As per Slarti's request, audio stuff.
Three transoceanic relocations have pretty much knocked the stuffing out of my audiophile desires, but hearing about it brings back lots of memories.
Tubular! (amps that is)
Feb 29, 2008 by libjpnComments
Feb 29, 2008, 23:09:30 cleek wrote:
for guitar amps, nothing touches tubes. but, to get things really working, you need VOLUME. and that's a problem, in a suburban subdivision.
on the other hand (40 pounds in one, 18 ounces in the other), my notebook sized digital multitrack recorder can simulate a music store full of different amps - directly into a pair of headphones. and it simulates them well enough that the big black Fender amp in the corner is only used when i feel guilty about [i]not[/i] using it.
Mar 01, 2008, 00:01:08 Slartibartfast wrote:
Tube amps have this cool microphonic effect, where ambient acoustics start moving the tube elements around and affecting gain, I'd expect. I don't know that I've actually <i>heard</i> the microphonic effect and recognized it, but that's the theory.
I think there were some comments about soft clipping, too, on the other thread, and I think that applies to tubes. You can use soft clipping as a distortion effect, where on transistor amps it would tend to hard clip. Transistor amps give you gain right up to the power supply voltage, and so what happens is the output waveform gets, roughly speaking, sheared off at close to the rail voltage. The consequence of hard clipping is that all sorts of high frequencies are generated as a result; if you consider the spectral content of a square wave, for instance, as compared with a sinusoid you begin to see the picture.
Hard clipping is what tends to blow out tweeters, I think.
Mar 01, 2008, 01:27:35 Turbulence wrote:
Slarti, what is soft clipping?
Mar 01, 2008, 02:52:35 nous wrote:
Turbulence:
http://www.trueaudio.com/at...
Soft clipping produces a rounder wave form where hard clipping produces a square wave form when driven above the threshold amplitude.
Not sure what to think about the arguments about even versus odd harmonics in Fourier transformations.
Mar 01, 2008, 04:58:06 ferrydust wrote:
Only going to say hi for now, and thanks Slartibart for the guidance. (It’s felix that was, under a new nom de keyboard) It’s difficult being a happy fault while confined to nine letters. All the really cool people go around reinventing themselves, right?
Be back soon.
Thanks again.
Mar 01, 2008, 06:11:18 Slartibartfast wrote:
That article looks about right to me. I'd like to see why tubes clip gradually; I'd guess they lose gain near rail voltage, but _why_ they lose gain would be interesting to know.
NAD supposedly has soft clipping on their amplifier design, but it's, as far as I can tell, designed-in, and not intrinsic to transistor amplifiers in general.
Mar 01, 2008, 11:35:21 Turbulence wrote:
Hi Felix!
I hope I didn't offend you too much with my last post at the mothership; looking back over it, my reply looks a might bit cranky, or at least crankier than you may have deserved. Anyway, I'll reply to your last comment here because, well, I feel like it.
[i]I suppose I’ve swallowed the idea that the best thinking comes outside the box, and the smaller engineering teams are the most innovative. I’m sure you’ve encountered that meme many times.[/i]
I'm not sure that outside-the-box thinking is always best, but I do think you have a good point with regards to small teams being more innovative. In fact, I'd go further and say that small teams are more effective.
[i]Is it useful to distinguish between positive and negative feedback? A widespread conviction exists in the audiophile conversation that zero negative feedback is a defining ideal.[/i]
I don't think so. Positive feedback systems tend to either explode up or explode down. I can imagine they might have subsidiary uses in a large amplifier, but they wouldn't have a place on the primary signal path where the frequency response mattered.
[i]While I’m allowed to continue wandering in obscurities, some of the most interesting designs, at price levels ranging from (relatively) inexpensive to exorbitant employ rechargeable batteries. A good example here. They by report always produce superior sound, presumably due to purity of the power source relative to rectified normal (polluted) electrical service. Especially given its green side, I can see it becoming more widespread.[/i]
Your link wasn't working so I couldn't see the page, but I'm skeptical. Filtering out the problematic effects of line current is something we're quite good at. It might be the case that cheap amplifiers have crummy isolation circuits, but in principle, we know how to solve this problem.
Battery based systems are usually bad for the environment. You need to spend at least as much energy charging the batteries as you would spend powering your amplifier from the wall socket; in fact, you need to more energy since the energy transfer into batteries is not 100% efficient (that's why chargers get hot: they're dissipating energy into the air). So right of the bat, you're using more energy. In addition, construction and disposal of batteries is not so good environmentally. Strong acids and heavy metals are generally not so good for the environment.
Mar 01, 2008, 13:22:26 ferrydust wrote:
Large emphatic thanks.
This could be very interesting. It’s leap day here, and the first of March there. Maybe MIT or CalTech or JPL or ARPA would find it valuable to observe. There’s that physics prof at U Mass who’s been working on something like this, but he hasn’t gotten the funding yet. (If it doesn’t ring a bell I can dig up the link.)
I’m going to have to do some searching and reading before I can offer anything worth biting into re engineering; but I’m delighted for the opportunity.
I’ve been strolling through EE as, as I said, a (bemused) spectator in the hope that some of it would stick; but such is my ineptitude, the whole handful doesn’t satisfy the requirements for the simplest coherent pattern.
Which is one excellent reason to be delighted with this circumstance; in thirty-five years of fascination, I haven’t encountered an opportunity to sit down to a lengthy conversation exploring what I regard as its most fascinating aspects— or their technical grounding. I’ve done theater sound. drawn blueprints for a guy who helped design the sound system for the former symphony/opera/ballet/etc. hall here, and a friend won an Oscar for sound (Chicago), and yet.
The link was to Red Wine Audio. I thought they had white-paper-type stuff there but I just checked and they don’t. 6moons.cam has some reviewers who seem to my ill-informed eye to have some EE competence, and they have a taste for off-beat stuff.
The point about batteries is well taken.
Ah, now it’s March here too.
Gee.
My trustworthy reasoning/perceptual apparatus, such as it is, is on the wane. Up all last night writing or maybe just blowing verbal bubbles, but in any event my brain (metaphorically speaking) is imploding and I am impelled to effect repairs.
Again much obliged for the courteous conversation.
Mar 01, 2008, 21:06:23 Slartibartfast wrote:
"I'm not sure that outside-the-box thinking is always best"
Not always, no. It has to have some baggage along with it, such as understanding of underlying principles.
Positive feedback is mostly, as Turbulence says, a bad thing. You hear positive feedback sometimes over public PA systems; if allowed to continue, it can blow amplifiers and speakers. Negative feedback is almost always used. I say "almost" because I don't know for sure.
My background is I've got a BS in EE (BS is much of my oeuvre, some folks will tell you) and of what I learned in college, most of what I use on the job is feedback controls and calculus, with some perturbation theory thrown in. I haven't done circuit design for 25 years or so.
About power supplies and line noise rejection, I have to disagree with Turbulence on general principles. Most power supplies on mainstream consumer audio are designed to convert 60Hz AC to DC at some rail voltage, rejecting 60 Hz as a priority. If all that was on your power line was clean 60 Hz, those would work rather well. Unfortunately, there's noise and distortion, which the power supplies can pass through to the rail voltage, and so it makes it out to your speakers as modulation on the music.
Which is not to say you have to go out and buy an expensive power-line conditioner, just that, like nearly everything else, it's more complicated than you'd expect.
Mar 02, 2008, 01:37:30 crionna wrote:
Is McIntosh still the cat's PJs? That gear (in the display rooms at House of Music) is the only reason I deal with the parking hassles at the Trader Joe's on Bryant.
Oooooh, green light, clicky knobs, glass fronts...
Mar 02, 2008, 01:39:57 OCSteve wrote:
I have to say that I feel like I’m missing something with all this discussion. I mean literally, I can’t hear a lot of this stuff you guys are talking about.
Reminds me of a guy I knew in school who had perfect pitch – I was jealous of that ability.
Mar 02, 2008, 01:48:49 Turbulence wrote:
Slarti,
I'm not sure we disagree too much on power conditioning. My only claim is that we know how to do a very good job in that area. But that doesn't mean that consumer grade equipment has power filtering worth a damn, or that costs more than a dollar.
Mar 02, 2008, 01:51:53 Turbulence wrote:
OCSteve,
If it makes you feel any better, I've never heard any difference either. Which is why I've been trying to find out what mechanism people thing causes transistor or feedback amplifiers to sound...not good. Also, I've got to defend the honor of my beloved transistors. Plus I spent way too much pain and agony learning feedback design skills that I can never practically use, but I'll be damned if I'm going to let some pissant audiophiles make those skills even more useless ;-)
Mar 02, 2008, 02:20:43 OCSteve wrote:
[i] I mean literally, I can’t hear a lot of this stuff you guys are talking about.[/i]
Hmm. That likely explains why I’ve been pretty disappointed with the 5.1 surround sound system I bought.
Mar 02, 2008, 02:55:42 Turbulence wrote:
If you can do it, I recommend visiting a Bose store; many of them have a sound proofed home theater room off to the side where they try to blow you away. I'm guessing its a good place to hear medium good sound technology for free.
Plus I have an irrational fondness for Bose. The idea of secret military research lab hiding in plain sight under the guise of being a consumer stereo company just cracks me up.
Mar 02, 2008, 03:02:58 Slartibartfast wrote:
Um...what?
I've never particularly liked Bose speakers. They do nifty things reflecting music off of walls, but that pretty well messes up any spatial content. It does make some stuff sound better that never had spatial content, though, and their high-end response...well, sucks.
If you can't hear that, though, buy them.
There is some guy that works in my field named Bose, and he is related to the loudspeaker Bose, but I don't think what he does is a family business.
Mar 02, 2008, 03:07:05 Slartibartfast wrote:
IOW I believe the claims of Bose every bit as much as I believe the phrase "The genius of Matthew Polk".
Mar 02, 2008, 12:01:23 ferrydust wrote:
Me too. With whatever apologies are due to Turbulence, Bose is not The Source; they were fairly common in clubs and theaters 15-20 years ago, much less so now. Those that are still there are there because they’re still there, not because they’re a current upgrade.
Depending on where you are, there might be an emporium(?) relatively nearby where you’re able to viscerally contrast and compare, with a variety of gear to choose from,
Not necessarily a panacea for inexperience though. There’s one high-end place across town I’ve enjoyed patronizing, bought my Cambridge amp there, but often what I’ve heard is pallid and unconvincing. The guy is forthright, even to his acknowledged disadvantage, and not condescending, which can be rare. So good luck.
A couple of experiences from ancient times which have framed my audio ruminations for forty years.
I was visiting a friend in his music studio. Jerry Ricks, who had a twelve-foot shelf of 1/4" tapes of old blues recordings. Completely amazing. He’s got a couple or so recordings, obscure; but sonic references. At the time his good friend Doc Watson was visiting too. The legend. We were sitting in a triangle, four or five feet apart. Doc picked up his D-18 (30’s as I recall) and gave us treat, eminently qualified as he was.
The guitar was a bit of a legend too, but what I’ve remembered all these years was the gorgeous richness of his voice.
That night at the club where he was playing, I sat in the front row, roughly eight or ten feet from him, and by comparison with that afternoon his voice could have been cut out of cardboard; because it was coming through a sound system. Granted this was forty years ago, but I doubt you’d be able to do much better today in the average club.
This suggests that attention to sound quality is well spent if, as if the case, there are options, sonically speaking.
After all, the stated goal of a great sound system is to convincingly portray the musician(s) as in the room with you, or you with them in their room,
The second framing experience occurred regularly a couple of years later. This is more obscure and less direct or conclusive, only serving as a suggestion, but every time I put a new disc on the turntable for the first time, the recording had a freshness and immediacy which didn’t last through repeated playings.
I’ve always supposed it implied the microscopic rough edges of the pressing that were worn away over time, those minute, not-yet-eradicated edges, added high frequencies with their effects, and are related to the perceived benefit of tubes. As it’s sometimes referred to, euphonic coloration.
Mind you, recently tube designs are described not as benignly distorted but simply clearer and less obscured. I think the intuition has to do with the same idea behind single-ended-triode designs, that with fewer elements in the circuit, primarily but not exclusively active elements (a big deal is made over the quality and sound of passive components), you are closer to the mythical holy grail of straight-wire-with-gain.
More controversially here, one of my root fascinations with audio is the mysteries attending it, the unaccountable subtleties that make all the difference, and the surprises that run contrary to settled orthodoxy.
That’s why, to a degree, I feel I can get off easy treating it as a thought experiment, a speculative fascination.
It doesn’t mean I have nothing to contribute, but it does leave me open to accusations of not having all my marbles or at not least playing with a full deck.
Finally— and this is where I shoot myself in the foot conversationally speaking— is my thought experiments lack significant empirical verification. To help the ball along I should at least be able to offer evidence by way of listening assessments; here, my quiver is almost empty.
In fact I’m hoping that I’ll feel myself compelled to spend some time in search of focused listening. Feel free to push.
A last probably essential point; the consensus, confirmed in that limited experience, is the most acute hearing is not the product of hard work, in the sense of tensed. We hear more if we are relaxed and immersed rather than trying hard. It tends to be why double-blind tests can be found suspect. Freedom to be surprised rather than forcing perception to conform to expectation.
Like that.
Again, all hail, and thanks.
Mar 02, 2008, 12:24:57 ferrydust wrote:
As if I haven’t promoted a sense of exhaustion with a surfeit of words (and confusingly misplaced commas), another significant contextual element:
I’ll take live over recorded any day if it’s in a relatively intimate setting. Crowded is OK, but close is where its at.
I like to be able to rest my feet on the monitor. I treasure a memory of a legendary free-jazz trumpeter blowing directly at my heart from two feet away. Absence of electronics apart from the necessary guitar amp. In front of, sitting next to, effectively on stage.
Not always viable but far, far preferable.
Nothing else like it.
Mar 02, 2008, 12:26:38 libjpn wrote:
Just out of curiosity, but why is it so important to hear where the instruments are? I think that it is, but if someone asked me that question, the only reason In could give is that it is more realistic and realism is good. But is it really?
Mar 02, 2008, 16:23:12 ferrydust wrote:
It has to contribute to the visceral excitement and thus reinforce the immediacy. The realism.
And from the above mentioned experience of live music I feel immediacy is one of music’s founding pleasures. I’m persuaded that goes back to well before the existence of hominids. Only question is how far.
Another element is timing, in the sense that if the vibrations reach your ears at precisely the femtosecond they would traveling from the ‘voice‘ to you, with both of you in the same space, and that’s true of each and every ‘voice’; voilá, a more immediately persuasive and pleasurable illusion.
Thus the ideal of straight wire with gain; the more artful the engineering, the shorter the signal path, and the less chance for the time and phase relationships in the recording to be screwed up.
Turbulence, did your endeavors leave you feeling fully informed about eddy currents? They come to mind through considering signal purity. Denis Morecroft of DNM has been talking about their importance for twenty years or so; recently there’s bee a fair amount of press around connectors, with Eichmann and WBT reducing the amount of metal drastically to limit them.
Also as I understand it all digital inaccuracies and thus distortions are timing errors. Hey, it’s true of computers; the shorter the signal path, the more signal can get through the circuit more quickly, and the better integration of data streams. (Yes?)
And those signal paths were already measured in micrometers, max.
One of the unimaginable things edging over into mystery that fascinates me is the reflection that the conversation can inch its way, or leap, over into quantum physics: But even more that those things are [i]audible[/i]! IOW, you can hear things tha defy measurement. (Well, maybe with a tunneling electron microscope or its acoustic equivalent. Not many of the small firms endemic in the high end are using them yet.%) )
Do read, by the way, [i]This is Your Brain On Music[/i], for some hint of the neurological foundation for hearing. <a href="http://www.psych.mcgill.ca/...">Levitin </a>has worked as a major-league producer and sound engineer, commuted from LA to SF to study neuroscience at Berkeley, and is now a leading researcher. Knows his chops.
The direct point of mentioning it is the incredibly fine timing signals the brain operates with. Nothing like wetware, I tell ya. An example:the average brain contains one hundred billion neaurons. Each neuron is connected to other neurons— usually to one thousand to ten thousand others. He writes quietly that the number of potential connections in the brain is greater than the number of known particles in the known universe. (That’s directly copied from the test.) Wetware rules.
Resonances.
Last thing I’ll mention now. Acoustics. Familiar as a practice and discipline. Experts are highly respected when the can get it right, and the stories of failure are major and legion. It helps to have FFT and gobs of computing power. But last I heard there are still aspects that make it an arcane art.
I guess the whole thing hangs on phase and timing relationships in reflected sound, and manipulating the space so they don’t get mucked up and turn the sound to unappealing lifeless mud.
For our purposes the same thing can happen in your listening space. A fabulous system placed badly is a ruinous waste. Arranged carefully and knowledgably a modest system can be brilliant.
Mar 02, 2008, 16:28:56 ferrydust wrote:
Sorry about the typos and my incompetence with tags.
Mar 02, 2008, 22:57:43 Slartibartfast wrote:
"Hey, it’s true of computers; the shorter the signal path, the more signal can get through the circuit more quickly, and the better integration of data streams. (Yes?)"
No. You can put a five-second delay on music, and destroy none of the _relative_ timing information.
Mar 04, 2008, 00:00:38 ferrydust wrote:
I’ll try this again, hoping my splashing about doesn’t add muddle.
Parenthetically to that point, I’m hoping not to allow my potential pretensions to derail the discourse. And apologies to the degree they may have already intruded.
Anyhoo, I spent a half-hour scrolling through Ars Technica’s archives trying to find a too-vaguely remembered piece detailing a new chip, that may have offered a useful example of what I had in mind, but came up empty, so I’ve got to go with what I’ve got, with hope for ongoing improvement.
I was under the impression that one of the fundamental advantages of 45nm over 65nm chips was shorter circuit lengths. I translated this, whether in my imagination only, or by virtue of explicit description in the cited AT piece, as meaning that a shorter circuit implied more rapid, and therefore higher volume per time unit, data transfer. I also (vaguely) recall that new forms of circuit-block integration were involved. ..
This synthesizes in my mind the notion that my essential point, perhaps too poorly expressed to be useful, remains intact (if fragmentary).
It is that the greater speed implements timing improvements (which obviously depend on clock speed, details of which haven’t stuck with me) with the idea closer, less error-prone signal tracking is enabled.
The improved accuracy (in this construct) would hold in the case of signal delay, assuming the delay circuit keeps all the bits in proper order. If so, it’s identical chunks of data i/o, with improved precision intact no matter what the delay.
(Maybe?)
I concede my evident pleasure in blowing verbal bubbles and indulging conceits, and only hope they don’t disable useful advance.
Mar 04, 2008, 00:19:37 Slartibartfast wrote:
Oh, sure: distance equals time.
More directly, I'd imagine, device dimension is more directly a consequence of mask (if that's even the term, anymore. Did I mention that I was going IC design, at one time?) resolution. For microprocessors, the trend is for die dimension to increase, because the number of transistors for a given device increases as the design evolves. Counter to that trend, feature resolution is always improving, so that you can squeeze more devices onto the same (or smaller) die. When I was a senior in college, HP (I believe) had just put a quarter of a million transistors on a die, which was kind of a breakthrough. Now you're talking about a couple of billion devices. It's entirely possible that a given microprocessor design could move to a smaller die, necessitated by advances in fabrication technology.
In terms of physical dimension, keep in mind that for current devices, one clock cycle is equal to a couple of inches in linear dimension. So, a hazard of having large-dimension circuits is, in addition to delay, that the leads start acting like antennas, radiating and interacting.
Small linear scale makes many things easier to cope with, as the clock rate goes up. For audio circuits, small scale is overkill.
Mar 04, 2008, 04:32:42 crionna wrote:
This is pretty funny <a href="a comparison of Monster cables vs. wire hangers at Gizmodo"></a>.
Mar 04, 2008, 04:33:59 crionna wrote:
Oops, dang my commenting skills have totally slumped.
<a href="http://gizmodo.com/363154/a..." >Gizmodo comparison</a>
Mar 04, 2008, 04:34:14 Phil wrote:
Dammit, crionna, you beat me to it!! But you messed up your BBCode, so :P
[url=http://gizmodo.com/363154/a...]Monster Cables vs. coat hangers[/url]
Mar 04, 2008, 04:34:33 crionna wrote:
Geez, I give up. Try a cut and paste.
Mar 04, 2008, 04:35:57 crionna wrote:
heh, thanks Phil. Together we are both quick and accurate...
Mar 04, 2008, 04:56:08 Jeff wrote:
crionna -- BBCode uses [ and ] where HTML uses < and >.
Mar 04, 2008, 05:40:37 ferrydust wrote:
Thanks.
My own search hasn’t provided what I was looking for though of course some other good stuff surfaced along the way.
What I did find was that reducing current leakage (by a factor of 10), with its corresponding heat loss, along with a consequent gain in freedom from signal-obscuring noise was a large element in squeezing more transistors into a smaller space; and that the real driving force in shrinking chips is the financial benefits. The more transistors in a given square cm, the higher the profits.
Surprise.
But one iron rule, and it’s not [i]c[/i].
Now I’ve recalled the likely source of the distance=speed idea was in a story about high-rise chips, because stacking circuits keeps distances shorter. On with my search. (Update: searching..searching..argh! I know it’s there somewhere
A couple of linear inches? Wow. I would never have imagined.
& what’s the story on the temporal disjunction here, speaking of timing?
Mar 04, 2008, 06:09:51 ferrydust wrote:
Hey, of [i]course[/i] Monster Cable=coathanger! Monster=Bose=Polk=Rudy Giuliani. QED.
To be fair, Matthew has picked up a few modest bouquets of late.
Mar 04, 2008, 07:49:12 russell wrote:
To paraphrase the immortal Butterfly McQueen, I don't know nothing about no FFTs. What I know about the physics behind acoustics would fit in a thimble.
I do know a guy I play with now and then who has a little 5 watt Class A SET guitar amp that will take the top of your head off when the knob is on, like, 4.
The amp is red and white and about the size of a small suitcase. Kind of cute, actually. The tone is pretty clean, actually, and it doesn't have to be that loud to get a good sound. But listening to it is like sticking your finger in an electric socket. It's just *forceful*. It has cojones like nobody's business.
I'm not sure who made it, it might be a Bedrock.
Mar 04, 2008, 08:34:10 Phil wrote:
When I was living near DC, I played in a band with a guy who built his own amps. Same deal as russell describes -- I saw him use an amp smaller than a Fender Champ to fill rooms from living-room size to ballrooms.
Mar 04, 2008, 16:42:55 ferrydust wrote:
One of the [A href ="http://www.spectrum.ieee.or..."]things [/a] I found during my search.
“For audio circuits, small scale is overkill.”
Slarti, that’s what I was talking about, what I’ve thought is so amazing. Human hearing, as a whole package, is so outrageously finessed, so fine in its discriminations on levels we regard as unworthy of attention that in fact are crucial; that everywhere you look in the circuits between the mike and speaker you will find obviously inconsequential elements, nooks and crannies, which when altered are, so the testimony goes, adequate to degrade the life in the musical signal.
Depending on your disposition it’s a recipe for craziness or a prospect of endless potential for improvement.
That’s (one) theory.
Mar 04, 2008, 16:47:26 ferrydust wrote:
eventually... code will no longer be my enabler in public incompetence, maybe.
Mar 04, 2008, 23:03:41 russell wrote:
slightly OT....
In looking at the source for this page, it seems like markup for anchors uses good old HTML angle brackets, but markup for everything other than anchors is square brackets.
Is that right?
Mar 04, 2008, 23:20:34 libjpn wrote:
Well, what it is, is that I could only find a plug in that allowed the comments to be marked up with BBCode (the square brackets), but the engine generates html source. I really hesitated, because I thought the markup should be html, but there were no plugins to do that in the comments. I finally gave up and went with the BBCode.
I've also found two comment preview plugins, but they are a bit problematic. The first uses javascript to create the preview, which creates problems in parsing markup (and I imagine that the BBCode won't work at all) The second looks very good, but it is from 2004 and it requires you hack your skin, but it looks like the php has changed, so it is not clear what you need to replace.
I've also just installed a plugin for a recent comments RSS, but I cannot find where it should go. Anyway, I'll probably have to leave this till the weekend, and I don't want to fire up the comment preview without taking the site offline while I test it, so a couple of days to cool off is probably a good thing