Tuesday, June 30, 2015

Shielded Box for line level signals

Bud seems to make more insulated alloy boxes than Hammond.  Mu Metal is a trademarked name, but many companies make similar alloys which combine EMF and RFI shielding.

Anyway, my current plan is to adjust levels digitally upon switching amps, therefore no need for attenuation in line.  But ultimately a Bud box may serve as the Y-adapter for the signal to both amplifiers even if there is no attenuation.  And the box may switch the signal from one box to the other to minimize loading.  And it may have an optional attenuator.  It may even have switching relay inside.

Why not modded DCX's ? Why not miniDSP ?

Instead of getting DEQ units for each range of digital crossover I need, another alternative is to get a DCX with a modification to permit digital outputs for each range.

Unfortunately there are not many companies offering this mod (and in fact the company that did appear to offer this mod according to old posts in various places does not list it on the company website anymore).  The last time I saw this mod listed, they wanted $999 just for the modification, or $1299 for a modified unit.  Since the price of a new DEQ in the US is now $299, the modification price equals 3 brand new DEQ's, and the cost for a fully modified DCX with 3 digital outputs equals 4 brand new DEQ's.

To add to the uncertainty, the only picture shown for the modification only shows high and low pass outputs, with the mid connector marked N/C.

And then it's a modification by technicians of unknown skill, compared to a brand new factory made unit with warranty.

And having the 3 DEQ's means that you can have meters or spectrum on each one, which is cool.  Though perhaps it's not so cool to have a pile of three units when one would do the same job.

Also, the DEQ's have AES/EBU output and Toslink output running in parallel, whereas the modified DCX's only have coax spdif.

On the other hand, the downsides that have appeared with the DEQ are:

1) Space, takes more space
2) Reliability?  I've had two DEQ's fail since 2005 but not one DCX fail since 2008.
3) Balance only to about nearest 0.2-0.3dB
4) Level only to 0.5dB
5) It takes some calculation and testing to set up a 24LR crossover.
6) You are limited to 6dB, 12 and 24 LR, and a few other crossover choices.  Notably you can't do 48LR because that would require making two 4 pole Butterworth filters.
7) Less convenient to mute.

Fortunately, delay is settable (though the delay menu is hard to find).

I'm still planning to go ahead with DEQ's for each crossover range.

An entirely different option would be to use miniDSP's.  They would cost more than the DEQ's and I'd still need one per range, they require a computer to set up (the DEQ can be set up entirely from front panel controls), I'd have to learn how to do what I want, and there's no "meter" option.

I still like things I can control directly without having to plug in a computer.

Monday, June 29, 2015

The Power Box

Cullen Cables makes a relatively inexpensive 6 outlet power box that I might get for my amplifiers.  This design would let me plug in up to 3 Insteon on/off modules to turn amplifiers and/or trigger supplies on/off.  I've also been looking at flat rack mount boxes by Tripplite that feature built-in ammeter, however those flat strips aren't friendly to the Insteon modules which would have to be plugged in through stub cords as I am doing now.

My dream version would have two wires wrapped around the hot inlet capturing info for a remote ammeter.

I came across this looking for serious upgrades to Sonos.  Over the weekend I enabled my 6th Sonos Connect, for the Turntable/Tape/Masterlink pod in the master bedroom.  This new Connect lets me more conveniently play turntable or tape throughout the house, and without going through multiple preamps that used to be required so it sounds better.  Plus it will allow me to record FM without worrying about what I last listened to on the main bedroom Sonos Connect.  It will also make it more convenient to monitor digital levels.

But as good as the Sonos connection just sounds now (listening to tape on the living room system was mind blowing…about as good as it gets…despite going through so many things) I wonder about upgrading the analog inputs I use ubiquitously.  Modifiers seem much more interested in modifying the analog outs which I don't even use, and the clocks which I think make little difference.  Unfortunately the old Cullen Circuits mod (now available through Wyred4Sound) is mainly a clock/dac/analog-out upgrade, and its key feature is up sampling (something I don't want).

The balancing act continues

I've found I can balance both main power amplifiers pretty close to the 0.1dB spec simply by using the Behringer DEQ 2496 image "tilt" control.  My current adjustment is 1 degree Left for the Krell, and 0 degrees for the Aragon (the Aragon already now being balanced to .15dB or better).

So then level matching the two amps comes down to level, and my current level setting is +1dB for the Aragon.  That's as good as I can do through the DEQ because now I find out it can only set gain in 0.5dB steps.  Fortunately, that almost matches the Krell.  Actually the Krell seems to change it's gain any about 0.3dB during the first hour of warmup.  The +1dB setting puts the Aragon at the low end of the range of the Krell, almost identical with the first minute of warmup.  If the DEQ had finer control, I'd likely set the level to 1.1dB or 1.2dB.

This 0.1dB matching for ABX testing is much more stringent than usually done in audio, notably more stringent than I've been doing for electronic balance (which I hadn't been paying attention to and just assuming it was correct since I have no variable resistance elements in my system…and now I see I had unintended differences larger than 0.3dB) and in setting up crossovers for triamplification (which I've done since around 1979 btw).  For many years I used a Pioneer D21 crossover (a Series Twenty beauty) and thought of it as about as perfect as such a thing could be, and it is only settable in 1dB steps.

For a few minutes, I was thinking my project of using DEQ's as crossovers was doomed because of the limited "tilt" balance control and 0.5dB steps.  But now I'm thinking "close enough for crossover" since even with the 0.1dB adjustability of the DCX units, I can't say my adjustments are anywhere near that accurate and I often play with changing levels by 1dB or more to see if a new setting would be better.  I recall differences of 1dB as being hard to tell apart and have never expected my crossover adjustments to be better than 0.5dB.

Strangely, IIRC, small differences in timing are more audible, with 0.5dB in level difference being not unlike 0.5 msec of delay, with even 0.1 msec being important (more important than 0.1dB).  0.1 msec is close to the difference of moving speaker or listening position 1 inch.

I just won a Denon 5000, my 3rd new R2R DAC.  I now have enough of them to do all three ranges in the living room system if I'm willing to use the same DAC for both midrange amplifiers, which I now think I will do since I have useable digital level adjustment with sufficient accuracy for a crossover (and almost sufficient for ABX).  I could even control the DEQ settings with a trigger voltage via a trigger-to-midi device.

I think I will implement a level-box as my ultimate Y adapter for the two amps, but in the meantime an Audioquest Y adapter will do, with low cap cables possibly from Blue Jeans.   None of the cables I am currently using are particularly low cap, though I could have done worse.  I was shocked to find that the 4 meter Monster 1000i or something like that (it was over $100 if not $200, IIRC) had 800pF and I decided that made it totally unusable.  But even the Straightwire 1m cable I have for the Krell is 300pF, and the combination of cables I have to the aragon is over 400pF.

I will need to get a rack.  The piles of equipment behind each speaker are already too high, and I needed to stack the turntable above the Aragon amp already…with a rack I can also support a couple of boat anchor DAC's and other stuff.  I've picked out a very nice audiophile grade rack and turnable stand, it will put the turntable at 48 inch height which is about all I can deal with.  Before finding the one I want now from Mapleshade in solid maple, the previous ones I was looking at had glass shelves.  It's looking hard to find the Lovan Sovereign I used to lust after, and even the Lovan Legacy that is even featured at Houzz.  There is a four point stand by VTI that looks OK for a mere $400, and even $100-200 glass stands at Wayfair look tolerable.  I don't much like the Sanus stand I have now under the keyboard.

I've been matching the Aragon to the Krell + Audio GD level because the latter has sounded so good, despite being different from my last DEQ Calibration in January.  I think I'm playing the midrange at a higher level now by about 2dB or so and that sounds better because it dominates the subwoofer bass, giving the sound a tighter electrostatic bass sound.   Or perhaps I should say it now equals the bass, whereas previous the bass dominated giving everything a plate-amplified subwoofer sound.  This adjustment needed to be done by ear anyway, it's hard to say precisely where one terrain of 10dB hills and valleys should be joined to another.

And so if I was wrong by 2dB during an actual calibration--with careful measurements and listening, just how critical is it to set level better than 0.5dB?

Monday, June 22, 2015

Determinacy is useless

One of the kind of thing I've been thinking of a lot about the differences in analog and digital…which generally means LP and digital…is that the analog sources apply their varying flaws (typically audible and near-audible levels of time variation, aka wow and flutter) so that not only do you hear something different every time (as you would do if everything were absolutely identical, merely because You have changed, especially having already having just heard it…) there is actually something slightly different there, and each time this difference may serve to highlight things differently and bring something different out.  Given that we are always going to hear something different anyway, it's not useless for it to actually be slightly different each time.

Of course the main reason why LP's sound better when they do is the mastering, and digital recordings of LP's sound very good to me (preferable to some CD's because the mastering was so much inter).  But they don't have That Magic, because the analog flaws at one time and place have been frozen, immortally.

It's not simply that hearing through the flaws lets us interpolate back to the ideal original--similar to the digital recording.  It's that they trigger more difference each time than we would otherwise hear, when we are listening to the analog.  The analog is, in it's way, a live performance.

Anyway, I think it is worth studying the psychology of of it all, applying science to the magic of the Placebo Effect and Audiophile Experience.  Not that I want to give the magic up, I want it all!

So I'm sure the techno freaks would say fine, we'll just add LP distortions to your sound.  I'm sure that will be a disaster for a long time.  But let me describe how it might work.

All the flaws should be idealized rather than sampled.  We don't want to simulate the things we don't really understand.  This is not sound, it is modification that attempts to retain the integrity of the original.  Therefore, it must have simple changes, not an accumulation of mud in order to be biologic.

(I criticize many of the presets of my Kurzweil K2661 synth for this.  They go for too much dirt and realism.  Of course you can program anything you want, make it all sine waves or whatever, or any sample, or just tone down the dirt channels.  But I'm lazier than anyone and I like the presets right according to me.)

And it's done in super high resolution, at least 48 bits of resolution, then dithered and so on.

And what I want is the ABX learning effect, where it choses from affects at levels not believed to be audible, or just above audibility, and you can guess at any time and move up to the next.  I want to be able to program the effects and their levels, perhaps specifying common wow frequencies and so on.

Of course I alway think there should be a geek version.

99 wins, one fail for DEQ

Discovering that I could easily program one passband of an electronic crossover for a multi-amplified system into a Behringer DEQ 2496, I have been planning to buy 3 more units, or something like that.  It turns out these are (mostly) the most versatile units I know of, nicely made, and cheap.  The have AES digital I/O, which is how I use them (forget the analog bits, those I want to avoid as much as possible, though I still am, mostly, and with little technical loss, they are objectively good enough when not used above 4V output not to have any proven audible flaws, but of course they are not built with audiophile design or jewelry).  And optical, so you can have two parallel outs (optical and digital…another fact I take advantage of).  Just the most useful little tool for me now, though I could dream of something similar and fully programmable, pop up a programming interface when you plug it into your compeer via USB, …

The DCX 2496 crossovers, which I still use except for the panel passband, don't have digital output.  If they just had that, they would be invaluable…though I don't really like the SRC they use, they convert everything to 96kHz.  The DEQ's don't do SRC, they pass the input to the output without changing sampling rate.  I prefer that, so I'd prefer that to having a modified DCX with digital I/O (I expect that would cost about as much as 3 DEQ's too).  The DEQ has a bigger screen, it has meters and spectrum analyzer…it's just cooler to have one of these for each passband than a DCX, except of course for the added cost and complexity.  But if you need it, you need it.  (Wait.  Why do I need it?)

Except, except one thing I wasn't expecting at all.  I guess it doesn't actually derail my plan, but it is a big disappointment, and it also in-general-now (but not wrt my Big Test, in the Long Run after I Work This Out) ruins my plan to use midi-programming of a DEQ to implement level changing for ABX testing.

The Behringer DEQ 2496 doesn't have separate level controls for each channel.  It's mind boggling that a thing with so many features lacks this one basic thing.  You can set overall "makeup gain" to 0.1 dB in the Utilities panel, which is fine though it seems peculiar to relegate so critical a thing to the back door, but there is no separate Right and Left or whatever, anywhere.

What the Behringer does have is a "Rotation" control, that allows you to set "Rotation" (something like balance) to 1 degree.  But that is not close to being good enough, has a non-linear relationship with the dB difference, and inconveniently affects both channels at once (in the opposite direction).  Even with a mere 0.3dB difference between the channels of my Krell, 1 degree of rotation overshot the mark.  I was easily able to dial in the correction on my Tact RCS 2.0, which had been set to 0/0.  (I had been worried that previous adjustment had been the cause of the 0.3dB difference.)

So this won't work for fine tuning levels for ABX testing.  It's basically good enough for crossovers in my system.  I can adjust the all-critical wide range middle speaker using the Tact, the subs can be adjusted with their own controls, and I have no idea how many dB wrong the current super tweeter balance is, and they are effectively inaudible at the listening position.  (If I do need to rebalance the super tweeters because they are noticeably different, as they look, quite possible the rotation control will be sufficient--that's about audible difference generally--ABX testing is actually to a much higher standard.)

So, for ABX testing, I'll have to use some other level adjustment-in-general-and-now, AND I have the sad knowledge that the Aragon has quite a bit less output than the Krell, not the reverse, so the Krell would need to be attenuated, and that Won't Do, so any arrangement will have to be temporary or something.

However for general use, the Midi control with 0.1dB control regarding level (if not L-R) will do fine, and that may do (with some adjustment/correction) even for a permanent Krell/Aragon installation.  I'll get the L-R balance of both perfected through other means, then changing the overall level is all that's needed.

Other things should have their balanced fixed first also.  So the Only thing we absolutely need is level (real L/R control would have been nice, somethings won't be testable, even right now they aren't, and I'm not going to put permanent attenuators of any stripe on the line to the Krell, I just won't.

The Krell has been sounding marvelous, btw.  Over the weekend, listening to FM mainly.  I was even drawn to Opera for awhile.  I had to setback the thermostat first 2 degrees and finally 3, taking back the 3rd when I put the Krell in standby.  Standby ended on Sunday night, though I've had the Krell on tonight for testing as described above.

I'm OK with raising the thermostat if I can have my Krell.

Apologies to those who say everything must be scientific.  At home, in my closet of a living room, I practice magic.  Magic requires suspension of disbelief.  The magician makes big sacrifices, lifts the Krell up onto the altar, and everything has a purpose…

Sunday, June 21, 2015

The Hi Def Recording from June 7

I am starting to post the recordings I made of Priscilla at the River City Audio Society entertainment event on June 7, 2015.  I will update this page as additional files are added or deleted.

The CD quality mono recordings made with studio grade equipment are here:





The High Definition (24bit/96kHz) mono recordings made using studio grade equipment are here:


The original recordings were made in 24/96 and resampled to CD Quality using Triumph and iZotope resampling and dithering.

Bonus Track: The first take of first song was recorded too high and has clipping near the end, but is a breathtaking performance.  CD Quality.


Thursday, June 18, 2015

The Yggdrasil

I'm pleasantly surprised to see a new PCM DAC on the market from Schiit.  At $2299, this is much less expensive than the least expensive PCM DAC from MSB, and there aren't many others being made now.  And it sounds very sophisticated, perhaps more sophisticated than any preceding PCM DAC in many ways.  I'm somewhat interested.

Though I need to have at least 6 DAC's in my home (4 in the living room system, 2 for the bedroom system) as standalone DAC, not counting the endless DAC chips in other equipment (ONLY the living room and master bedroom systems are High End enough to warrant actual DACs…if any do…).  So I'm not sure if I'm going to spring for $2299 for any of them.  If I get one $2299 DAC for the living room, I'd automatically have to get two, so both amplifiers under A/B testing would have the same kind of DAC.

Here's a discussion at ComputerAudiophile.  Here's an endless rambling discussion at Head-Fi that generally favors PCM DAC's as compared with Delta Sigma and DSD.

And they say Mike Moffat was working on this for 5 years (along with a mathematician and others).  I only got on the PCM bandwagon last year (though it seems like a dozen years since then).  I see now that he's been on the PCM side of the debate now for a long time.

I met Mike Moffat very briefly in 1979 when I was an audio technician at Audio Directions.  He came by with his Theta Preamplifier, which was demonstrated at one of the first meetings of the San Diego Audio Society which I later became a President of.  He is very charismatic.  The preamp sounded good to us, but the tech staff I was part of felt it was sloppily built.  Well that was a long time ago when Mike didn't really have a company yet.  I could hardly have imagined what he would be up to in the future--a towering and pioneering figure in the history of high end DACs, though he has not been without many critics including some who might well call the whole field a fraud.   Hypester or Genius?  Honestly I could never tell, though I did lean toward believing he was more on the hypester side.  But maybe I was wrong then, or maybe I've become one too.  Anyway, I'm glad to see he put his Jobs-like force to make something I now find interesting.  Moffat seems somewhat like Jobs to me.  Jobs was of course the penultimate hypester, but he had good taste too, chose the right directions, chose the right ingredients, not so much the genius inventor but the genius selector.  I myself use Macs for that reason…though I'd always felt a well configured Gnu/Linux could be better, getting there isn't worth it and I for sure don't have the time.

And I'm pleased to see others thinking about digital technology in a similar way to me.  Though many are still clinging to DSD, in the last year or so (about since I got on board…proof of my overwhelming influence) PCM is back, and more and more designs are recognizing the virtues of real PCM.  I consider DSD a fraud (and have always done so…going back to David Rich's take on it circa 2001 in Stereophile), originally designed to give Sony lots and lots of money by owning the music business and owning the technology.  Well they did get to own those things, as it turned out, perhaps by convincing certain people that DSD was a key, even if it turned out DSD (and SACD) has never been a huge success, it ended up getting Sony the things they thought would buy them endless profits.

I nevertheless consider SACD relevant enough that I have SACD players that get resampled by high end ADC's to 96kHz.  When you're resampling an analog signal, as I do, you can choose any target frequency and it works great.  In fact, I believe strongly that it's better to physically resample in the analog domain generally than to do it digitally, and you might as well sample at the best rate you can support, it's like super-dithered, but it's especially true when converting DSD to PCM.  And DSD can be converted to PCM with little loss--native PCM end to end is a relatively low loss process for a bandwidth limited signal--whereas creating DSD initially causes a huge loss in information in the middle highs up to the inaudible.  I've heard that (not in DBT) but of course I've always been strongly expecting it.  There was only a brief period around the time I got my first SACD player that I had hope for the format.  I didn't bother to even get an SACD player until after I'd gotten a DVD-Audio player, and had gotten frustrated with the Stereo-Unfriendly menuing and lack of titles.

But it seems in the cult world, the status of N.O.S. DAC's is looming larger in the imagination generally than PCM as such (though many combine the two concepts).  IMO a DAC without a proper reconstruction filter is a noise effects processor.

Moffat has never been that kind of fool, all his digital equipment has had considerably thought out digital filters…that's been the hallmark of his products.  So at least he's no fool in that way.  He's focussed on time coherent filters--a concept which makes abundant sense.

How do I know this?  Well of course I don't "know" anything, at least in the sense I could "prove" it.  But I hardly get the bulk of my ideas from listening tests.  Basically all of the times I've attempted to do a double-blind test on something I absolutely believed would be audible, though different from the total objectivist canon, I not just failed, I couldn't even get started.  When confronted with level matched "A" and "B" I found them identical, so the best I could do would be to guess on getting the first impression. Whenever I bothered to go that far, results were close to chance.

But who needs to do such listening tests?  Subjectivists of course!  And I'm basically not a full throttle subjectivist, by any means.  I've experienced myself, strongly, and taken to heart the findings of the DBT testing and attempts at DBT testing I've taken.  Perhaps I haven't taken them to heart the way others do.  But I don't believe proper listening tests actually tell you very much, and are very hard work for what little they do, reliably, tell you.

Better off when possible, then, frequently obtaining and mostly relying on objective test results!

In many cases it's not hard to know what audio circuits do, when they are linear and when they are non-linear.  Of course we want them linear all the time, and in every way, if possible, and if not, we'd like to sacrifice as little nonlinearity as possible.  Do you have a problem with that?  Most subjectivists do!  Even though what I'm describing is not just "a number", it's a concept, it's an objective idea.  They start with hearing what X "sounds like", by performing tests that most likely are producing little more than a random number compared with what one would be getting in a blind test.  I take the blind tests as showing what we actually, and provably, hear.  Everything else is speculation, and in my view speculation about how things work is far more meaningful than performing sighted listening tests.  In fact I fully believe that sighted listening tests are pretty much useless.  They are unreliable and their main product is superstition.

So I want to shove off the idea that I'm just a "measurements guy," though measurements are a fine way to check things out--generally much more reliable, and more interesting, than listening tests.  But better yet is understanding how things actually work.  Did you know that with very high bias (so high basically nobody uses it) MOSFETs can be perfectly linear?  That's what we want!  Too bad nobody does things that way.  It seems everything is compromised.

And it is true that often the most useless things of all have been specifications.  Manufacturers specifications have often been outright lies, or very calculated incomplete truths.  Even at their very best, they have never told the whole story.  The best story of any piece of equipment basically involves a full technical exploration.  The closest to this I know of has been the measurement work by John Atkinson.  I respect him enormously for that.  Meanwhile, I continue to imagine something far better, a far more complete story being told.  And the best story of all includes context.  Fortunately with the web we can get much context quickly, though often we lack the contexts before around 1996.  There is some good old information on Usenet…is that now archived on Google Groups?

Back in April, atomicbob at Head-Fi did a very impressive technical exploration of the Yggdrasil, and he says he plans to check out other DAC's also.  (I had done a lot of technical investigations when I started this blog, and the intent was to do more and publish them here.  My old ones tended to get written down in notebooks that got lost in the pile, and stuff like that.  Sometimes I'd come across a pile of old spectrum analyzer results and find myself amazed that I had ever done something like that.)

So then we can continue making things better, primarily based on understanding how things work, and secondarily by checking them out technically with measurements of various kinds, and ultimately, but very incompletely over any finite amount of listening, with listening tests.  But unless you're going to do the most careful level matching, I wouldn't bother with any A/B testing, just put new thing on and see if it works.

The science tells me I can do that.  I need not bother with the "listen for yourself" mantra of subjectivists (although I do, sometimes, do just that, but hardly ever in a testing context), or anything I consider overpriced, or relatively unimportant.

The science tells me I can do what I want to do.  I need not be shamed by others.

Except possibly by those who say I should be devoting far more attention to the speakers and acoustics than to the electronics.  I do feel somewhat embarrassed I haven't spent more time on the acoustics.

But that's more work.  I like thinking about electronics.  That's what a big part of my hobby is about.

And as I think about how things could be made better, I often like to implement those ideas as well.  It seems to me often that that is what I spend too much time doing.  But there's often no way around it.  If you want to listen to the stereo you got to connect the speakers.

AES splitter

The new plan for wiring crossover modules and alternative amplifier pathway in the living room will use a 4 way AES/EBU splitter from Henry Engineering.  That will allow me to connect 4 separate DSP units to the main digital output from my digital preamp: subs, panels, super tweeters, and alternative panel amplifier.  With a separate DSP unit for the alternative panel amplifier, I can set the gain digitally for a 0.1dB match.  If I'm not doing amplifier testing, I can use the "alternative" feed for other kinds of testing.

XLR Shielding

About 10 days ago I had an opportunity to record a young girl singer.  She sounded fabulous and I will eventually be releasing the recordings in various formats.  When gathering my gear together, I could not find the required TRS to XLR cable I need to connect my microphone preamp (an M-Audio DMP3) and the recorder (an Alesis Masterlink).  I quickly brewed up a cable.  Actually it wasn't so quick.  First I had to research the TRS and XLR pin connections (TRS is as intuitive as it seems--the tip is hot).  This was the first and only soldering job I've done all year, and I didn't do many last year.  I've never settled in with a good temperature on the RadioShack digital soldering station I now use, it's only a few years since I ditched my old Weller station with broken thermostat, so I even looked online for what temperatures other people use for Eutectic solder.  Just under 700F seems about right, or it could be over 700 if you have "confidence."  Actually it takes less than 400F to melt the solder, but the higher temps let you work quicker--which is better (better fast and hot than less hot and slow).  This was all being done in the wee hours of the morning, 2-4am, with the recording scheduled for the next day at 1pm (an early hour for me).  I made the cable, and then found I had done it completely backwards so I had to take it all apart and put it back the right way.  Finishing just before 4 am.  Whew!

Actually what I did was to cut a 6ft TRS to TRS, and solder in an XLR.  But if you look at the official AES recommendation (as written up by Rane), it appears that a TRS to XLR should be wired lifting the shield at the XLR end.  (This is very strange and seems to deviate from all their other recommendations.)  I wired pin 1 to the shield, which I measured at being connected to the TRS sleeve.

I just ordered a new TRS to XLR cable from Sescom through Markertek…it will be interesting to see how they do it.

Wednesday, June 17, 2015


I believe that jitter is not as big a problem as most subjectivist audiophiles believe.  For one thing, the best science, published by JAES, is that it takes jitter differences in the 10's of nanoseconds to be audible--barely audible.  Meanwhile typical equipment such as my Sonos Zoneplayers (not normally considered "high end") has about 220pS of jitter, at least 50 times less than what would be barely audible.  A few months prior to reviewing and measuring the Sonos Zoneplayer, John Atkinson tested a 2004 vintage DCS stack (about $80,000 or so) and found it had 230pS jitter, at 160 times the cost.

I found a related and interesting thread at DIYAudio about jitter.  The OP (now apparently banned) was saying basically that SPDIF is not the problem (well, it is actually) the problem is with poor receivers.  He showed several different receivers had very different levels of recovered jitter as appears on spectral graphs (he did not reduct it to a number).  Later posters claimed his worst case (a NOS DAC with 8 parallel 1543's--which the OP said was an audiophile favorite) looked bad mainly because of lack of digital filtering mainly and jitter had little to do with it.

Anyway, one poster in this thread rightly points out how simple it is to show the termination of a SPDIF line.  Measure voltage unterminated with scope, then connect 75 ohm load, and it should drop 50%.  The receiver can be measured by being sure it causes voltage to drop 50%, and the actual impedance can be estimated from how much it does drop.  This poster is somewhat concerned about jitter, pointing out that when an analog signal is turned into digital, timing is everything.  That's a somewhat weak argument IMO, but his better argument is that impedance, etc., does affect the digital signal slope, and therefore clock recovery.  It is not unimportant, even if downstream clocks and servos can mostly repair the damage.

Tuesday, June 16, 2015

Another way to adjust level for ABX

Before being converted to (or back to) analog, my system runs digital signals through several processors: the Tact RCS 2.0 (I don't use the RCS, but this is a most useful device anyway with digital selection and volume control), and, in parallel, two Behringer units, one is a 2496 DCX and the other is a 2496 DEQ.

The midrange section runs through the DEQ, and the DEQ has a midi control feature which lets you select presets.  So all I need to do is have two presets which set the overall gain differently but are otherwise identical, and use midi to switch the presets exactly when the amplifiers are being switched.

It wouldn't be a bad idea at all if all ABX switching units were also midi triggered.  Midi was designed for doing stuff like this.  The typical way of controlling amplifiers with a  12V DC trigger signal is relatively limited.

Quite awhile ago, when I first got the QSC ABX box, I did something very much like this.  I used the ABX box to switch a DC signal on and off.  The I bought a midi controller box which triggered a pre-programmed midi signal when an voltage changes.  I programmed the DEQ with a "normal" program and a "reversed polarity" program, which could be selected by the midi transmitted by the midi box.  Then I could switch polarity by remote control.  All I'd need to add to this would be to trigger another function to change the amplifier relays at the same time.

I remember the sense of accomplishment at getting it all to work, but little else.  I did not find polarity to be particularly audible in ABX testing, and put very little effort into testing after the challenging equipment setup was done.  And so it often goes…

But anyway, controlling levels this way means no signal degrading attenuation box, no worries about high frequency roll off or phase shift, no worries about microphonic or noisy potentiometers, etc.

Monday, June 15, 2015

Surprise Discoveries

Little can top the discovery recently that my Krell amplifier is working OK.  But the discoveries continue.

Last weekend I wired up a second set of speaker wires and terminated all speaker wires on the speaker end with banana plugs, so it only takes a minute to switch from Krell to Aragon or back.  In all the time that I have owned both amplifiers, I've never had the capability to compare either amplifier quickly since previously I only had speaker wires running to the center amplifier position, and they wouldn't reach the side, and moving the amplifiers around is a big job.  Attaching wires to the speakers is a very touchy and difficult job which seems to take longer than the 30 minutes or so it actually does.  I manage to insert the wire into the small hole in the binding posts and then tighten them down.

Since wiring the new cables, I've been just plugging either set of wires into the 5 way binding posts in back of the Acoustats.  My plan had been to have a short stub wire screwed on the the Acoustats with bananas on the other end to connect to the amplifier cable bananas--because I recall bananas slowly falling out of the Acoustat posts.  But the new bananas fit quite tightly into the Acoustat posts, and when disconnecting the old wires I discovered that some of the wires I had screwed onto the Acoustats a couple years ago were already loose anyway--the very thing I had been trying to avoid with a direct wire connection.  I wonder how that was affecting the sound.  You just can't make a permanent friction connection to loudspeakers!  Perhaps speakers should have hookup posts that you solder your cables to.

BTW I continue to use 16 gauge zip cord, though it does happen to be either Monster or PureSound wire.  It's actually hard to beat 16 gauge zip cord for "neutrality."  Thick wires have far more inductance, creating relative loss at higher frequencies.  That's far more important than a fraction of a dB in spectrum-wide SPL loss.  Few, very few, audiophiles understand this--that inductance is a key factor in speaker wire.  The solution is to weave multiple conductors as closely as possible, and many exotic wires do this.   But I also get away with 16 gauge wires because my wires are only about 5 feet, at least to the center amp position.  Somewhat unfairly to the side amp position, the Aragon wires run about 9 feet.  The Monster and PureSound zip cords I have use a polyethylene dielectric around the conductors, which protects the copper wire from coming in direct contact with the vinyl outer cord.  Polyethylene is one of the best cable dielectrics and vinyl is one of the worst, and you get most of the benefit from a thin layer.

My plan for speaker wire is ultimately to switch to a 4 cross wire such as Canare 4S11.  By having 4 conductors carrying signal in both directions, inductance is enormously reduced.  That also reduces the external magnetic field, and even Blue Jeans Cables says that is useful under some circumstances.  With all my jumble of wires, I fit one of those circumstances too.  Canare also uses polyethylene dielectric, as most serious or professional cables do.

Switching back and forth a few times between the amplifiers has only reinforced my previous opinion that the Krell is better sounding.  Not only does it seem to have more solidity, a more infinitely layered depth, it sounds more pleasant and relaxing, whereas the Aragon can have some edginess.  I attribute this to the no-feedback output stage of the Krell, and since the back energy from the speaker is handled by the class A output transistors and doesn't get looped through the whole amplifier as would happen with normal feedback.  This could be pure fantasy.

But another discovery was that neither amplifier was balanced to anything like 0.1dB.  I made a nice 880Hz -20dB test tone using sox, and then played it at -6dB.   On the Krell, the left channel was 0.33dB higher--not great.  On the Aragon, the left channel was 0.58dB higher--completely unacceptable.  Reversing the connections on the Aragon improved things considerably.  I'm now using the right side of the amplifier to amplify the left channel and so on.  That reduced the imbalance to 0.14dB on the Aragon, with the left channel still higher.  The fact that I was able to reduce the imbalance so well this way suggests that about half of the imbalance is in the DAC output and the other half is in the amplifier.  Given that the imbalance tilts left even when run through the opposite amplifier channel indicates that the DAC has more imbalance than the amplifier, so that even run through the opposite channel the amplifier can't cancel the difference--though it does come close.  I can't just reverse the inputs to the DAC--since it's an interleaved SPDIF signal.  Each amplifier uses a different DAC, the Aragon is running from the Onkyo RDV-1 and the Krell is run from the Audio GD DAC 19.

I have no potentiometers in the system at all.  The level was reduced to -6dB digitally, and the 880Hz tone was synthesized at exactly -20dB.  The midrange signal may be reduced very slightly by the 80Hz crossover (also digital).  The outputs of the DAC's feed the amplifiers directly.  The differences must come only from the DAC's and the Amplifiers themselves, and previously I would have imagined them as being well within 0.1dB.

The voltages after adjustment were: Krell 1.132 left and 1.092 right, Aragon 0.940 left and 0.925 right.  Before swapping channels, the Aragon outputs measured 0.970 and 0.906.  I suppose some of the left-right difference could also be in the loads that the speakers present to the amplifier.

It was also surprising that the Krell with it's DAC has more 1.6dB right/1.4dB left gain than the Aragon with it's DAC.  (Much of the difference could be between the two DAC's).  I had believed it to be the reverse…forgetting the DAC change I had even dialed in a 1.7dB reduction when switching to the Aragon (remembering that as the difference--perhaps wrongly).  The Krell did sound more solid…but I figured it was the character  not the loudness.  Well you can see here that not only are all my judgements so far wrong--they should probably never be believed!

Fixing the imbalance on the Aragon made it sound a little more like the Krell.  The image was better focussed and the depth seemed to have more layers.  The Aragon still was a bit edgy compared with the Krell.

This was not something I observed when doing the last room eq measurements because basically I don't pay attention to differences less than 1dB, and I don't trust my ability to measure balance that well acoustically.

I decided to take a chance running both amplifiers through Insteon on/off modules, which are rated at 15A and 1800W.  That should be OK though Krell advises only a direct wall connection.  Both amplifiers also go through a single wattmeter so I can keep track of how much power they are using.  The Krell typically uses 550-1200W, typically 750, and the Aragon uses 150-200W.

Two of my Insteon wall keypads, in the kitchen and in the lab, have two buttons for the amplifiers.  The buttons light up when the amplifier power is toggled on.  When I turn on the power for the Krell, it goes into standby, but I can fully turn the Aragon on or off from the keypads since I'm leaving the amplifier switch turned on.  This makes it much easier to keep the amplifiers from using too much power, since I can turn them off walking away from the living room in either direction.

Despite it's monstrous power and looks, the Krell is very soft on the power.  When turned on it only slowly charges the capacitors, using a slow-start circuit that is really slow and deliberate.

I've determined I need about 2 degrees setback for the thermostat in the living room when I am running the Krell.  This prevents the AC from super chilling the rest of the house to keep the living room cool.

After a surprisingly good recording session turned up by surprise the previous weekend, I had all the bedroom side equipment disconnected so I could remove the Masterlink for the recording session.  As I started moving equipment back last Sunday, I started thinking about hooking up the "new" Sonos zone player I rotated out of the living room after I bought a new zone player the previous month.  The new one was placed in the most critical spot (living room) with the old one renamed for being the second master bedroom unit.  It started to occur to me I had only one more Cat6a connection in my bedroom wall panel, and I already had many other ideas as to how I would use it.  I'd have to wire the new Sonos in daisy chain to the first master bedroom zone player.  But while I had the second bedroom connected with a UTP wire (all the ceiling wires are STP) to avoid hum, I used the shielded wire in the bedroom.  Wait!  Maybe that was causing the hum I had noticed in the previous few months, and had me worried that my Parasound HCA-1500A was going bad.

I managed to find one remaining 1ft Cat6a UTP wire and a Cat6a shielded coupler, and connected it just before the already-in-use zone player.  The hum completely disappeared!  Now I know I should not be running shielded ethernet wires to stereo systems.  I'm thinking I'll also use an unshielded Cat6a for the connecting the two zone players in the master bedroom, since that would also avoid a ground loop.

Tuesday, June 9, 2015

Other ABX testers

I'm very distressed that PCABX disappeared before I even had a chance to use it.  I remember eagerly viewing the PCABX website for years and looking at all the hearing tests I could perform on myself.  And now it's gone.  Even if it were there, I suppose that if the author hadn't updated it for a later version of windows, it would be useless anyway.  (I hate endless OS updates.)

I've seen people recommend Foobar, which has an ABX plug in.  This is apparently The Standard at HydrogenAudio.com which distributes Foobar.  But I have no reason to want to mess with Foobar, giving that I primarily use Macs.  I don't need another music browser given that I have Sonos and iTunes already.  I expect using Foobar to run ABX is first climbing the hill just to learn how to use Foobar, then finally, if I ever get that far, learning to use Foobar/ABX.  And all using an OS I can barely stand to look at and rarely use and would never use regularly for music.

Here's an essay on ABX which mentions the ABXTester by Takashi Jogataki.

It can be downloaded from the Mac App store!

I ran into a cross-platform Java ABX called ABX/Shootout-er that looks very interesting.  It runs anywhere with Java, including linux.

Here's an Open Source (Gnu GPL) ABX tester.  The author doesn't know if it works other than on linux systems using ALSA.

Here is Simple ABX Tester for Android.  It allows the comparison of any two audio files as the second mode.

Here is ABX Audio for Android for $0.99.  It defaults to letting you compare any two files on your phone.

Here is a discussion about building hardware ABX comparators.  One poster says he has built several, but they were too expensive to make a kit.  He's interested in an open design, where anyone could build the required switch boxes to go with a computer controller which he could sell.  YES!!!

SOME of the links on David Carlson's site still work.  (The link for NEW ABX Comparator available October 1999 does not work, but fortunately all the other links do seem to work, and there is still interesting information here.)

Arny Krueger talks about making blind test CD's.  When that essay was last updated, I had already done this, making a set of polarity test CD's for testing by my friend George Louis, the self-described Polarity Pundit.  He took a year to finish the test, but the results did not reach p < 0.05 significance.  I literally bought a Mac laptop in 2010 specifically to enable me to make the tests for George while I was staying at his house.  He envisioned becoming the distributor of these tests to help people learn how to hear whether polarity is correct.  The Mac has become part of my computer collection and is still very much appreciated.

Here is a great discussion on ABX testing at AVS Forum featuring Arny, a well known subjectivist named Amir (who is a moderator at What's Best Forum and almost universally well regarded there (a place Ethan Winer has been banned), but less so at AVS forum), several newbies, and old hands.  Arny continues to say the best place to get started with ABX is to download Foobar2000 and the ABX plug in.  Here he also recommends the audio test files from Ethan Winer.

Following that link back to Winer, he used these test files in a video recorded seminar on audio myths.

I'm not sure if I linked this above, there is also the Lacinaot ABX software written in Java.  This should be the same as the Shootout-er above.

Friday, June 5, 2015

Adjusting level for ABX

My QSC ABX box does have a built-in potentiometer system, though I'm not sure I want to use it for a whole host of audio perfectionist reasons.

Instead, I want to use a small passive attenuator box right before the power amp.  This will allow the high impedance output of the attenuator to have as short a cable as possible, perhaps 1f or 0.5M.  That's important because with a given amount of source resistance added by the attenuator, an RC network is created, and the longer the cable the more the C and the lower the frequency at which rolloff or phase shift of a particular magnitude appears.

I also want the passive attenuator impedance to be as low as possible.  5k sounds good.  It shouldn't be too low, such as 1k or lower, because that often causes more distortion in the source that feeds the attenuator.

The potentiometer should NOT be stepped as it will be necessary to adjust to 0.05dB or better.  Fake steps might be OK but I'd rather avoid them.  Wirewound may add inductance (but might actually be OK).  Two mono pots and not stereo (though two stereo pots would also work).

I've seen rankings like this from cheapest to best.

0) Fake Blue Velvet pots from China (this may be what I got on ebay a few years ago).  Some say these aren't too bad, if not as good as the genuine article.  Others say nothing like the real thing.  Said to be counterfeited more than Gucci handbags.

(Note: Even real Blue Velvet may have fake "steps," so I'm not sure I'd want it at all.)

1) Real Alps RK27 Blue Velvet from trusted source (e.g. PartsConnection)

(Note: Black Velvet is a real stepped attenuator.  I think Blue is fake steps.  I'd rather have no steps whatever.)

2) Nobel AP25

2.5) Bourns NOS, Allen Bradley NOS

3) TKD

Here's a Vishay 0.75W 26 turn bulk foil 5k trimmer.

Wednesday, June 3, 2015

ABX comparators

A short review here of the QSC ABX comparator.

The latest and greatest ABX comparator is now being made by Audio by Van Alstine.  AVA is a long established perfectionist audio company in USA that is most know for its preamplifiers.

David Clark invented the ABX comparison method around 1980 and published it in JAES in 1981.  The ABX method addresses long-standing objections by audiophiles to objective testing methods by permitting the Subject to compare the X (randomly chosen from A and B) against A and B as long as they wish and switching among them as many times as they choose before making a decision.  The decision is a forced-choice objective question of whether X is A or B so there are only two possible answers, objectively verifiable (and not a "which is better" which might rest on multiple shifting dimensions).  The purpose is specifically to find whether an audible difference even exists, since well before the 1980's many engineers and a number of famous audio reviewers had claimed that electronic components such as amplifiers and preamplifiers were already good enough that no audible differences existed between objectively good ones, while others claim to find huge differences--often favoring components with less good measured performance.  The ABX comparator was created expressly to investigate such claims regarding whether differences are actually audible.  (Of course, to true-believing subjectivists who believe in their previous sighted tests, to them it is axiomatically  In Your Face and useless.  But true belief is the opposite of science.  Science requires blind testing.)

From 1981-87 an ABX Company made ABX comparators (apparently not many).  I believe this company was run by David Clark and his friends in the SWMWTS audio society in Michigan, one of the few (along with the Boston Audio Society) which favored scientific methods.

QSC was a company that started by making instrument amplifiers for musicians but graduated to mostly making general purpose professional amplifiers (for musicians, PA systems, etc.).  Generally such amplifiers were inexpensive for each watt of power produced, and especially compared with the exotic and expensive amplifiers used by well heeled audiophiles.  So it stands to reason they would be interested in a comparator that would show that their amplifiers sound just as good as much more expensive audiophile amplifiers.  This is to suggest why they were motivated to build an ABX comparator in the absence of overwhelming customer demand, and not to suggest they "cheated" in any way.  I assume and it has always appeared that they made the best comparator they could, and it was relatively high priced compared with their amplifiers.  It appears that their comparators were designed to meet an emerging international standard.

The QSC ABX comparator came out in 1998 and QSC likely made more comparators than anyone else.  The manufacturing run has been said to be 200-300, and it does not appear than any more were made after that.

This appears to be a product type which most audiophiles don't want to bother with, and in many cases they might understandably positively hate the idea of having their sighted comparisons shown to be weak or wrong.  They are satisfied with their sighted and unscientific comparison methodologies and/or they fear that testing devices themselves will corrupt their highly purified systems (and/or the latter may be used as an excuse initially).

There exists a smaller (?) group of audiophiles who embrace ABX testing and it's known history of ubiquitous negative results which tend to discredit many audiophile claims, but this group of audiophiles may not generally spend kilobucks on equipment like this either.

I think this is sad…and I have always felt every audiophile who contemplates changing electronic equipment in their lifetimes (which means ALL) should own and use some kind of DBT device.  (I own two QSC ABX comparators, one is on permanent loan to a subjectivist audiophile friend who has never used it.  Sadly I haven't used it much either.  And for audiophiles of perfectionistic bent (like myself) there is no reason why an ABX comparator can't be made or modified to the highest quality level.

Using a comparator is not necessarily identical to performing an ABX test.  Actually I have mainly used my ABX comparator simply to perform sighted tests with quick switching.  If I don't think I can hear a difference in a sighted test with quick switching…it's not worth bothering with an ABX test as I'd only be guessing.  Or sometimes I want to make a few blind tests.  I don't always need statistical verification.  If I feel as though I'm guessing…I don't need to go any further.

What are the arguments that can be made AGAINST an ABX testing methodology?

A Negative Result is Not Proof.  (Of course, experts already know this, but sometimes exuberant objectivists forget.)  A negative result in a particular ABX comparison does not prove that an audible difference does not exist between items being tested.  It only judges the evidence presented in a particular testing run.  There is a long list of reasons why a negative result may have been obtained even when there are audible differences:

1) Not enough trials were predetermined to reach significance.  Note: the number of trials MUST BE predetermined in a classic ABX test and most statistical methods.  Hard-to-hear differences require larger numbers of trials.  Of course, subjectivist audiophiles believe nearly all the differences they hear, even the ones from their latest tiny tweak, are enormous and life-changing, so they overestimate the audibility of effects (or imagine them) and don't think they need many trials.  In tests I have conducted, one subject had no doubt up front that he would get every single identification correct so the minimum number of trials would suffice, and he didn't want to bother with more than that.  This overestimation of effect itself has a Bayesian interpretation which is not kind to the subject's POV when he is shown wrong.  In fact, it adds to the evidence the effect is imaginary in it's effect or at least much more limited than believed….therefore requiring even more trials to really confirm next time.

2) Not sufficiently trained or capable listener(s) (if the finding is to generalize to others or the same person in the future).

3) Listener lacking patience, motivation, or in bad mood and therefore unable to do test correctly.

4) Musical selections not completely representative of what is possible.  Or switching simply not done in a way as to display the possible differences.

5) Equipment errors.  (Of course, this is what subjectivist audiophiles have almost always argued when confronted with a negative result.  In particular, the comparator itself is blamed for obscuring the differences.)

Errors 2,3,4 are often discounted by advocates of ABX.  They say "well, they had all the time they wanted," etc.  But that kind of dismissal is inadequate for science.  In real science, we must take 2,3,4 very seriously even when Subjects believe they did OK.  This is not a contest to see who is more manly.

The Peter Aczellian "All Decent or Better Amplifiers Sound the Same" does not come from any one ABX test, but rather over 3 decades of failure to produce an ABX or equivalent test showing the reverse (with countless ABX tests having been performed--mainly by people who believe much like Aczel).  And in that, I think he has a fairly good case (though at the same time, another contrary argument, that an amalgamation of endless poor tests does not equal one good test, arises).

But where I have to come down is at the very nature of scientific truth.  It is not ever absolute!  It is conditional and contingent and endlessly subject to revision.  So I do not believe that even decades of research has "proven" anything except that a contrary proof has not yet arisen.  But obviously what decades of ABX test failures to show differences has robustly shown is that "If Audible Differences Exist, they Must Be Rather Hard to Hear or Very Inconsistently Heard."  This actually is very important in itself and not a meaningless quibble with TBS's.

So ultimately it's a matter of the degree and effect strength.  It is actually very hard to discriminate between weak effects and nonexistent effects.  And that's where we are now in The Big Question.

If you wish to relate this back to scientific philosophy, the basic problem is that an open ended question such as "What Can Humans Hear" is not a scientific question.  The conditioned question "What Can Humans Hear Most of the Time" is a scientific question.  Routinely such conditions are left out, but leaving them out without understanding they are, finding in science the solution to unconditioned questions--therefore all questions--that is scientism, not science.

So then suppose you are the designer of the new ultimate audio system, intended to actually be (but not necessarily make people think) the best ever, and you have an unlimited budget.  Do you design your amplifier with approximately 0.5% distortion because anything less than that has not been shown to be audible in ABX testing?  Hell No!  You do the best that you can in every way.  You throw every idea at it that doesn't conflict with a more well established idea.  You make it perform as objectively well as possible, or at least far better than needed, while not breaking suspected audiophile superstitions more than necessary either.  You don't ignore things merely because they haven't been proven necessary.

This is what artists and competitors do, within their budgets, within their capabilities, they don't hold back, they do all they can.

The Objectivist Triumphalists claim to represent the ordinary cost-conscious audiophile.  They want to get the best value, and not waste their money on the unnecessary.  Audio is simply the machinery that enables people to enjoy recorded music, they say.

Then look at their systems and how they spend their time.  It shows as much as music something else--a cause, a hope, a body of shared understandings.  Much as it is to subjectivists.  I would certainly side with the objectivists has having the better understandings nearly complete, but not fully.  And however misguided in ways…subjectivists have good times as much as others.  There is far more to life than truth.

Clearly audio itself is a human art, as well as human society, not merely science or engineering.  It should serve music, but what all it might serve cannot be known or studied by science.

This is not that many shouldn't rely on cost-conscious and scientific engineering when as it is called for.  On tap if not on top.

Tuesday, June 2, 2015

ABX amp switching

I'm now planning to pull my QSC ABX box out of storage to serve as my amplifier switcher in the living room.  AFAIK it has 30A relays for speaker switching, and I have verified that it uses binding posts (and not XLR's, as I had been thinking) to attach speaker connections.  Given that QSC was making professional power amplifiers with many hundreds of watts of power, their ABX box ought to be able to handle fairly high voltage as well as the impressive 30A, and not mess up balanced amplifiers that don't want either output terminal grounded.

I had been planning to use external relays to switch amplifiers, like the Schneider-Magnecraft 199X-12.  I still think this is a cool idea, I can then run minimum length cables to the speakers, with the relays behind each speaker.

Monday, June 1, 2015

Interview with Dan D'Agostino


The Momentum is Class AB, not Class A.  Dan says that Class A buys very little additional performance with very much power and heat.  When asked what is special about his Momentum amplifiers, he points to his 35 years of experience designing good sounding amplifiers.

Very Ironic, given that Dan became famous for Class A amplifiers.  That was certainly how I first heard of him, when I first saw the man himself give a demo of his KSA-50 amplifier at an audiophile society meeting in San Diego, CA around 1982.  It has seemed to me and others that Dan made Class A transistor amps acceptable as they had never been before.  Now the leading seller of Class A amplifiers is Pass Labs.  Nelson Pass had previously been associated with the quasi-Class A that some Threshold amplifiers used (but it was not really anything like real Class A) during the 1970's.  But after starting his own company again around 1990, Pass went right to Real Class A with the Aleph amplifiers, and now the XA and First Watt amplifiers, which are all all real Class A.  Dan being Dan the Class A man made it possible for Nelson to be Nelson.

Dan also suggests there are interesting new ideas in his new amplifier he's not prepared to discuss.

That's fine for his business and customers, though what I like is the interesting ideas, so I'm disappointed he's not revealing more.