Sound signatures in integrated amps?

[mtrycrafts]

1. Subtle timing cues are smeared when nfb is used in a circuit.

2. In nature sounds are richly embedded with even order harmonics. A single ended circuit is more adept at reproducing the harmonic structure of live music and real sounds. A p/p or class AB amp is less adept at reproducing the natural harmonic structures, and squashes whatever distortion exists out to much higher order at lower levels, something that just doesn't exist in nature and is perceived as artificial.

3. Low order harmonics are involved with human perception of the acoustic space of a performance (or reproduction).

These could be completely off base, feel free to tear 'em to shreds. (And thanks to all for the lively and civil discussion!)

Timing cues? Boy, you are confused about harmonics in nature and harmonic distortion created by such amps.
Nothing really to discuss, is there?

 

[PENG]

1. Subtle timing cues are smeared when nfb is used in a circuit.

2. In nature sounds are richly embedded with even order harmonics. A single ended circuit is more adept at reproducing the harmonic structure of live music and real sounds. A p/p or class AB amp is less adept at reproducing the natural harmonic structures, and squashes whatever distortion exists out to much higher order at lower levels, something that just doesn't exist in nature and is perceived as artificial.

3. Low order harmonics are involved with human perception of the acoustic space of a performance (or reproduction).

These could be completely off base, feel free to tear 'em to shreds. (And thanks to all for the lively and civil discussion!)

I did read quite a bit of what you cited here, and some, including how the talk of inaudible but perceptible high (as well as low) frequencies/harmonics. Again, instead of wondering whether or not those things matter, I try to focus much more or enjoying the music. I do believe that those things may in fact matter somewhat, though not a whole lot.

 

[ski2xblack]

I did read quite a bit of what you cited here, and some, including how the talk of inaudible but perceptible high (as well as low) frequencies/harmonics. Again, instead of wondering whether or not those things matter, I try to focus much more or enjoying the music. I do believe that those things may in fact matter somewhat, though not a whole lot.

I'm with you PENG, it's all about the music.

And mtry, I'm quite disappointed in your response. Given a golden opportunity to share your vast knowledge, and all you do is throw out petty insults? That was some prime woo-ish red meat I offered up! I suppose you think that the use of nfb has no effect on an amps sonic signature? It does, and specifically it smears subtle timing cues which results in a collapse of soundstage depth, giving the flat, two-dimentional presentation. It measures better, but sounds artificial. (But what do I know, I just build my own amps to my taste and have first hand experience, not some outdated engineering textbook definition of what 'hi-fi' is from the dawn of ultralinear p/p circuits.)

 

[TLS Guy]

I'm with you PENG, it's all about the music.

And mtry, I'm quite disappointed in your response. Given a golden opportunity to share your vast knowledge, and all you do is throw out petty insults? That was some prime woo-ish red meat I offered up! I suppose you think that the use of nfb has no effect on an amps sonic signature? It does, and specifically it smears subtle timing cues which results in a collapse of soundstage depth, giving the flat, two-dimentional presentation. It measures better, but sounds artificial. (But what do I know, I just build my own amps to my taste and have first hand experience, not some outdated engineering textbook definition of what 'hi-fi' is from the dawn of ultralinear p/p circuits.)

Disappointed by mtry? I have just been alerted to this thread full of unsupported bilge.

Where to start?

First not for the first time lets take on SETS.

Whichever way you call it these are lousy amps.

They have limited power and nowhere near enough for modern digital sources.

They have limited frequency response and turn all square waves into sign waves.

Worse they add huge amounts of even harmonic distortion. Granted some "Golden Ears" seem to like that. However instead of proclaiming audio nirvana, they should be honest and just say "We like lousy low powered amps!" Anyhow I certainly don't want my Bach and Mozart courtesy of Mesa Boogie.

No lets deal with negative feed back.

What is your evidence that negative feedback causes time aberration. I want hard data. It is easy to test.

Now the design of everything is to an extent trade offs and judgments. However all amplifying devices have distortion. To make an excellent amplifier that has to be addressed. If memory serves me right negative feedback was published by Bell Labs in the late 20s.

So what is is negative feedback. Well it is a comparator system. A portion of the output via phase inversion is fed back to the input, to cancel errors, in other words distortion.

As with everything there is no free lunch. The downsides are reduced gain, a tendency to introduce instability of not applied correctly.

The biggest downside, is that while it is very effective at controlling distortion within the power range of the devices, ones clipping occurs then distortion rises very quickly and much quicker than it would without it.

However judiciously applied negative feedback is a a time honored way of achieving low distortion. That is a fundamental requirement for high fidelity reproduction. The honor for breaking the 0.1% THD barrier has to go to Harold J. Leak in his TL 10, TL12 and TL 20 amplifiers with the judicious use of negative feedback in the late forties.

You might be interested to know that I listen to amplifiers that have +ve rather then negative feedback. The justly famous Quad current dumpers.

Again the advantages and trade offs. The big advantage class A performance without the disadvantages and zero crossover distortion.

The downside is that in order to make the amp stable relatively slow output devices have to be selected. The only consequence of this is to limit frequency response to the audible range. In other words no stratospheric frequency response to please a passing bat.

Now I going to return to the issue of phase and time aligned speakers.

Now as I have said often, good speakers are the sum of countless intelligent compromises. Now if you pursue just one problem to the exclusion of others you create a huge problem for yourself. Time and phase aligned speakers had a huge vogue in the early eighties. B & W even had a design that looked pregnant. A lot of us took the religion without really stopping to count the cost.

Now the first problem is that once you separate driver and add a crossover, you separate fundamental tones from their harmonic content. That is a big trespass and a huge compromise right there with every speaker using multiple drivers

So what are the solutions. A full range driver. Well done these can sound excellent at low to moderate power, Lowther excepted. However there is a problem, especially with deep low frequency extension. The other problem is doppler inter modulation distortion. This latter problem also affects coaxial designs.

Now the only crossover that can be summed to give a theoretical perfect transient response is the first order filter. However this is highly theoretical.
There is huge driver over lap and the acoustic center is not constant with frequency. You also have huge comb filtering issues to contend with.

The drivers must be aligned so that their acoustic axes are lined up. This can never be done perfectly and the time paths for the drivers will be different for each with changes in listening position.

If the tweeter is on top bass driver at the bottom there is a downward 15 degree lobing error. If the other way round there is a 15 degrees upward tilt like in the Dynaudio first order designs.

The next issue is that this really is only practical at relatively high frequencies, as protruding drivers cause bad unwanted reflections from the recessed driver above.

So lets take this speaker I wrestled with for 10 years, to get a superior speaker.

The tweeter is at such a height as to form a direct path to the preferred listening position allowing for the 15 lobing tilt.

Only the top three drivers have first order crossovers as the lower two KEF B 139s are not suitable for use with first order crossovers. The hand over is third order at 180 Hz active, so the 15 degrees tilt is maintained.

The next crossover is 900 Hz. So the calculated time delay is 3.6 inches, in practice it turned out to be less than that. In mounting the 15 degree tilt has to be allowed for.

The next crossover is 5 kHz. The calculated time delay is 0.7 inches, which turned out to be close.

So the array in theory is only good for the preferred listening position. In fact three to four can get close to optimal results. In the old studio it was fine. For this application it works out well for the front row, less so for the rear row. However since this is mostly an ambiance channel it works out well.

At the end of it all however listening tests have shown a preference for speakers with symmetrical lobing patterns. So this technique had never really caught on and is used significantly less than it was over 20 years ago.

The moral of the story is to think carefully before trying to solve one problem to the exclusion of other and creating a boat load more to solve.

I certainly leaned more than a few lessons along the journey with that speaker.

 

[ski2xblack]

Thanks Dr. Mark for the thoughtful reply. Sorry for the unsupported bilge, but it did get you to chime in, so it served it's purpose.

You mention your Quad currend dumping amps again which I find quite intriguing. While they give you class A goodness, what are the potential weaknesses/compromises/trade offs?

 

[TLS Guy]

Thanks Dr. Mark for the thoughtful reply. Sorry for the unsupported bilge, but it did get you to chime in, so it served it's purpose.

You mention your Quad currend dumping amps again which I find quite intriguing. While they give you class A goodness, what are the potential weaknesses/compromises/trade offs?

I don't think there is a significant trade off. The amp has a low part count and no internal adjustments. The circuit can compensate for a wide drift in the value of components with age.

The downside is, as I stated, that the output devices have to be relatively slow to keep the circuit stable. However the only effect of this is to limit bandwidth, so HF response extends to the limits of human hearing and not beyond.

Now Audiophools will tell you have to extend response beyond 20 kHz to get "air". More nonsense from this crowd who have done such a grave disservice to the development of audio.

The fact is you can test human hearing with brain stem evoked audiometry. That is a technique way beyond the comprehension of the compressed minds of audiophools. But what it means is that if you stimulate the ear with a sound you can tell if the brain receives it. This is how the hearing of newborns is tested. So if the brain does not respond there can be no comprehension. That is the end of the story.

I posted about current dumping feed forward amps recently.

I'm glad to note your keen interest in the English vernacular.

 

[ski2xblack]

Thanks for the link to your Quad amp commentary. Why they haven't licensed the technology is a real head scratcher.

I wish I had some peer reviewed support for some of the experiences I've had for your benefit, but I don't. But perhaps you will accept my personal anecdotal experience for what it is. I'm kind of an oddball around here, severely turned off by woo, but with experience that confounds the prevailing dogma.

Regarding SETs, sure, they are severely limited in many ways. If you cater your system to their strengths, however, they make fine amplifiers. I use mine to power the mids/highs on a pair of LaScalas, with a ss A/B amp pushing the bass bins, augmented with a couple sealed twelves w/ 300w AB plate amps. In my use, bandwidth limitations don't come into play because I simply don't ask them to reproduce bass frequencies; limited power doesn't come into play because it only takes 3 watts to get >100db at my chair, which is louder than I ever listen. I get the class A goodness at minimal expense, at the cost of some midrange embellishment, which is a benefit rather than a loss in my opinion, particularly with horns. I would have to spend thousands more to achieve this w/ ss, and possibly never get there. (My other systems are testament to that.)

I realize such a system involving SET amps and high-efficiency horns would make Dr. Mark and others cringe, but it really does provide supremely effortless, all-day non-fatiguing, lifelike reproduction, with 3D imaging to die for. Microdynamic nuances are exposed that other systems just can't reveal. No horny shout or cupped-hands sounds, just pure music. It's something special. Maybe not 'hi-fi' according to the engineers, but damn it's impressive in sheer attention grabbing liveliness and psychological/emotional engagement (which is what I aim to achieve, as opposed to aiming for good on-paper specs). It's capable of subtlety and nuance, yet also able to make your pants flap and plaster you into your chair as in that old Maxell tape ad, and never falls apart when doing so. And it was not expensive, considering its all used or DIY gear. That whole system cost less than a Pass amp I auditioned.

Recently, a musician friend was in town to produce a CD. Two weeks in the studio and it was about complete. Near the end of this, I found my friend up late night, listening to tracks on the SET system. I got up and went downstairs to see what sort of festivities I might have been missing. He was just sitting in the dark listening to his work. I told him that this was the 'toy' system, not to be used for mastering or serious evaluation (several other rooms with conventional systems more suited to that purpose around here). His comment was, "But my guitar [as reproduced on the recording] sounds better in here. That is the Taylor!" It is this sort of experience, coupled with my own, that confounds me. Admittedly, I have more trust for musicians than engineers when it comes to 'good sound'.

Dr. Mark, I agree with your assessment of "air" and high frequency nonsense. If you can bear just a bit more bilge, I feel it just happens to be a good subjective description of the differences I was hearing between different levels of nfb in my SET building experiments. It seemed like there was more 'air' between the source and myself, a greater sense of depth/space to the soundstage the less feedback used. I can only hear to 17K hz or so, and the speakers don't go above 20K, so it clearly is not related to extended high frequency capability, but it was perceptible and I found that I have my specific personal preferences. "Smearing of timimg cues" is just a subjective description of what seems to be going on, and probably a horrible term to use in this forum...unsupported bilge, perhaps, but an attempt to describe a real phenomenon as I heard it.

I generally tune my sytems for a nice, thick presentation at 95db reflective of my own personal Fletcher-Munson preferences, which results in pretty 'flat' sound at normal levels (~85db). At such levels it, and if not listening from within the sweet zone, it's really difficult to tell much difference between tube and ss power, they are more similar than different. The SETs are definitely well within their most linear range at such levels, no where close to clipping. But when in the sweet spot, with eyes closed, the soundstage seems to receed back 30' or more. The conventional amps I tried just didn't do that, and the no-nfb single ended tube circuit did it the best. I can't be sure exactly what is causing it, but it seems to me that A/B amps take something away that class A allows through, with more spatial detail from the recording seemingly arriving at your ears.

Would I encourage anyone else to go down the SET rabbit-hole as I have? Probably not, unless you have a pair of 'Scalas or equivalently high efficiency speakers sitting around, know what you're in for, and realize the limitations going in. That's why I would not encourage the approach generally. But they are not as bad as many here would make you believe. Or maybe they are, and it's just personal preference. All I can say is that I've heard Pass ss class A, I've heard SETs, and I have a distinct preference for their, uh, idiosyncratic sonic personalities, that supremely non-fatiguing silky smoothness and soundstage that extends well beyond my room's actual dimensions. If the Quads have that without the heat, expense, and other compromises, they sound like superb amps. I think I see one added to the collection in the future, for curiosity's sake if nothing else.

Sorry for the lengthy post. I guess the point is that subtle differences do exist between amps, and I have found it rather fun and educational to explore this aspect of our hobby.

 

[TLS Guy]

Thanks for the link to your Quad amp commentary. Why they haven't licensed the technology is a real head scratcher.

I wish I had some peer reviewed support for some of the experiences I've had for your benefit, but I don't. But perhaps you will accept my personal anecdotal experience for what it is. I'm kind of an oddball around here, severely turned off by woo, but with experience that confounds the prevailing dogma.

Regarding SETs, sure, they are severely limited in many ways. If you cater your system to their strengths, however, they make fine amplifiers. I use mine to power the mids/highs on a pair of LaScalas, with a ss A/B amp pushing the bass bins, augmented with a couple sealed twelves w/ 300w AB plate amps. In my use, bandwidth limitations don't come into play because I simply don't ask them to reproduce bass frequencies; limited power doesn't come into play because it only takes 3 watts to get >100db at my chair, which is louder than I ever listen. I get the class A goodness at minimal expense, at the cost of some midrange embellishment, which is a benefit rather than a loss in my opinion, particularly with horns. I would have to spend thousands more to achieve this w/ ss, and possibly never get there. (My other systems are testament to that.)

I realize such a system involving SET amps and high-efficiency horns would make Dr. Mark and others cringe, but it really does provide supremely effortless, all-day non-fatiguing, lifelike reproduction, with 3D imaging to die for. Microdynamic nuances are exposed that other systems just can't reveal. No horny shout or cupped-hands sounds, just pure music. It's something special. Maybe not 'hi-fi' according to the engineers, but damn it's impressive in sheer attention grabbing liveliness and psychological/emotional engagement (which is what I aim to achieve, as opposed to aiming for good on-paper specs). It's capable of subtlety and nuance, yet also able to make your pants flap and plaster you into your chair as in that old Maxell tape ad, and never falls apart when doing so. And it was not expensive, considering its all used or DIY gear. That whole system cost less than a Pass amp I auditioned.

Recently, a musician friend was in town to produce a CD. Two weeks in the studio and it was about complete. Near the end of this, I found my friend up late night, listening to tracks on the SET system. I got up and went downstairs to see what sort of festivities I might have been missing. He was just sitting in the dark listening to his work. I told him that this was the 'toy' system, not to be used for mastering or serious evaluation (several other rooms with conventional systems more suited to that purpose around here). His comment was, "But my guitar [as reproduced on the recording] sounds better in here. That is the Taylor!" It is this sort of experience, coupled with my own, that confounds me. Admittedly, I have more trust for musicians than engineers when it comes to 'good sound'.

Dr. Mark, I agree with your assessment of "air" and high frequency nonsense. If you can bear just a bit more bilge, I feel it just happens to be a good subjective description of the differences I was hearing between different levels of nfb in my SET building experiments. It seemed like there was more 'air' between the source and myself, a greater sense of depth/space to the soundstage the less feedback used. I can only hear to 17K hz or so, and the speakers don't go above 20K, so it clearly is not related to extended high frequency capability, but it was perceptible and I found that I have my specific personal preferences. "Smearing of timimg cues" is just a subjective description of what seems to be going on, and probably a horrible term to use in this forum...unsupported bilge, perhaps, but an attempt to describe a real phenomenon as I heard it.

I generally tune my sytems for a nice, thick presentation at 95db reflective of my own personal Fletcher-Munson preferences, which results in pretty 'flat' sound at normal levels (~85db). At such levels it, and if not listening from within the sweet zone, it's really difficult to tell much difference between tube and ss power, they are more similar than different. The SETs are definitely well within their most linear range at such levels, no where close to clipping. But when in the sweet spot, with eyes closed, the soundstage seems to receed back 30' or more. The conventional amps I tried just didn't do that, and the no-nfb single ended tube circuit did it the best. I can't be sure exactly what is causing it, but it seems to me that A/B amps take something away that class A allows through, with more spatial detail from the recording seemingly arriving at your ears.

Would I encourage anyone else to go down the SET rabbit-hole as I have? Probably not, unless you have a pair of 'Scalas or equivalently high efficiency speakers sitting around, know what you're in for, and realize the limitations going in. That's why I would not encourage the approach generally. But they are not as bad as many here would make you believe. Or maybe they are, and it's just personal preference. All I can say is that I've heard Pass ss class A, I've heard SETs, and I have a distinct preference for their, uh, idiosyncratic sonic personalities, that supremely non-fatiguing silky smoothness and soundstage that extends well beyond my room's actual dimensions. If the Quads have that without the heat, expense, and other compromises, they sound like superb amps. I think I see one added to the collection in the future, for curiosity's sake if nothing else.

Sorry for the lengthy post. I guess the point is that subtle differences do exist between amps, and I have found it rather fun and educational to explore this aspect of our hobby.

The LaScala.

They can sound good in large space. The top horns are certainly on the fierce side.

If you are using the SET to just run the horn, I agree in a home that would take little power and limit the even harmonic distortion, but it will still be there. That SET amp will be rolling off the highs and taming the horn though.

The Quad does give true class A performance, you can see that in Peter's math in his paper that I linked. Peter had a wonderful knack of making his math and science easily understood, even I can follow it.

 

[KEW]

His comment was, "But my guitar [as reproduced on the recording] sounds better in here. That is the Taylor!" It is this sort of experience, coupled with my own, that confounds me. Admittedly, I have more trust for musicians than engineers when it comes to 'good sound'.

A couple of relevant questions:
Is his Taylor acoustic or amplified? If amplified, what type of amp?
If he defines the sound of his guitar after amplification it would be interesting to know how his amp and your SET compare in nature.
If he defines the sound as acoustic, it may be that the immediacy of a pick-up or microphone alters the sound (compared to what he hears at 90° to the sound board) and your particular combination of components "undoes" that alteration.
Please understand, I'm just throwing out notions which might help bridge the gap between this experience and TLS Guy's knowledge.

 

[ski2xblack]

Doc, I woudn't characterize mine as 'fierce' at all. If anything, they are kind of laid back sounding, mellow, very easy going sound. This surprises me as the impedance is quite high up in the treble regions, and the amps should be good to over 20Khz, so I would expect a rising response if anything. Even using ss power, they maintain thier easy going character, only slightly more edgy than with tubes, a bit less soundstage depth, but no glaring differences at all. The speakers are stock except for new caps in the crossovers, and their response falls off above 17K or so. I only wish they looked as good as the pic you posted...as with much of my DIY stuff, once the audio portion is complete, the aesthetic finishing stage faces indefinite procrastination. I also own a pair of old forte ii's, which are slightly more forward, but still a pretty even keeled speaker. Klipsch's rep for strident shouty horns is certainly not based on the Heritage models I own, in which the glorious midrange dominates with remarkable smoothness from top to bottom. They may even have lower measureable distortion than the tube amps pushing them!

Kew-straight up acoustic, recorded with mic. It would not seem his preference was based on similar amp signatures used in production vs reproduction. It wasn't because of the tubes or other reproduction equipment; he couldn't give a crap about hi-fi gear as long as it sounds good. To add context, he enjoys coming over whenever doing recording work to hear his stuff on a variety of systems, with only that one occurrence seeming to explicitly favor the tube setup. It could have been due to his mood at the time, who knows, but it seemed to be a rather telling event.