Do amplifiers really sound the same?

[sergeauckland]

Serge,

you didn't reply to my questions. You just regurgitated your views.

A valve amplifier designed to normal engineering standards, as opposed to one made to be a retro effects box, has not too much trouble. Ask Johnson, Beard, TdP, ... oh, Williamson, Walker, Leak, ...As for the speakers: think about the measurement methods available in 1968. How much inherent smoothing would have been going on back then?

Completely agree with the thrust of your arguement. Vintage valve amps and a few modern ones can indeed meet DIN 45-500. A few can be considered transparent in that they will most likely pass a SWBT. What gets me is that whilst a 50 year old Quad, or Leak or Williamson will pass, many modern ones won't. A few will, Mackintosh for example, but many more won't especially SETs.

One would have though (wouldn't one?) that after 50 + years, if one wanted to design a valve amplifier one could do a better job than those past manufacturers. Clearly not.

As to loudspeakers, I'm not sure of your meaning about smoothing.

Loudspeakers then were usually measured with a B&K pen recorder, where the writing speed determined the amount of smoothing. Using a slow sweep speed and a high pen speed, it was possible to capture a pretty much unsmoothed trace, just as now. Now of course, the software will apply as much or as little smoothing as needed for the purpose, third octave is a good compromise as it more or less stacks up with what we hear in terms of the energy contained in the peaks or troughs. For publicity /brochure purposes, octave smoothing can do wonders for a peaky loudspeaker.....

Looking at test reports as published in HFN or on Paul Miller's web site, too many modern loudspeakers seemingly make little attempt at a flat frequency response even with third-octave smoothing or accurate pair-matching. Many do of course, but like with valve amplifiers, where's the progress in accuracy over the years?

S.

 

[sergeauckland]

So you're saying it's possible to make a valve amp that works just as well as a solid state one?

It depends what you mean by "just as well". It's possible to create a valve amplifier that is audibly transparent. Difficult, because of the requirements of the output transformer. At low frequencyies it needs to be large to avoid core saturation increasing distortion. However a physically large transformer will have more parastics which affect high frequencies. Then there's the issue of phase shift which limits the amount of feedback that can be safely applied.

So, whilst it's almost trivially easy to make a SS amplifier that is transparent, it's difficult, but possible, to do the same with valves. Much like watches. It;s trivially easy to get a quartz watch to keep good time, it's very difficult to do the sdame mechanically.

This brings me to the issue of why bother. If it's easy to achieve accuracy with SS amplifiers, why would one ever bother to try and do it with valves, unless of course transparency isn't the objective.

S.

 

[Werner]

but many more won't especially SETs.

So you assume they are meant to comply to a certain notion of blamelessness, while IMO they really really want to offer a distinctive sound, so as to appeal to a certain market.
Retro effects boxes.

One would have though (wouldn't one?) that after 50 + years, if one wanted to design a valve amplifier one could do a better job than those past manufacturers. Clearly not.

The key is: if one wanted.

As to loudspeakers, I'm not sure of your meaning about smoothing.

Wonder how much smoothing DIN45500 allowed.

of the energy contained in the peaks or troughs. For publicity /brochure purposes, octave smoothing can do wonders for a peaky loudspeaker.....

Now. And then. AFAIK the whole lot was based on self-certification, so any half decent plot would do.

 

[hammeredklavier]

This brings me to the issue of why bother. If it's easy to achieve accuracy with SS amplifiers, why would one ever bother to try and do it with valves, unless of course transparency isn't the objective.

Expensive valve amps were all the range when I was just getting into hi-fi in the early '90s, and I never understood it. As far as I could tell, if you threw huge amounts of money at the problem, it was possible to make a valve amp that sounded like budget SS amp. Then there were various SS amps that ostensibly aped the sound of valve amps... actually, aped them in terms of the amount of heat they chucked out as well, think of the MF A120 for eg!

 

[NRG]

Modern valve amplifiers often have rather more than 3% distortion at LF and HF at anything more than a few watts.

S.

No they don't. You cant tar all 'modern' valve amps with the same tar brush.

 

[Yank]

Looking at test reports as published in HFN or on Paul Miller's web site, too many modern loudspeakers seemingly make little attempt at a flat frequency response even with third-octave smoothing or accurate pair-matching. Many do of course, but like with valve amplifiers, where's the progress in accuracy over the years?

S.

I can't think of many in the 1970s that attempted flat response either, outside of the small sphere of BBC-influenced British boxes.

 

[sergeauckland]

I can't think of many in the 1970s that attempted flat response either, outside of the small sphere of BBC-influenced British boxes.

Those were the only ones that mattered. ]

S.

 

[sergeauckland]

No they don't. You cant tar all 'modern' valve amps with the same tar brush.

They don't all do that, but just look at the reports of valve amplifiers on the HFN (Miller) site. All valve amps show an increase in distortion at LF and HF, some, but I agree not all, exceed several %.

S.

 

[sergeauckland]

Expensive valve amps were all the range when I was just getting into hi-fi in the early '90s, and I never understood it. As far as I could tell, if you threw huge amounts of money at the problem, it was possible to make a valve amp that sounded like budget SS amp. Then there were various SS amps that ostensibly aped the sound of valve amps... actually, aped them in terms of the amount of heat they chucked out as well, think of the MF A120 for eg!

That's been my view of valve amps since the 1970s when silicon complementary pairs like the 2N2955 / 2N3055 became available. Why use expensive valves and expensive output transformers to do what a simple SS amp will do at far less cost, far less electrical consumption and hence heat, far less bother with loading.

Unless of course, transparency isn't the aim, but a "nice" sound is, i.e. the amplifier as effects box.

S.

 

[Yank]

Those were the only ones that mattered. ]

S.

That's not what you said before:

As to loudspeakers, many today make no attempt at a flat frequency response or accurate pair-matching, and so wouldn't meet DIN 45-500. They seem to sell on style and trying to get people to like their sound rather than going for accuracy. In the 1970s, manufacturers strove for accuracy pretty much above
all else, so meeting DIN 45-500 was relevant.

I don't think this claim can be supported. With the widespread availability of inexpensive measurement tools and crossover modelling software, I doubt today's speakers are, on average, less flat than in the '70s.

 

[sergeauckland]

That's not what you said before:


That was a joke!


I don't think this claim can be supported. With the widespread availability of inexpensive measurement tools and crossover modelling software, I doubt today's speakers are, on average, less flat than in the '70s.

You would think so wouldn't you! Yet, some of the horrors I've seen reviewed in the past few years makes me wonder whether these people have ever heard of a measuring microphone or a pseudo-anechoic impulse response.

S.

 

[sq225917]

or if indeed your methodology and equipment is up to the task...

 

[Yank]

You would think so wouldn't you! Yet, some of the horrors I've seen reviewed in the past few years makes me wonder whether these people have ever heard of a measuring microphone or a pseudo-anechoic impulse response.

S.

No doubt, but there was no shortage of those in the past either. And from big names with large engineering departments too.

 

[James]

No. You don't create the music, record it, produce it, press it or make it available for download.

Can you not see the difference?

You are missing my point completely.

The photography analogy I was using compares a lens with an amplifier.

Both amplifier and lens *should* be transparent/accurate/uncoloured. Just as you can measure THD/IMD/FR and every other amplifier spec known to serge, you can measure some aspects of lens performance such as resolution, linear distortion etc. But these aspects alone don't always tell how the amplifier or lens performs subjectively.

Of course you can choose your preferred amp or lens on specs alone, or you can try them out and see/hear if you like the way they perform.

Is that clear enough for you?

 

[James]

No doubt, but there was no shortage of those in the past either. And from big names with large engineering departments too.

... not to mention some people absolutely love the way they perform, despite their woeful technical measurements.

 

[hecd2]

You are missing my point completely.

The photography analogy I was using compares a lens with an amplifier.

Both amplifier and lens *should* be transparent/accurate/uncoloured. Just as you can measure THD/IMD/FR and every other amplifier spec known to serge, you can measure some aspects of lens performance such as resolution, linear distortion etc. But these aspects alone don't always tell how the amplifier or lens performs subjectively.

Of course you can choose your preferred amp or lens on specs alone, or you can try them out and see/hear if you like the way they perform.

Is that clear enough for you?

It is clear but misleading. A photographic lens is a linear optical system (ie optical media are linear and the superposition principle holds) and its output can be completely determined given its input and its transfer function. The transfer function of a lens at any field and focal position is determined by just three things:

• 
  The geometric aberrations of the lens which are wholly determined by the lens design - these aberrations consist of the first order aberrations such as spherical aberration, coma, astigmatism, longitudinal and transverse chromatic aberration plus higher order aberrations. The geometric performance of the lens design can be determined by ray tracing.

• 
  The diffraction of the lens, normally determined by diffraction at the lens aperture - in the case of camera lens this will generally be at the polygonal lens diaphragm although there might be some vignetting at large apertures and edge of field conditions.

• 
  The former two effects are inherent to any lens design - on top of that the performance of an individual lens will be affected by errors in fabrication - departures of lens components from design (departures from sphericity, curvature, thickness, surface alignment etc) and errors in assembly (component spacing, off axis assembly, component tilt etc).

The effect of these three classes of imperfection are determined at any aperture, field and focal position by the optical transfer function, which can be measured. The optical transfer function completely specifies the performance of the lens - the image is determined by the convolution of the transfer function and the object. There are no hidden effects. If two lenses have identical transfer functions then their images (and "artistic" descriptions of the appearance of the image such as bokeh) will be identical and indistinguishable. If a subject can reliably distinguish between the images produced by two lenses then you can be sure that they will have different transfer functions. (And of course lens designers know what effect different choices have on bokeh, and are perfectly able to design lenses with good bokeh if that is important compared to many of the other conflicting requirements for the design. Bokeh is not magical but is determined largely by the out-of-focus two dimensional point spread function)

So this is where the analogy does hold - Serge holds that if two amplifiers measure to have identical transfer functions (ie impulse responses) or frequency responses into real loads within the limits of perception then they will be indistinguishable to observers in a controlled test - similarly two lenses that measure identically will be indistinguishable to observers in a controlled test. You can measure all that is important in both cases as there are no unmeasurable physical effects present that can affect perception.

One thing is sure - it's a damn sight easier to design a practically perfect audio amplifier than it is to design a practically perfect 100-400mm F4.5 zoom. The former is routine, the latter impossible.

 

[James]

That's a fascinating and concise description of optical design theory. Thanks.

I agree that if transfer functions are identical, then there should be no difference in either amplifiers or lenses. I can certainly hear no difference between my "lesser" Densen B-340 4-channel amplifier when compared with the "better" B-350 mono-blocks; nor can I discern any image difference between my Pentax-A 50/1.4 and FA50/1.4, which have identical optics.

But when comparing two differently designed, but similarly specified, items - their measured performance may show a difference (if measuring the right things) or it may not. Serge's contention that most SS amplifiers are practically transparent and therefore should sound the same may not hold if he's not measuring, for example, their transfer function. In any case, one of his criterion for transparency is a flat response from 20Hz - 20kHz, but does this mean an ultra-wide bandwidth design that is flat from DC to 2MHz, but identical in every other measureable aspect to another that has a brickwall filter at 22kHz, should sound the same? I'm not sure.

When I changed from Naim NAP135s to Densen B-350s, I was struck by a very distinct change in sonic character that I could pick them apart blind. It's probable the Naim has a different transfer function to the Densen, but both very competent amplifiers can't be considered transparent if they sounded different.

I don't design amplifiers or lenses, but the final frontier for hifi transparency has always been, and will always be, the loudspeaker. These, I do design - in vain.

 

[avole]

You are missing my point completely.

The photography analogy I was using compares a lens with an amplifier.

Both amplifier and lens *should* be transparent/accurate/uncoloured. Just as you can measure THD/IMD/FR and every other amplifier spec known to serge, you can measure some aspects of lens performance such as resolution, linear distortion etc. But these aspects alone don't always tell how the amplifier or lens performs subjectively.

Of course you can choose your preferred amp or lens on specs alone, or you can try them out and see/hear if you like the way they perform.

Is that clear enough for you?

Not what you said at all, old son. I've highlighted the bits so it'll be easier for you to understand:

"Transparency is both absolute and relative. In the analogous world of photography, the amplifier is a camera lens. It should transfer what it sees from one side and re-create an identical image on the other side on film or sensor."

The world of photography is not analogous at all, for the reasons I've stated. Moving the goal posts, as you have in your last answer, suggests you think I'm actually right.

 

[James]

Not what you said at all, old son. I've highlighted the bits so it'll be easier for you to understand:

"Transparency is both absolute and relative. In the analogous world of photography, the amplifier is a camera lens. It should transfer what it sees from one side and re-create an identical image on the other side on film or sensor."

The world of photography is not analogous at all, for the reasons I've stated. Moving the goal posts, as you have in your last answer, suggests you think I'm actually right.

You can think what you like, but you're wrong.

 

[sergeauckland]

The analogy with photography fails because the photograph is analogous with the recording, not the reproduction,

A photocopier is analogous with HiFi reproduction.

If one takes a recording and analyses the audio information therein contained, one can compare it with what comes out of the amplifier, and assess accuracy of reproduction.

In the same way, one can photocopy a photograph and assess the accuracy of reproduction using suitable optical instrumentation.

SImilarly, one can take the original photograph and compare it visually with the photocopy. Can they be told apart? If original photographs and photocopies are shuffled, can they be identified visually as to which is which? If not, with any statistical accuracy, then the photocopier is, for those observers, transparent.

Similarly, if a SWBT fails to identify the path with the PAuT, then the PAuT must be transparent for those listeners.

S.