Determinants of amplifier performance

[Sawyer]

From many years of listening to music at home using many amplifiers, my thought are as under:
Quote
What then should decide the purchase of an amplifier?
My views:
1. Enough power to do the job, defined by the speaker needs, plus something more as a power reserve. For a very large majority of applications at home, 60-80 wpc is enough. Even 100wpc, solid state amplifier watts aren't very expensive now.
2. Features such as input/output sockets, USB sockets, etc., based on individual needs.
3. Availability of parts and after sales service
4. Brand reputation for reliability
5. Tone controls - another thing isn't hard to implement without, as some think it does, "corrupting" the signal chain. The ability to tweak bass and treble levels is invaluable in customising the sound of a speaker to the room acoustics/environment.
6. Integrated amps are as good as any. With solid state designs and properly implemented power supplies, the preamp separation to protect the delicate low voltage signal at the preamp stage from distortion causing effects isn't as important any more, where the power outputs are in the ranges mentioned above.
7. Footprint criteria
Sonic signature doesn't figure any where in this list.
Unquote

Cut/pasted from post#14 of another thread I started here:
http://www.hifivision.com/what-should-i-buy/51819-essentials-getting-yourself-good-music-home.html
I am not an engineer, but I am one curious layman, and how things work is an area of intense interest. This aspect is reflected in how I have reached my conclusions, and the fact that I have not referred to quantifiable data anywhere.
The purpose of this thread is to get inputs from more experience people here, or more qualified in the discipline of electrical/electronic/audio engineering, for my benefit. And for the benefit of others here trying to decide which amplifier to buy.
PS: To keep things simple, let us talk only about 2 channel amps here. The principles apply to AVRs too, but those devices have a lot of sound shaping electronics in them, which make things more complex.
Also, I realise that the thread title should be something on the lines of assessment of performance - does anyone know if this can be corrected?

 

[Sawyer]

What spurred the creation of this thread

In my thread referred above, in post #60 I have dismissed the worth of current day hifi magazines.
Even so, once in a year or so, I read the latest issue of What HiFi, to keep up to what is happening in the market.
What I read in the latest issue again amazed and irritated me, prompting me to start this thread for reasons referred to above, and also to test and validate my thoughts via contributions from other rational and better qualified members here.
The latest issue has a review of the Marantz PM 6005 amplifier. The review states that their assessment of this, after "testing" is that this amplifier is more musically accomplished than any of its predecessors. Somewhere else it says - it sounds better than its award winning predecessor, the PM 6004.
The other reason that prompts this thread is seeing that here too there is so much talk about amplifier matching and selection based on fuzzy things like how it sounds - warm, bright, lean, dry.
To both the above, my reaction is - if this is true for an all electric/electronic device like an amplifier, explain to me the meaning of these words in precise scientific terms.
The magazine article doesn't. Nor does it list anywhere the technical specifications of the amplifier, such that if one does a comparison with those of say the PM 6004, one can see where this better sound is visible in terms of technical specifications.
I don't have these to do this myself, but I am almost certain that if I was to do a comparison of the specs of the two, there would be no difference.
If that is the case, the only thing the magazine is doing is to create dissatisfaction in the minds of PM6004 owners.
There is one important reason to upgrade, I agree. The 6005 has an extra feature, it has a digital inputs, because unlike the 6004, it has a DAC built in. If that feature was pointed out to be the reason for an upgrade, no problem with that. But to say that the 6005 is better sounding or more musically accomplished? In my view till this point in time - a shameless lie. Until the time it is shown to not be so by a difference in objectively measurable specification parameters for the outputs of the two, I have no hesitation in calling it that.

 

[Sawyer]

Amplifier tech specs

I have the manual of my trusted Quad 909 power amp open and I see these parameters:
1. Power output
2. Total Harmonic Distortion
3. Output impedance
4. DC offset voltage
5. Frequency response
6. Power response
7. Input sensitivity
8. Input impedance
9. Crosstalk
10. Signal to noise ratio
11. Stability

I don't think there is anything missing that will have a sonic signature impact - am I right?
My assertion, open to rational attack:
If any other amplifier has the same measurements that the Quad 909 has for the above items, replacing it for the 909 will result in no change in the sound heard. Provided that no other variables are changed.
My guess:
Of the above, the items that will influence sonic signature to any degree, small or large, are 2, 5, 6, 9, 10 and 11.
The others will affect source and speaker selection for electrical - not sonic - matching and sound levels produced.

 

[Sawyer]

Re: Amplifier tech specs

My assertion, open to rational attack:
If any other amplifier has the same measurements that the Quad 909 has for the above items, replacing it for the 909 will result in no change in the sound heard. Provided that no other variables are changed.

If the above is a valid assertion, what rational reason do I have to spend hard earned money to change a working amplifier for another, except for reasons 1 to 7 mentioned in my opening post?
And outside of these measurements, where do the supposed changes in sound from cold to warm, lush to analytical, bright to dark, lean to full bodied come from?
By the way, the days when globally reputed amp makers knowingly cheated in their spec sheets are gone. If anything, they are conservative in this area, there is too much to lose. And they perhaps know that they really don't need to do this, the media and peer pressure influences are enough to sell the so called latest greatest thing since sliced bread without having to resort to made up spec sheets, which no one is going to read anyway.

 

[Sawyer]

Re: Amplifier tech specs

If the above is a valid assertion, what rational reason do I have to spend hard earned money to change a working amplifier for another, except for reasons 1 to 7 mentioned in my opening post?

Taking this a step further, the eight reason would the tech specs. I don't include this in my list only because I believe that the nature of 2 channel audio amplification technology is such that every global maker of repute makes amps that comfortably exceed the measurements needed to be achieved to have an audible impact on unbiased human hearing. If I were to change my amp today, I would not look at the spec sheet, nor would I be obsessed with its sound signature. A quick test to see if it works would be all that I would need.
However, not everyone will see things this way.
In which case, what are parts of the spec have an influence on the sound quality? And what are the parameters for these that have to be met/exceeded for the amp to not have an audible sonic signature by way of distortion?
Anyone? I haven't a clue as to the second question's answer.

 

[Sawyer]

A Summary

There are two parts to this.
First, anything that yields a sonic signature attributable to a 2 channel amp that does not have tone controls/or with tone controls set at the centre has to be visible in objectively measurable spec parameters of the amp. My assertion of course, prove me wrong, I am happy to change it.
Two, once the human audibility threshold has been passed, better performance on spec parameters doesn't matter. Again, my assertion, but common sense really. Except it would be useful to know what spec criteria and what parameters are the cut off points.
Speakers are a different species of beast altogether...more on why so later.

 

[Sawyer]

Prat

Pace, Rhythm, Attack and Timing.
I have a few Brit friends and I enjoy their company. Britain still holds its own in many high tech industries too as well as some legacy ones like high end audio. But Britain also is a pioneer of trash tabloid journalism and this has entered their hifi magazine industry as well.
A great example of this is PRAT. Originally coined for speakers, this term now gets applied to all system components. Recent discussions here on digital audio may tend to lead to thinking in some places that PRAT may be a valid term for DACs as well!
But using PRAT to describe the output of an audio amplifier that takes analog inputs and with gain, hands these over as analog outputs? Utter rubbish. Probably started tongue in cheek by someone somewhere - the Brits are good at that - and then idiotically imitated in all seriousness later.
I haven't seen it used here...perhaps the acronym isn't popular anymore.

 

[Nero9]

Wow very nice lot of information i ll grab a tea and start reading lol.

 

[Sawyer]

Wow very nice lot of information i ll grab a tea and start reading lol.

Enjoy!
Another ludicrous statement in the magazine:
"The Marantz separates are tuned to play together so perfectly, we can't tell where one ends and the other starts."
The only advantage of such pairings is a common look, and use of a common remote. And a common service centre - hopefully!
And the same magazine will not hesitate at another time to recommend different make components to be part of what is supposedly a system with a synergistic sonic signature/philosophy. More drivel. I think they just do eeny meenie myna mo, and write such articles.
The whole idea of separates was to mix and match, based on feature needs and budgets. All that has to be taken care of is electrical matching in the areas of signal sensitivities, impedances and power needs. And makers build in enough play in these aspects to allow for a wide range of choices for good enough electrical matching.
And yet, following the lead of some magazine that says that Brand X CDP goes best with Brand Y amp, which goes best with Brand A speaker, this advice is slavishly followed, and disseminated. A lot of that goes on here too on many threads.
Commercially too, to do otherwise would be hara kiri. In what is already a shrinking market, the rational thing to do would to make sure that a component works with as broad a range of others as possible. And the more the makes of components out there, the more the reason to not tie yourself down. Business management 101.

 

[Sawyer]

Small digression on speakers

A digression, only to highlight the amp subject via the contrast with speakers.
Speakers are electro-mechanical devices, that also directly interface with air to deliver the output. There are so many known and unknown variables at play that tech specs can tell only a small part of the story. Things like how the cone material behaves, or how the cone movement interacts with the air inside the speaker box, or how that air interacts with the speaker box cannot be captured in a spec. Speaker voicing remains an art+science+listening acuity on the part of the designer subject, and with many many degrees of freedom to a designer that affect the nature of the sound produced. All the terms referred earlier, from PRAT to things like lush, dry, analytical, natural, lean, and the like apply even while remaining undefined, because many of these are also subjective. And all of this, before room acoustics comes in to take over the show, in an unbreakable bond with the speakers. I would never buy an expensive speaker without first listening to it, in a familiar room, with familiar music of preferred genres. For an extended length of time. Specs can only be a starting point to narrow down the selection.
Back to simple(!!!) electrons and 2 channel amplifiers...

 

[asliarun]

There is also the amp class - class a, class ab, class d, chip amps etc.
Class d amps are especially interesting because they got a bad rep, but of late, the newer tech amps like hypex ncore and ice power have got rave reviews. I have not had the good fortune of hearing them meaningfully though. What I also find interesting is that class d is so energy efficient and also is apparently completely immune to speaker impedance swings.

Plus, most studio amps like crown are class d, so not sure why they have a bad rep. Kanwar in this forum can probably talk about this for hours ;-)


Good power delivery at all impedance levels is a huge deal in my book. I used to have totem model one which were power hungry mini monitors with impedance dips up to 3.6 ohms. People spoke about amp pairing but in this case, I think it was all about being able to supply good quantities of current at all impedance swings.

 

[jai1611]

Re: Amplifier tech specs

I have the manual of my trusted Quad 909 power amp open and I see these parameters:
1. Power output
2. Total Harmonic Distortion
3. Output impedance
4. DC offset voltage
5. Frequency response
6. Power response
7. Input sensitivity
8. Input impedance
9. Crosstalk
10. Signal to noise ratio
11. Stability

I don't think there is anything missing that will have a sonic signature impact - am I right?
My assertion, open to rational attack:
If any other amplifier has the same measurements that the Quad 909 has for the above items, replacing it for the 909 will result in no change in the sound heard. Provided that no other variables are changed.
My guess:
Of the above, the items that will influence sonic signature to any degree, small or large, are 2, 5, 6, 9, 10 and 11.
The others will affect source and speaker selection for electrical - not sonic - matching and sound levels produced.

Unfortunately the list is longer than that. I'm sure Kanwar can add more. Here's few I thought off:
1. Slew rate
2. Rise time
3. The breakup of THD into first, second order etc

These all have a material influence on sound.

 

[deba]

Question : What then should decide the purchase of an amplifier?

Answer: Your EARS only.

 

[Sawyer]

Good power delivery at all impedance levels is a huge deal in my book. I used to have totem model one which were power hungry mini monitors with impedance dips up to 3.6 ohms. People spoke about amp pairing but in this case, I think it was all about being able to supply good quantities of current at all impedance swings.

Ok, here we go.
Yes, in this case amp pairing just means enough power to cope with the impedance dip for the required sound levels. 3.6 isn't bad, there are some that drop to 2 for low frequency reproduction, demanding and sucking out huge power/current from the amp.
Layman question: My amp is 140 wpc, and also has peak current supply of 11 amps a channel. Don't the two go together - will there be another 140wpc amp, that does not supply the same current level if needed to? Assuming both amps are honestly specced.
PS: On the amp to speaker matching subject, afaik, power is the only factor to be considered. Once one has decided on what speakers to buy based on the various attributes needed, including the capability to handle the power needed to produce the desired sound levels, amp selection follows by seeing if it has the capability to supply, without clipping, the power that the speaker will draw for all frequencies, to produce those levels. More power delivery capability than that in the amp isn't an issue as long as one can use the volume control responsibly. Less power delivery capability is, because that will introduce clipping caused distortion and beyond a point, speaker damage.

 

[Sawyer]

Re: Amplifier tech specs

Unfortunately the list is longer than that. I'm sure Kanwar can add more. Here's few I thought off:
1. Slew rate
2. Rise time
3. The breakup of THD into first, second order etc

These all have a material influence on sound.

Ok, lets see what more turn up. I assume that all three can be measured via instruments and the thresholds where these start having a material influence is known via experience.

 

[asliarun]

Oh, and damping factor.

 

[jai1611]

Oh, and damping factor.

Damping factor is just another way of measuring output impedance

 

[jai1611]

Re: Amplifier tech specs

Ok, lets see what more turn up. I assume that all three can be measured via instruments and the thresholds where these start having a material influence is known via experience.

They have a known impact. However, different people like different things. For example, tube amps have higher THD but have different order distortion, which sounds different.

Output impedance is also speaker dependent. Lowthers love damping factor of around 8-10, Dynaudios like damping factor of 50 or higher.

This is all without getting into the issue of published specs on amps being often unreliable. The simplest way I've found is to try the amp with your speakers. Some speakers are not really fussy about amps (eg Harbeth) while others are very picky (eg Dynaudio).

 

[nishikant]

I quote from HiFIWorld UK
"I first measured and heard a 'perfect' amplifier somewhere around 1974 and I've been measuring and listening to them ever since. A 'perfect' solid state amplifier sounds little different from an imperfect one; sometimes it can sound worse. For example, I pitied Hitachi with their HMA7600 MOSFET amplifier; it had the most astonishing measured performance, yet nobody liked it much. The Japanese design engineers must have cried! " Here is the link
WAD 300B AMPLIFIER

 

[Sawyer]

Re: Amplifier tech specs

For example, tube amps have higher THD but have different order distortion, which sounds different.

Some speakers are not really fussy about amps (eg Harbeth) while others are very picky (eg Dynaudio).

If some one likes to hear distorted tube sound as a preference, that is one thing. The word distortion doesn't by definition mean unpleasant. Some people also like the distortion caused by low powered amplifiers - clipping mildly this may be a very rational subjective preference. Question is - are they being as true to the recorded signal as another amp, say a modern solid state amp from almost any reputed global brand?
By the way, I once had a very well engineered Unison tube amp with 6 EL34s. I think it was about 35wpc. I found no difference in my sound, replacing it with a 50wpc solid state. Lost the lovely glow of 6 tubes at night though.
As to the Harbeth/Dynaudio difference - my view is that it is just a matter of one needing a lot more power than another to perform as designed. This is not any reflection of how they sound, or which one sounds better. The only difference is that the power hungry speaker will need a more powerful, and therefore more expensive amp to work well. I can't imagine that over two adequately powerful amps, the Dynaudio will be picky in terms of preferring one over the other. If I am wrong, I would love to hear a scientific reason for this.

I quote from HiFIWorld UK
Here is the link
WAD 300B AMPLIFIER

I read it. I am afraid I can't lend too much credibility to a person that claims to be able to hear cable differences.