Do all amps sound (nearly) the same?

[superczar]

First a small anecdote
Some months ago, I decided to wire up a unused amplifier to my AVR to drive the front L/R speakers to offset some of the power delivery when I added ATMOS to my setup. While I do believe that offloading the two primaries to a dedicated amp would open up headroom for the overused multi channel amps on the AVR, the end result got me thinking along a different line.
After I had added the stereo amp to the setup , I ran Audyssey (multeq XT32) on the AVR and was a bit surprised when I noticed that Audyssey boosted the gain on the FL/FR channels by 4db
I was under the assumption that gain settings on all amps (at least mid to high grade) would be set to a reference level so the result took me by a bit of surprise.

Now it’s probably well know that psychoacoustically, a few dB extra loudness makes us perceive the sound as fuller/richer - so It got me wondering if the so called amp sonic differences stem on account of the default gain variance.

Thus I carried out a small experiment on my own.
Common thread - All sessions were set to output a 78db loudness level on a pink noise file by using a calibrated microphone wired to a macbook to set the gain on all tested amps so that they output the same amplitude

Amplifiers tested -
1) Marantz PM17 SA integrated amp
2) Marantz SR6010 AVR in direct mode
3) SMSL SA-160 class D amp
4) Harman Kardon AVR 5000 in direct mode
5) Paradigm PW Amp in direct mode (I.e. Anthem room correction disabled)
6) Sure Electronics TAS5630 class D amp board
7) ultra cheap TPA3116D2 board

I am not sure if its my ears or if it’s something else but honest to goodness, other than (7) above, subtle differences aside, I would be hard pressed to select one over the other
Even (7) wasn’t bad at all but seemed to be running out of steam at what is a fairly high dB level.. I guess it would end up being at par (at least for me ) at more normal listening levels

Source for all the tests was a M-audio DAC hooked to a MacBook Pro playing a 24/192 version of The National’s About Today

Hypothesis : Is it possible that a lot of (not all) perceived differences in subjective assessments of Amps are on account of a gain difference rather than a true sonic difference?

PS: (1) and (2) are the AVR and amp referred in the back-story
PS2: Two sets of speakers were used for the tests , KEF Q300 and Phase Tech Velocity V12 - Arguably neither is a top of the line set but both are critically acclaimed mid-end speakers

 

[Fantastic]

Well designed amps played below clipping and with reasonably low distortion ( almost a given nowadays) 'tend' to sound 'similar'. However there could be subtle differences usually only noticeable with critical listening and on certain tracks/sounds.
You can check this out by trying different brands and types of input capacitors on a dc coupled amp. They can sound different. But taken as a black box and listened to a different times, most amps could sound fairly similar.
Differences will appear when the load becomes difficult ( like going below 4 ohms etc. ) and ability to handle large dynamic signals. Loudspeakers usually don't have a uniform impedance with respect to frequency. They vary quite a lot. So a nominally 4 ohm rated speaker could drop to say 2.5 ohms at some part of the bandwidth. This would be difficult to drive with some amps. That's when all amps are no longer equal.
For people who play music at low level , even a well designed 20 watt per channel chip amp would sound great ......and maybe save a boat load of money ! Do a blind test, you will be surprised!

 

[Nitin K]

First a small anecdote
Some months ago, I decided to wire up a unused amplifier to my AVR to drive the front L/R speakers to offset some of the power delivery when I added ATMOS to my setup. While I do believe that offloading the two primaries to a dedicated amp would open up headroom for the overused multi channel amps on the AVR, the end result got me thinking along a different line.
After I had added the stereo amp to the setup , I ran Audyssey (multeq XT32) on the AVR and was a bit surprised when I noticed that Audyssey boosted the gain on the FL/FR channels by 4db
I was under the assumption that gain settings on all amps (at least mid to high grade) would be set to a reference level so the result took me by a bit of surprise.

Now it’s probably well know that psychoacoustically, a few dB extra loudness makes us perceive the sound as fuller/richer - so It got me wondering if the so called amp sonic differences stem on account of the default gain variance.

Thus I carried out a small experiment on my own.
Common thread - All sessions were set to output a 78db loudness level on a pink noise file by using a calibrated microphone wired to a macbook to set the gain on all tested amps so that they output the same amplitude

Amplifiers tested -
1) Marantz PM17 SA integrated amp
2) Marantz SR6010 AVR in direct mode
3) SMSL SA-160 class D amp
4) Harman Kardon AVR 5000 in direct mode
5) Paradigm PW Amp in direct mode (I.e. Anthem room correction disabled)
6) Sure Electronics TAS5630 class D amp board
7) ultra cheap TPA3116D2 board

I am not sure if its my ears or if it’s something else but honest to goodness, other than (7) above, subtle differences aside, I would be hard pressed to select one over the other
Even (7) wasn’t bad at all but seemed to be running out of steam at what is a fairly high dB level.. I guess it would end up being at par (at least for me ) at more normal listening levels

Source for all the tests was a M-audio DAC hooked to a MacBook Pro playing a 24/192 version of The National’s About Today

Hypothesis : Is it possible that a lot of (not all) perceived differences in subjective assessments of Amps are on account of a gain difference rather than a true sonic difference?

PS: (1) and (2) are the AVR and amp referred in the back-story
PS2: Two sets of speakers were used for the tests , KEF Q300 and Phase Tech Velocity V12 - Arguably neither is a top of the line set but both are critically acclaimed mid-end speakers

Maybe in Source Direct mode most of them may sound similar but with room correction in normal mode may sound different. Could be there are subtle differences in their sonic signature due to the electronic components used but in the end all of them are supposed to play back music faithfully the way they were recorded.

Here's some food for thought. .

Do Hi-Fi Amplifiers Sound Alike?

Note: This is a long read. Scroll down for a TL;DR at the end. Links to sources are also at the end. If you have time, please read it all. What Does An Amplifier Do To The Sound? What difference do…

jakekuyser.wordpress.com

 

[elangoas]

After I had added the stereo amp to the setup , I ran Audyssey (multeq XT32) on the AVR and was a bit surprised when I noticed that Audyssey boosted the gain on the FL/FR channels by 4db

No idea on the gain structure and other technical details..

I recently added power amps to my AVR, though yet to calibrate with XT32.. Logically IMO, if the speaker position & listening position doesn't change, the speaker level should remain the same..

The benefit of adding power amp to AVR is for robust power to speakers, headroom and dynamics.. I wasn't sure if i had to re-calibrate after adding power-amp.. Anyways, have planned to re-calibrate and check for speaker level before / after power amp..

 

[superczar]

Well designed amps played below clipping and with reasonably low distortion ( almost a given nowadays) 'tend' to sound 'similar'. However there could be subtle differences usually only noticeable with critical listening and on certain tracks/sounds.
You can check this out by trying different brands and types of input capacitors on a dc coupled amp. They can sound different. But taken as a black box and listened to a different times, most amps could sound fairly similar.
Differences will appear when the load becomes difficult ( like going below 4 ohms etc. ) and ability to handle large dynamic signals. Loudspeakers usually don't have a uniform impedance with respect to frequency. They vary quite a lot. So a nominally 4 ohm rated speaker could drop to say 2.5 ohms at some part of the bandwidth. This would be difficult to drive with some amps. That's when all amps are no longer equal.
For people who play music at low level , even a well designed 20 watt per channel chip amp would sound great ......and maybe save a boat load of money ! Do a blind test, you will be surprised!

While it would be difficult to do a blind test given the lack of switching equipment and availability of other interested parties , I agree that some subtle differences may exist even at low volumes.
However for most users who play at regular volumes, I must say that a subtle difference between an amp that cost 20 X of the other (#1 and #6 above) is a bit of an ani-climax, isn't it ?

No idea on the gain structure and other technical details..

I recently added power amps to my AVR, though yet to calibrate with XT32.. Logically IMO, if the speaker position & listening position doesn't change, the speaker level should remain the same..

The benefit of adding power amp to AVR is for robust power to speakers, headroom and dynamics.. I wasn't sure if i had to re-calibrate after adding power-amp.. Anyways, have planned to re-calibrate and check for speaker level before / after power amp..

You do need to re-calibrate after making any major change
What you have said above is right and that's what one would expect
My point isn's that a power AMP won't help with an AVR - My point is more on whether there is a true (and reasonably large) difference between a $1000 amp, a $50 amp and maybe a $5000 amp for a pure stereo setup

Maybe in Source Direct mode most of them may sound similar but with room correction in normal mode may sound different. Could be there are subtle differences in their sonic signature due to the electronic components used but in the end all of them are supposed to play back music faithfully the way they were recorded.

Here's some food for thought. .

Do Hi-Fi Amplifiers Sound Alike?

Note: This is a long read. Scroll down for a TL;DR at the end. Links to sources are also at the end. If you have time, please read it all. What Does An Amplifier Do To The Sound? What difference do…

jakekuyser.wordpress.com

Interesting article - It appears this fellow also reached the same conclusion as I
And yes, I am a firm believer in good room correction and have always maintained that for an enthusiast on a non-unlimited budget, the best bang for buck comes from investing in enabling a good input at a pre-amp level (good source, good DAC, good room correction) compared to spending on a expensive amp (and even more so on expensive cables )

 

[Nitin K]

Interesting article - It appears this fellow also reached the same conclusion as I
And yes, I am a firm believer in good room correction and have always maintained that for an enthusiast on a non-unlimited budget, the best bang for buck comes from investing in enabling a good input at a pre-amp level (good source, good DAC, good room correction) compared to spending on a expensive amp (and even more so on expensive cables )
[/QUOTE]


Absolutely true, fully agree with you.

 

[elangoas]

You do need to re-calibrate after making any major change

Ok.. Will re-calibrate..

 

[newlash09]

Hi superczar

Every amplification section of a power amp works at a fixed level of amplification. One of my tubed power amps has 29db of gain, where as the other SS power amp has 26 db of gain. So if I were to use them as power amps running out of the preamp output of the AVR. Then the AVR when doing the calibration at its reference 75db level will reduce the gain on the tube power amps by 3db to achieve the same volume level at your listening position, as compared to using the SS amp before .

I believe this is what is happening in your system too. The inbuilt power amp sections of the AVR have a gain of 4db more than the amp of the power amp. So to compensate for that, it is increasing the preamp output by 4 db .

 

[superczar]

Hi superczar

Every amplification section of a power amp works at a fixed level of amplification. One of my tubed power amps has 29db of gain, where as the other SS power amp has 26 db of gain. So if I were to use them as power amps running out of the preamp output of the AVR. Then the AVR when doing the calibration at its reference 75db level will reduce the gain on the tube power amps by 3db to achieve the same volume level at your listening position, as compared to using the SS amp before .

I believe this is what is happening in your system too. The inbuilt power amp sections of the AVR have a gain of 4db more than the amp of the power amp. So to compensate for that, it is increasing the preamp output by 4 db .

Hey @newlash09
I got that bit which is in fact what got me wondering if its the different gain levels between amps that make listeners wonder if Amp A is more forward sounding/bright etc etc than Amp B
Assume you were A/B testing 2 analog integrated amps that have a 4db difference at a given volume position. After setting both volume knobs to 12’o clock, you playback a soundtrack
Even if everything else were the same, it’s very likely that a listener would prefer the +4db amp because the gain difference is too small for the ear to realise in terms of loudness but large enough to make a perceptible difference in the details one would hear

 

[sound_cycle]

Even if everything else were the same, it’s very likely that a listener would prefer the +4db amp because the gain difference is too small for the ear to realise in terms of loudness but large enough to make a perceptible difference in the details one would hear

I'm not in the debate on amps. Or into scientific testing. I'm comfortable with both the objective and subjective camps.

While hardly masquerading as na level matched perfectly, objective "experiment" an account to entertain, below.

Now I run room correction to somewhat get the room out if the way. (I'm very clear on what DRC can / not do)

And we heard a chifi pre- power combo and a cheapshit int DAC amp (with an external DAC) the 25 k extra for the much bigger dabba was wasted money.

May be we liked the valve distortion and harmonics maybe the amp but it was a very clear preference that led me to a pre- power combo that completely comprehensively blew the you wonder away. There was smoothness and detail which neither chifi nor cheapshit showed us.

Later we tried these in a class a/b vs class d amps, in an even more unscientific manner. Using different rigs and rooms and whaddya know both listeners are happy with the sound from either (at least true for me)

Conclusion:

Neither speakers or amps can show you what is not there.

And that is subject to the room.

I find great hilarity in reading accounts of non treated non corrected rooms having great mids and highs or that trouser flapping bloated one bass is great for two ch. The room boom will NOT let you hear it. No <bleep> way.

But the greatest leveller the brain comes in filling in detail and missing bits.

So, what's the point ? whatever you want to make of it.

For me, now in order of importance it is room, source, DAC, pre- and amp/speakers.

There's my bumbling and highly opinionated subjective money squandering experience in a few pithy words. And while I am happy and content in my state of imperfection, I am reasonably certain that for my money and constraints that's more or less what I can get.

Ofc that is provisional and FTR I am not chasing dabbas.

Ciao
GR

 

[newlash09]

Hey @newlash09
I got that bit which is in fact what got me wondering if its the different gain levels between amps that make listeners wonder if Amp A is more forward sounding/bright etc etc than Amp B
Assume you were A/B testing 2 analog integrated amps that have a 4db difference at a given volume position. After setting both volume knobs to 12’o clock, you playback a soundtrack
Even if everything else were the same, it’s very likely that a listener would prefer the +4db amp because the gain difference is too small for the ear to realise in terms of loudness but large enough to make a perceptible difference in the details one would hear

Hi superczar..

Thats very true..and a lot of companies do indulge in that fickenry too...same with dac's. Every dac has its own output level. So a dac with a higher output level will immediately sound more detailed. So it is important to level match every component in terms of absolute listener volume if you really want to compare between them.

Another trick is to enhance the treble...So it will look loads more detailed, but will be a tiring listen in the long term.

 

[arj]

There is a theory which says that the Amp should only be a wire with a gain but unfortunately most amps are coloured in some way and that usually means that more the detail the differently it is perceived.
I remember reading about an experiment where nelson pass made the same amp sound like a tube amp and an SS with some minor tweaks. While there are many amps which do a great job irrespective of their high or low price, it is very curious that all the amps from an AVR to a chip amp sound the same .

It could be an interesting experiment to plug in a dedicated CDP and repeat this experiment with maybe just Amps 1 and 2 and see if this is still the case ?

 

[mpw]

Neither speakers or amps can show you what is not there.

And that is subject to the room.

I find great hilarity in reading accounts of non treated non corrected rooms having great mids and highs or that trouser flapping bloated one bass is great for two ch. The room boom will NOT let you hear it. No <bleep> way.

But the greatest leveller the brain comes in filling in detail and missing bits.

So, what's the point ? whatever you want to make of it.

For me, now in order of importance it is room, source, DAC, pre- and amp/speakers.

There's my bumbling and highly opinionated subjective money squandering experience in a few pithy words. And while I am happy and content in my state of imperfection, I am reasonably certain that for my money and constraints that's more or less what I can get.

Ofc that is provisional and FTR I am not chasing dabbas.

Ciao
GR

good one...

 

[liverpool_for_life]

it’s very likely that a listener would prefer the +4db amp because the gain difference is too small for the ear to realise in terms of loudness [...]

Hmm, I'd differ with you on this. A difference of +4db in levels will be quite likely fairly obvious in terms of loudness.

I'd agree with you on the general premise re. amps, as long as they're competently designed and built to do the thing they're supposed to do (they don't have to be expensive, FWIW). My reading suggests that headroom does make a difference. Of course, in those cases, we are talking about high listening levels that most untreated rooms can't support.

Thanks for doing this, btw!

 

[superczar]

There is a theory which says that the Amp should only be a wire with a gain but unfortunately most amps are coloured in some way and that usually means that more the detail the differently it is perceived. I remember reading about an experiment where nelson pass made the same amp sound like a tube amp and an SS with some minor tweaks. While there are many amps which do a great job irrespective of their high or low price, it is very curious that all the amps rom AVR to a chip amp sound the same .

It could be an interesting experiment to plug in a dedicated CDP and repeat this experiment with maybe just Amps 1 and 2 and see if this is still the case ?

The last CDP I had was circa 2008 and while I did rip my CD collection back then, that physical collection has been long gathering dust in an attic at my paternal hometown (and probably already damaged beyond repair)
I kind of doubt 1 and 2 would have a perceptible difference even in a critical session even though I maintain that there is value in having a dedicated integrated amp
What is surprising to me is the relative lack of difference between two vastly different amps like 1 and 4 - although I have to say that 1 is dead silent with no input while 4 does generate (a minute amount of ) some white noise that can be heard with the ears close to the tweeters (when there is no input)

Hi superczar..

Thats very true..and a lot of companies do indulge in that fickenry too...same with dac's. Every dac has its own output level. So a dac with a higher output level will immediately sound more detailed. So it is important to level match every component in terms of absolute listener volume if you really want to compare between them.

Another trick is to enhance the treble...So it will look loads more detailed, but will be a tiring listen in the long term.

In the absence of an oscilloscope, the only way I could do a level matching is with a SPL meter - which is what I did for this little experiment.
Would be curious to see if you could try the same with your two amps (that are quite different topologically speaking)

 

[superczar]

Hmm, I'd differ with you on this. A difference of +4db in levels will be quite likely fairly obvious in terms of loudness.

I'd agree with you on the general premise re. amps, as long as they're competently designed and built to do the thing they're supposed to do (they don't have to be expensive, FWIW). My reading suggests that headroom does make a difference. Of course, in those cases, we are talking to about high listening levels that most untreated rooms can't support.

Thanks for doing this, btw!

There is quite a major difference in the headroom.. The PM17 for example runs circles around the rest (with the exception of the HK AVR) on tracks with deep bass at high volumes
My only contention here being that the extra headroom didn’t seem to make a difference at regular volumes

The most curious aspect - In production (I.e at their regular location) , the best soundstage/clarity (by a good margin) of all my setups comes from an ageing KEF iq5 and a relatively cheap streaming amp (Paradigm Link) - the soundstage is almost 3 dimensional vs the relatively flat experience I get with the others

Not sure if it’s the acoustics of the room or the anthem room correction or something else

 

[square_wave]

First a small anecdote
Some months ago, I decided to wire up a unused amplifier to my AVR to drive the front L/R speakers to offset some of the power delivery when I added ATMOS to my setup. While I do believe that offloading the two primaries to a dedicated amp would open up headroom for the overused multi channel amps on the AVR, the end result got me thinking along a different line.
After I had added the stereo amp to the setup , I ran Audyssey (multeq XT32) on the AVR and was a bit surprised when I noticed that Audyssey boosted the gain on the FL/FR channels by 4db
I was under the assumption that gain settings on all amps (at least mid to high grade) would be set to a reference level so the result took me by a bit of surprise.

Now it’s probably well know that psychoacoustically, a few dB extra loudness makes us perceive the sound as fuller/richer - so It got me wondering if the so called amp sonic differences stem on account of the default gain variance.

Thus I carried out a small experiment on my own.
Common thread - All sessions were set to output a 78db loudness level on a pink noise file by using a calibrated microphone wired to a macbook to set the gain on all tested amps so that they output the same amplitude

Amplifiers tested -
1) Marantz PM17 SA integrated amp
2) Marantz SR6010 AVR in direct mode
3) SMSL SA-160 class D amp
4) Harman Kardon AVR 5000 in direct mode
5) Paradigm PW Amp in direct mode (I.e. Anthem room correction disabled)
6) Sure Electronics TAS5630 class D amp board
7) ultra cheap TPA3116D2 board

I am not sure if its my ears or if it’s something else but honest to goodness, other than (7) above, subtle differences aside, I would be hard pressed to select one over the other
Even (7) wasn’t bad at all but seemed to be running out of steam at what is a fairly high dB level.. I guess it would end up being at par (at least for me ) at more normal listening levels

Source for all the tests was a M-audio DAC hooked to a MacBook Pro playing a 24/192 version of The National’s About Today

Hypothesis : Is it possible that a lot of (not all) perceived differences in subjective assessments of Amps are on account of a gain difference rather than a true sonic difference?

PS: (1) and (2) are the AVR and amp referred in the back-story
PS2: Two sets of speakers were used for the tests , KEF Q300 and Phase Tech Velocity V12 - Arguably neither is a top of the line set but both are critically acclaimed mid-end speakers

I may be wrong here but I am assuming you were using the preamp of the AVR while doing the comparison. If that is the case, that would be your bottleneck.

From my experience so far, there are substantial differences between amps. Especially from different brands and topologies. Even if you compare power amps using a level matched procedure from the same brand, if you have a bottle neck in terms of preamp quality, cables, speaker resolving capability, speaker setup, basic listening room acoustics, the difference will be negligible. Once you solve all these, then differences start showing up. Some of these differences might seem small to some while for others it will make or break a system.

 

[G401fan]

The methodology of the original test is seriously flawed. That's because auditory memory is extremely short - barely a few seconds. It would be possible to use the same methodology to conclude that all CDPs sound sinilar, all speakers sound similar, so also cables, and we might all go home and listen to music on our phones.

The only way (though flawed too) to arrive at statistically significant results is with a test group and an A-B-X comparator double blind methodology, carried out over a period of time, with levels perfectly matched. This was the method used by Stereo Review magazine in their infamous 1981 issue (it can be googled) that concluded that ' for this test group, and for this set of conditions, all the amplifiers (8 models from a receiver to an OTL behemoth) sound alike.

Of course, that set the cat among the pigeons, with the audiophile media quick to debunk the test methodology. Among the criticisms were issues related to psycho- acoustics, and the way any 'test' brings in its own issues of stress in the participants, with corresponding results. It was also pointed out that music, like other sensory arts, is not a yes/no situation. It is enjoyed and appreciated differently depending on one's mood, the time of day, whether in company or not, and also based on one's indepth knowledge of the piece in question, AND also the provenance of the playback equipment. Double blind tests only show statistically significant results that have no reference to whether or not the test group had participants that had musical background, or even musical interesots, let alone 'golden ears'.

Among the debunkers of DBT have been people like Art Dudley, Michael Fremer and John Atkinson ( of Stereophile) who have participated in such tests, and have correctly called out one high-end amplifier from another, in double blind testing.

All this does not even approach the issues of loudspeaker load, clipping levels, topology, cables used, and plenty more. And finally, music is for enjoyment, not for stringent comparisons. If you're an aspiring gourmet, would you want to make a case (or even care about) whether all chicken tikka masalas in a bunch if restaurants taste the same?

 

[arj]

I may be wrong here but I am assuming you were using the preamp of the AVR while doing the comparison. If that is the case, that would be your bottleneck.

Quite true..I was actually thinking more in terms of the dac. As I really don't know much about it. Hence the suggestion for a cdp

 

[Kumar442244]

Whoa...The stage for a discussion with difference of opinion is set. M sure we all will have experienced it in the journey in hi-fi. I m from the school of thought who firmly believe that each amp within the class of amplification and otherwise sound different. I have recently moved from a class A amp to a class D and have experienced a lot of difference from the same chain that existed with class A amp. The complete set of presentation is largely different. So from my personal experience, I can safely say that yes, the amp sound different.
I also have a related question which is slightly off topic but pretty much within the preview, it's often considerd that the class D amp is not very widely accepted in the audiophile community, wanted to understand y is it so? Also, is a class A 60 wpc is equivalent to ~110-120 wpc in terms of ability to drive speakers?