's theoretically impossible, . . . .
John,
It is interesting to me (BA in Psychology, BET in EE) that some people don't realize that hearing and perception is an individual thing. I have friends who can't hear anything behind the speakers. Their vision dominates their hearing. It is a natural occurrence.
What I learned in a Perception class was that 20-20k Hz was ear hearing. We "sense" much more.
So, "whether it is audible or not is a different issue," is a matter of perception not hearing. That isn't semantics.
-
Grainger is right on the money here... Our ability to perceive music, like most of our abilities spans a huge range and pretty much follows a standard Gaussian distribution. I like to call it our MQ - or Musicality Quotient and just as a short list is made up of various abilities such as the ability to percieve rythm, tonality, pitch differential as well as other perceptive and cognitive parameters. Then there is the physiological aspects of individual ears, shape of ear canal, condition of cochlear sensing hairs, and a zillion other things, not even touching on the neurological issues, which also span a large range in humans.
Most people fall into the 3rd standard deviation, some of us are in the 5th and others in the 1st. Note this has nothing to do with musical enjoyment -- if you like what you hear, then nobody can question you on that, but it is totally absurd to tell somebody else what they can and can't hear, whether it be digital artifacts, differences in cables, dacs, speaker materials, etc. If somebody perceives those things then nobody has the right to call them snake oil or bullshit, etc., it just means that by luck of the draw and possiibly other influences over one's lifespan that their MQ is just higher and like intelligence quotient, they are working in a different zone. This stuff is not news to cognitive neuroscientists and can apply to every aspect of life that people perceive, it's not psuedoscience, it's hard science with a huge body of work to support it. I know people who cannot tell any difference between a car radio and a $100k system -- they simply cannot percieve the things that set them apart. Then there's another fact of life that some people are more visual, some more auditory, others more haptic, etc. and these things just represent a resultant vector of the various "Q"s. So yes, interestingly enough hearing and sound perception is not a uniform trait in human beings andd that all people have the same hearing abilities and the same perceptive and cognitive responses to those.
Also in my case as an adventitiously blinded person, it is also well known and documented that the previously active visual cortex is still plastic enough as an adult to be co-opted by the auditory processing system and why I may be more sensitive to these things than most people -- I simply have more processing power dedicated to auditory input as well as tthe converse, less dedicated to visual processing, which is normally 95 to 98% of our waking cognitive load. I didn't ask for this and I did nothing special to train myself, it's just the natural response of the brain and body to the loss of eyesight.
I will read John's article with great interest, as I did Gordon Rankin's but I'm certainly not going to engage in some of the ridiculous back and forth in the comments, especially those from people that have zero background in this stuff and are only parroting what they read on other forums by other armchair digital experts.
Bottom line is that everybody is right -- for them -- if they can't hear a difference between usb cables, then so be it, consider yourself lucky and stick with a $5 belken cable, but don't call it snake oil because there are people out here who can easily tell the differences between usb cables, as only one example.
-- Jim
-
Jim,
Eloquently said!
-
I agree, Jim. I'd add that one of the influences of "perception" is expectation bias. One truly believes they hear a difference because they expect to hear a difference whether or not a difference can be in fact measured. False memory is another example.
-
There are a couple other biases that go way back in the literature. One is that you like what sounds similar to what you are used to. Another is that if you are not used to a particular sonic, it sounds more real until you are able to recognize it - hence Edison's very early recordings truly sounded like the real thing at the time; today they are interesting artifacts which sound like crap. Notice that these two effects seem mutually exclusive - welcome to the strange world of psychological "science"!!!
-
Yes, definitely, but again, the level and type of expectation bias is also highly variable in the population. Optimist or pessimist? Also, and related to what Paul said, it tends to weaken or change over time and with familiarization. I've often seen people say that they expected to hear no difference but did, as well as the converse, but the first seems much more prevelant, at least in this hobby/pursuit. I always get a kick out of it when somebody tells me you hear a difference because you're expecting to, not realizing that I can say with exactly the same validity back to them that they hear no difference because they are not expecting to. Yes, it is astrange but IMO fascinating, and ultimately very human (with all it's faults and foibles) to look into any kind of pursuit where perception dominates over logic and measurability. And for me, just one more reason I am so fascinated with the human brain.
-- Jim
-
THIS, is a very interesting thread for me. I always knew that I was 'different' in so many ways, but hearing stood out as one of the most prominent differences. I was the kid that entered a room with a question, 'what's that sound I'm hearing?' Others could hear it, but only if the 'chose' not to ignore it. For me it was always the opposite. Then there is, 'it just sounds right'. what can I tell you? I don't have an explanation for it, and others just think that I make it up, but of course, we all know that's one of the differences between us and them. I'll not go further, for the moment, but will continue to read what others have to say that surprises me, meaning, it's not just me?
-
I find that no matter where I am, if there is music playing I hear it and recognize songs that I know.
In the 70s I was accused of having "Servo ears" because when the Servo-Statics were acting up I knew it before anyone else. Mainly because I had to be on top of a very unreliable speaker's quirks.
-
'G', Funny you should mention that about hearing trouble with equipment. In the old central offices, there a variety of solenoids clicking and clacking. Some of them were in a marching band kind drummer type of sequence. I could barely walk by them without stopping to hear the beat. If a solenoid was stuck or chattering, I was usually first on scene. And, don't you hear high frequencies in some places for no apparent reason? Makes you wonder.
-
My wife was an "Outside Plant Engineer" for South Central Bell when we met. She ended up in management with AT&T, till if fell apart.
-
O' Lord how things changed from the original Bell System.
I see false memory mentioned earlier in this thread. Man, that is one of the craziest games that our brains play on us! What's the deal with that? And it happens across the board. How big things seemed when we were young, or how amazing this or that sounded. Then you go back and the memory may be completely distorted. Selective perhaps, 'error correction', I really don't understand all that mental drama and it's purpose.
-
It's like a "desensitization process"..part of our adaptive abilities "survival"... Everything is well said, and very interesting.... I will add 2 very basic and yet powerful forces in shaping our already fine hearing abilities.... 1) is Body Chemistry, what we are , or have been lately, putting in our bodies ie. nutrients or lack thereof, too much coffee, booze, on a bad diet kick, lack of exercise, fatigue and on and on... (I am sure Jim already included these in the above Physiology category)... 2) the often overlooked and never mentioned amount of Earwax a person CAN have.... I for one, usually have to go to an audiologist from time to time and get it removed.. When, I leave it for 2 years or so, My ear drum is completely covered over.. I can hear very well, but, the funny thing is, just how much it robs my hearing of the lower midrange and makes me much more sensitive to high frequencies.. Many things sound a tad brighter to me and more annoying.. When, I came back from my appointment the first time, I played my system, expecting maybe nothing but a little clearer sound, but, BOY, the FULL, rich midrange that was filling the room that wasnt there before...I thought my speaks were just a little weak in that area, but NO, it was my ears....I never hear people discussing this, I just assume it must be me only..However, it surely doesnt hurt to make an appointment, If you dont do it already to an audiologist and make sure your lower ear canal is pretty wax free.... Maybe, this is a dumb post from me, but I am just making sure if anybody hasnt thought of this and has relatively plugged ears and doesnt know it... It is a CHEAP and simple upgrade to your TOTAL audio system experience.... just sayin..
-
A cheap and often amazing upgrade, I agree. I tend to lose top end when my ears have been waxy. Luckily it doesn't happen that often, every few years. But taking care of it is like spending $10k on great new gear.
-
Yep, I wonder how much gear people have gotten rid of due to this. When it maybe wasnt their gear at all.. Or rejected a perfectly nice sounding product due to this also...haha
-
I just read this article and it's great!
Question: does a USB cable experience the same ramp up mentioned in the article and do different cables therefore have the potential to introduce jitter?
-
One thing that you need to keep in mind concerning USB cables is that many of the "audible" changes attributed to them are highly unlikely. Most issues like jitter and other timing things are easily taken care of by simple buffering - a long solved problem, with an easy solution. Bit errors are either corrected - or detected and result in loss of packets (if not correctible). Loss of packets results in major audio glitches - pops, dropouts and so on.
More subtle things like changes in the "character" of the music or changes in midrange level (for example) will NOT be caused by cables - as they would have to result from actual changes in the digital "numbers" carried within the bitstream - and any such changes would be detected as errors by the checksum (or CRC) algorithm that is used over the entire packet. For a cable to change (for example) the midrange character of a signal, it would have to figure out which numbers are carrying that information, make the changes and then recalculate the checksums to make the packet valid again - which is a pretty tall order for a simple piece of wire.
Rich
-
I understand that in theory, but my experience differs and goes against placebo in this situation. I expected no change based on the same theory you just explained, but my results were significant, repeatable and undeniable.
The only explanation I can fathom is that ramp up or some similar trait must be involved.
-
This may well be a case where human and psycho-acoustic factors come into play (expectation rather than experience) as the science/engineering is pretty clear cut and very simple. However, if someone does have a technical explanation for other effects (with some data to back it up), I'm certainly open to learning more.
BTW: I think this applies to more than just USB - most digital audio connections don't simply pour bits over the "wire" - rather there's a packetized structure involved that does have error checking (and an establised timing/buffer model).
I come from the Digital Television world, mostly inventing and standardizing new transport structures...
Rich
-
This may well be a case where human and psycho-acoustic factors come into play (expectation rather than experience)
That can't be the case when the expectation goes directly against the experience. Had I expected improvement, I would question the positive experience, but expecting no change and then hearing a definite change (for better or worse) suggests that something is going on. My question then is not "Did I hear a change?", but "Why did I hear a change?"
-
It could be simply that the cable carries EMI that is injected into the signal ground of the DAC, and that carries over to the analog stage ground.
-
Are there any standard issues that this type of EMI would create? What would it do to the or could it vary dramatically?
-
Very interesting article.
When I was younger, I remember being so annoyed whenever someone would leave the TV on in another room. The CRT whine felt like it was penetrating my skull. It took me a good while to figure out that most people could rarely, if ever hear that noise when a TV was running properly. I assumed it was normal and everyone tolerated it :)
This may well be a case where human and psycho-acoustic factors come into play (expectation rather than experience) as the science/engineering is pretty clear cut and very simple. However, if someone does have a technical explanation for other effects (with some data to back it up), I'm certainly open to learning more.
BTW: I think this applies to more than just USB - most digital audio connections don't simply pour bits over the "wire" - rather there's a packetized structure involved that does have error checking (and an establised timing/buffer model).
I come from the Digital Television world, mostly inventing and standardizing new transport structures...
Rich
As an example, how come a faulty HDMI cable can cause video artefacts resembling random fuzzy picture noise (which going on appearance alone doesn't look like lost packets/data, but seems overlaid on what is functional video). The signal is there, just with added snow:) I've observed this on a Chinese cable which was beyond spec and far too long.
That's the only reason I wouldn't totally dismiss claims of one USB cable being "better" than another, even though it seems like the end results should be far more 0 and 1.