Jump to content
  • entries
  • comments
  • views

Rational audio....a subjectivist journey.


Rational Audio....a subjectivist journey.


I began audio as a rational consumer or so I thought. I didn't want to spend money that didn't provide additional performance. Mainly because I didn't have much to spend. I wondered for instance why those Magnepan speakers cost so much when they didn't go down below 50 hz and only went up to 15 khz. So many speakers went lower and higher with less power for less money. With my first real job came a chance to buy a good stereo. I purchased a receiver of modest power, speakers that sounded good from what I had heard, a turntable and cassette deck. Still have the receiver that functions perfectly 30 years later. The tape deck functioned until a couple years ago. The rest was sold over the years. Most of my 'knowledge' was from reading Stereo Review and Audio.


As fortune would have it I later worked with someone owning Magnepan speakers, driven by a G.A.S amp controlled by an Audio Research pre-amp. No it didn't play as loud as my speakers, didn't sound as warm(I of course now know my speakers were overblown and poorly controlled in the low end), but somehow it was clearly far more musical.This system was just flat more enjoyable. I learned about a few more brands and pieces of high end audio equipment that wasn't exactly mainstream from this coworker. Later I took to reading Stereophile and the very strange Absolute Sound at the local library.


Specs couldn't be relied upon to figure out which pieces of equipment sounded the best. Plus the soon to be released CD despite great specs far beyond any LP system clearly had some problems. It just didn't sound fun to listen to like a fine analog rig. I fell under the spell of subjectivists. The ear was the final arbiter of worth. The funny Absolute Sound discussing the problems with digital sound and which wire was the best didn't seem so funny anymore. A newspaper ad listed some Maggie 2C's at a price I could afford along with a Carver Receiver. I eventually made a deal for those as my receiver didn't have enough power for the Maggies and the Carver did. Funny that I got that Carver with the Maggies. Super specs, lots of power, an FM section that was quite remarkable in many ways. Yet it did not sound that great either. Not like my buddy's G.A.S. amp.


Not long thereafter a pair of really cheap Acoustat Two's were in the local trading paper. I had read of them in the Absolute Sound, and these were honest to god mythical electrostats. I was immediately smitten by the sound (and prefer electrostats to this day), but wondered if my Carver receiver would play them. The owner said no way. I went back with the Carver to see. I must say it played them quite admirably. The Acoustat owner found it hard to believe. So I ended up with the Acoustats. The Carver really had power to play this difficult load though the sound quality could be better. In time I found a McIntosh 752 I could afford. Though less powerful than the Carver it played the Acoustats well enough and sounded far nicer. The Mac owner threw in some nice speaker cable which I thought helped as well.


So now I had a receiver and some Maggies I wasn't using. Put an ad in the paper and eventually sold them to someone that would become a good friend. He also knew several other audiophiles in my town. He didn't buy the receiver however. But getting to know other audiophiles was important in getting to hear other serious highly enjoyable sound systems. I was able to hear how various systems did things well though none did everything well. Hearing horns, LS3/5a speakers and speakers like Thiels.


Skipping ahead rather than detailing every step up I made, in time I owned Quad ESL63's which would eventually be driven by tubes. Still making decisions subjectively upon my opinion of the sound of CDP's, DAC's, pre and power amps, cabling all of it. It still didn't sit easy with me however. That so much seemed unexplainable. Some of the equipment made large improvements in sound quality. These seemed magical. All good hifi rigs of course should seem magical. Only I didn't really believe in magic. In time I went to the bother of taking a junior college program in industrial electronics just to learn more about all this stuff. I earned a two year degree in electronics. I had seen too many explanations in high end mags that simply were wrong though espoused as if from some authoritative understanding. I wanted to know how this really worked.


Prior to this high end equipment seemed mysterious and magical. Special designers having figured out the mysteries of power amps, and component selection magic made musical high end equipment, while mass market stuff was like a paint by numbers enterprise designed by specs selling inexpensively while sounding worse. When I learned basic electronics it began to loose its mystery. As I applied what I had learned to the particulars of audio more mystery was lost. Not to say all such equipment was a farce or bilking consumers. Much of it was made to a much higher standard, and truly provided much better sound quality than mass market items. Even so much of the marketing and magazine writing really was so much bunk. Claims for what made one item better than another was actually no explanation or was simply wrong headed.


I did begin to apply my new electronics knowledge to modding equipment. Replacing components with better quality components which did seem to improve sound. My Quads were driven by tube amps by then. CD transports driving DACs had improved enough I thought them musical enough to replace LPs. I found pre-amps unneeded if your DAC is the source. I built my own switched resistor passive pre-amps eventually. One of my friends and I improved components in the VTL tube amps he and I owned, things like better caps (MIT's), and resistors (Vishays). Then we rewired them from Ultralinear to triode connection. Not to mention buying good sounding replacement tubes. At this point I learned enough basic electronics to alter equipment, understand how the circuits worked, yet made all my decisions subjectively as to whether a given change of circuitry or componentry was a step forward or not.


After these mods a few of my friends thought the trioded, upgraded, modded VTL was the most musical sounding amp they had heard. We took them over to their houses to try them out on a number of different systems. It seemed to convey rhythm, resolve space and fine detail better than other amps while being highly musical. Following words and music was simply easier. I had become addicted to regularly improving the sound of those fine tube amps. I at that time didn't know of any likely useful changes to make. I could think of a few that would be highly complex and expensive. I began to wonder, “how much better could it get?” I wondered did my amps at that time convey 99% of the music in the input signal or only 50%. Was there little left to gain or much more to be coaxed out of it?


I also had a Spectral DMA50 at that time. One of the few really good sounding transistor amps I had heard. Subjectively I would say it was 60% as good as the VTL at that time. An idea began to form as a way to figure out how much was lost in the VTL from input to output. The Spectral was good enough sounding to make good judgments in comparing upstream changes. What if I loaded the VTL with a power resistor at the output, then used high quality resistors to step down the voltage so the resulting signal was exactly unity gain. Then I fed that into the Spectral and listened to it. If the VTL was perfect straight wire with gain, feeding the Spectral I should hear no difference with it between my DAC and the Spectral amp. If the VTL was passing 50% of the information in the input signal, then the Spectral was good enough I could hear the loss in quality and musicality. I would have an idea how much room for improvement there was.


I did that experiment. The results were quite obvious, and completely unexpected. With a VTL between the DAC and Spectral power amp the sound out of my Quads simply sounded like the VTL. It was spacious, liquid, naturally smooth sounding, with glowing inner detail, and a highly musical quality. I was hearing exactly what the VTL always seemed to sound like. How could this be? I knew the Spectral lost some resolution and fine detail yet I was hearing that with the Spectral driving the speakers. In time I thought the sound was a little leaner, and the treble a little recessed vs the VTL, but only a little. The basic quality was still of a highly musical triode amp. I even quickly knew why. I knew the VTL because of the odd impedance of the Quads had its output impedance react with the speaker to give a 1.25 dB bump centered around 50 hz and covering nearly two octaves and the transformer from measurements I had done would have a rising response just above 20khz, but it was wide enough it lifted the last audible octave about .5 a dB. The Spectral was pretty much unaffected by the load of the Quads.


I decided to reverse positions. I loaded the output of the Spectral with power resistors and attenuated the result to achieve unity gain and fed it into the VTL. I could not hear any effect of the Spectral being in or out of the circuit. I listened for several days and put the Spectral in and out a number of times. Sounded the same either way. The Spectral should be straight wire with gain. It had DC-1 mhz bandwidth, a very high slew rate, S/N around 90 dB, very low distortion and excellent channel separation. It should be audibly invisible.


So not an answer I anticipated. The only reasonable conclusion was the Spectral did little if anything to harm the signal. While the VTL for all of its apparent superior sound was subjectively superior because it colored the sound in a pleasing manner. No other idea made sense.


I tried this same experiment with a few more quality transistor amps. No others quite matched the Spectral for being invisible. Some came pretty close some were surprisingly audible as being brittle sounding or having misty grainy highs. All when driven by the VTL however sounded at least 60% or more like the VTL. You would hear the qualities of the VTL overlaid with whatever inadequacies were in the various solid state amps.


Over a few years I had argued foolishly it now seems with people like James Johnston (known as JJ) on various internet forums or newsgroups. I eventually read a couple of the textbooks on psycho-acoustics he recommended. Those have plenty of information useful in home audio. Some things are readily explainable in scientifically proven terms. JJ wouldn't claim all solid state amps sound the same as he knows the basic specs aren't enough. But specs well applied are usable. Much of subjectivist audio I think has gotten out of hand . Not all audible phenomena can yet be fully explained, but quite a lot of it can. Much of this knowledge is regularly ignored by people who should know better. I think we could get further along and get better sound for less money by using this knowledge rather than ridiculing it.


Recommended Comments

I have not done any test like this on amps, but I have on digital and I did work in the R&D department at Garrard in the 70s where I had a lot of opportunity to measure, and listen to the effect of, various turntables and aspects of turntable design.


I had a similar experience.


Even early digital (I tried the Sony PCM F1, IIRC) in the tape loop was very close to transparent.


Vinyl has several inherent characteristics which are unavoidable but can be tuned to sound nice (though never accurate), as in the amp comparison.


I have not tried yet myself but I know others who have recorded their vinyl to digital, so called needle drops, which sound just like the LP.




This does create something of a dilemma, IMO, since hifi as a hobby which is dying out as nowadays music is usually listened to in the car or a portable. The enthusiasts who remain contain a huge number of people convinced that vinyl and valves are better, rather than just can be made to sound nice...

Link to comment

Yes, I have digitized a few LP's. Digital is not a problem for sound quality. Back over a decade ago I had become bothered by most analog pre-amps. I got an analog to digital converter which I used to convert all my analog sources to digital before sending the signal to a DAC. If it wasn't fully transparent it was awfully close. Though that AD unit would do 96 khz my DAC at the time would only do 48 khz. I am not so sure we need anything more than 48 khz at 24 bit for home reproduction. I certainly see little point in going beyond 96 khz.




I have written comments here before how several of my friends got together recordings of the same music done on LP, CD and reel to reel tape. The LP was always the odd man out. Sometimes LP was seemingly better, sometimes not, but it was always the one sounding very different than the other two sources.




I have nothing against making something sound better to your ear. If you mistake that for higher fidelity when it isn't you will end up chasing ghosts. I would like to see more effort in consciously altering sound for better subjective sound quality. That can be done in DSP quite inexpensively if we know what we want.

Link to comment

First, a really, really fine post. I wish I had been as diligent at following my curiosity about audio electronics as you were.




I am not so sure we need anything more than 48 khz at 24 bit for home reproduction. I certainly see little point in going beyond 96 khz.




Keith O. Johnson, as you know the lead designer on that Spectral amp, does his recording and playback media at 176.4 or 192kHz, so perhaps there is some point in it. I have no personal exposure to the issue, as my current DAC is limited to 48kHz and I don't have the theoretical background (other than a very, very rudimentary exposure to Shannon/Nyquist) to understand the important electronic and audio factors at play.




In a couple of weeks, however, I'll be receiving the inexpensive ($450 with USB, $350 without) Schiit Bifrost DAC. At that point, I look forward to listening through its various inputs to a Linn download or two available in both 24/96 and 24/192, to see whether I can consistently hear any differences in sound quality.

Link to comment

After reading your post, I am left with the feeling that I want more info.




Could you provide some links that would complete my quest of expending my knowledge of HiFi components.





Link to comment

"I am not so sure we need anything more than 48 khz at 24 bit for home reproduction. I certainly see little point in going beyond 96 khz."




Call it the law of diminishing returns. I somewhat agree with you. I have lots of 24/48 that is pretty much indistinguishable from higher res sources. Great read.

Link to comment

  • Create New...