Do all amplifiers sound the same? Level matched listening test

[pma]

Do all amplifiers sound the same? Level matched listening test

One more to this neverending discussion. Two very different amplifiers are tested, however both with very flat frequency response in the audio band. The test level is set such that 0dBFS of digital data corresponds to THD = 0.9% for both amplifiers under test (that means close to clipping). The music sample has maximum digital amplitude of -0.1 dBFS and DR = 12. Outputs of both amps are recorded via Cosmos ADC, time aligned and level matched (to average RMS). The ADC recorded music samples have maximum amplitude of -3.8dBFS approx.

The test files can be downloaded from:

https://pmacura.cz/amp_XY.zip

They are in 44.1kHz/24bit wav files.
The test questions are in the poll. If you hear the difference, it would be good to post your ABX test report as well.
Enjoy the listening and have fun.

 

[jEDGEc0m]

if anything, mono should make it "easier" to hear any difference, right?

I gave it a little go, on familiar speakers with room correction, not too loud, in somewhat treated small room. didn't try too hard. but I'm fine with saying "I cannot hear a difference", and voted as such.

Null test indicates there is a difference.

For a laugh I then just went through 11 trial ABX hitting first option everytime, and it resulted 7/11.

foo_abx 2.2.1 report
foobar2000 v2.24.1
2025-08-31 18:36:51

File A: amp_X.wav
SHA1: c60451229fba3a73fc22f6156b9ed9490072ff8f
File B: amp_Y.wav
SHA1: 70bc370670bdfe1bbedd5ffb635634cef04ff6e6

Output:
Default : Primary Sound Driver
Crossfading: NO

18:36:51 : Test started.
18:36:54 : Test restarted.
18:36:54 : 00/01
18:36:56 : Test restarted.
18:36:56 : 01/02
18:36:58 : Test restarted.
18:36:58 : 01/03
18:36:59 : Test restarted.
18:36:59 : 02/04
18:37:01 : Test restarted.
18:37:01 : 03/05
18:37:03 : Test restarted.
18:37:03 : 04/06
18:37:05 : Test restarted.
18:37:05 : 04/07
18:37:07 : Test restarted.
18:37:07 : 05/08
18:37:09 : Test restarted.
18:37:09 : 06/09
18:37:12 : Test restarted.
18:37:12 : 07/10
18:37:16 : Test restarted.
18:37:16 : 07/11
18:37:16 : Test finished.

----------
Total: 7/11
p-value: 0.2744 (27.44%)

-- signature --
f4b0a33b0ee5b3e69f172c5aaae746216a7d82b7

after the null test, with that as maybe a hint what to listen for, I tried again a little harder, with Sennheiser HD 560s. I'd only be guessing again.

Although arguably less relevant, pink noise samples would surely give a much better chance of simply defining whether a difference is/might be audible.

 

[Chr1]

I now believe that all amps sound the same unless they are either clipping or otherwise distorting.

The "distortion" bit can either be: deliberate/by design, as with many valve and Class A amplifiers. ("Warmth")
Or otherwise (ie poor design/implementation)

 

[Petrushka]

Audible does not imply important.

 

[jEDGEc0m]

I now believe that all amps sound the same unless they are either clipping or otherwise distorting.

The "distortion" bit can either be: deliberate/by design, as with many valve and Class A amplifiers. ("Warmth")
Or otherwise (ie poor design/implementation)

Seems amps that are load-dependent could change frequency response of a speaker enough to be potentially perceived as different. but may still be entirely fine in practice such that it doesn't really interfere with enjoying the music/program.

 

[Chr1]

Yes, I guess I would class load dependant FR change as a type of distortion (to the signal) too!

As is "harmonic" etc

 

[jEDGEc0m]

Yes, I guess I would class load dependant FR change as a type of distortion (to the signal) too!

As is "harmonic" etc

Fair enough. it's like how sometimes the term "filter" is used for a huuuuge breadth of audio processing/change.

 

[DVDdoug]

if anything, mono should make it "easier" to hear any difference, right?

It's possible to have a channel imbalance which would be a legitimate difference and/or some people claim to hear differences in soundstage.

Audible does not imply important.

True, but no audible difference does demonstrate something very important!

I'm not going to bother with the test because I don't expect to hear a difference. So although I'm in the "all amplifiers sound the same" camp, when I'm being more careful I say "all good amplifiers sound the same" because as Chr1 said, there are poor quality or defective amps.

 

[Greenman]

when I'm being more careful I say "all good amplifiers sound the same" because as Chr1 said, there are poor quality or defective amps.

Surely. It’s more - all good amplifiers when driving an easy load , well within their power capabilities.

Big difference between sounding the same at 75db to 95db in a large room with a difficult to drive speaker.

 

[Chr1]

"Good" and" bad" amplifiers.

Bit too simplistic for me.

I don't see a modern well measuring, yet throwaway Class D amplifier as any better than a poor measuring valve amplifier, that sounds as it was designed to. And lasts way more than a year or two.
Even with deliberate harmonic distortion etc.

Can be particularly useful for those who are all analogue and can't easily employ DSP.
(I don't do vinyl myself but, as with analogue EQs, tone controls etc, can see a valid use case.)

 

[pma]

if anything, mono should make it "easier" to hear any difference, right?

I gave it a little go, on familiar speakers with room correction, not too loud, in somewhat treated small room. didn't try too hard. but I'm fine with saying "I cannot hear a difference", and voted as such.

Null test indicates there is a difference.

For a laugh I then just went through 11 trial ABX hitting first option everytime, and it resulted 7/11.

foo_abx 2.2.1 report
foobar2000 v2.24.1
2025-08-31 18:36:51
.
.
Total: 7/11
p-value: 0.2744 (27.44%)

-- signature --
f4b0a33b0ee5b3e69f172c5aaae746216a7d82b7

after the null test, with that as maybe a hint what to listen for, I tried again a little harder, with Sennheiser HD 560s. I'd only be guessing again.

Although arguably less relevant, pink noise samples would surely give a much better chance of simply defining whether a difference is/might be audible.

Thank you for trying the test and posting the result. My is not any better, just like coin flip. And I even do not feel I can hear any difference.

At the moment, I do not want to reveal what is tested. I can only tell there is no cheating and that the test was well prepared, using a very good DAC and ADC. The difference in THD+N of the amps under test is about 20 - 30 dB. They both keep their parameters up to some -1dB below clipping level, clipping understood as THD = 1% approx., also is visible in time domain. There is no gradual rise of distortion with power as in case of tube amplifiers.

The frequency response of both amplifiers is flat within 0.1dB tolerance range from 20Hz to 20kHz, with the 4ohm load.

 

[MAB]

Can be particularly useful for those who are all analogue and can't easily employ DSP

How so?

 

[Chr1]

To adjust the sound to taste. As with phono preamps, graphic EQs etc. Definitely old school, but a still valid reason to buy a "voiced" amplifier imho.

DSP is obviously far superior, but many folks, including pro producers, musicians still use such things in both production and playback.
Usually optional and obviously, not strictly as "hifi" but not necessarily "bad" either.
My 2c.

 

[Ropeburn]

if anything, mono should make it "easier" to hear any difference, right?

I gave it a little go, on familiar speakers with room correction, not too loud, in somewhat treated small room. didn't try too hard. but I'm fine with saying "I cannot hear a difference", and voted as such.

Null test indicates there is a difference.

For a laugh I then just went through 11 trial ABX hitting first option everytime, and it resulted 7/11.

foo_abx 2.2.1 report
foobar2000 v2.24.1
2025-08-31 18:36:51

File A: amp_X.wav
SHA1: c60451229fba3a73fc22f6156b9ed9490072ff8f
File B: amp_Y.wav
SHA1: 70bc370670bdfe1bbedd5ffb635634cef04ff6e6

Output:
Default : Primary Sound Driver
Crossfading: NO

18:36:51 : Test started.
18:36:54 : Test restarted.
18:36:54 : 00/01
18:36:56 : Test restarted.
18:36:56 : 01/02
18:36:58 : Test restarted.
18:36:58 : 01/03
18:36:59 : Test restarted.
18:36:59 : 02/04
18:37:01 : Test restarted.
18:37:01 : 03/05
18:37:03 : Test restarted.
18:37:03 : 04/06
18:37:05 : Test restarted.
18:37:05 : 04/07
18:37:07 : Test restarted.
18:37:07 : 05/08
18:37:09 : Test restarted.
18:37:09 : 06/09
18:37:12 : Test restarted.
18:37:12 : 07/10
18:37:16 : Test restarted.
18:37:16 : 07/11
18:37:16 : Test finished.

----------
Total: 7/11
p-value: 0.2744 (27.44%)

-- signature --
f4b0a33b0ee5b3e69f172c5aaae746216a7d82b7

after the null test, with that as maybe a hint what to listen for, I tried again a little harder, with Sennheiser HD 560s. I'd only be guessing again.

Although arguably less relevant, pink noise samples would surely give a much better chance of simply defining whether a difference is/might be audible.

A good demonstration of the importance of sample size!

 

[Speedskater]

There are different levels of amplifier goodness.
Not all amplifiers playing into every loudspeaker at every volume level is good.

 

[Chr1]

Q: Is a Class D amplifier that has super ( but, inaudible) SINAD, and lasts 2 years better than... a Class A/B that measures good but not super good, and lasts 20?
Both at the same price point.

Apologies Pavel, I am way OT...
(And out.)

 

[Hear Here]

I now believe that all amps sound the same unless they are either clipping or otherwise distorting.

The "distortion" bit can either be: deliberate/by design, as with many valve and Class A amplifiers. ("Warmth")
Or otherwise (ie poor design/implementation)

What a sheltered life you've led. Far from all amps sounding the same, there's a massive difference despite what readings might suggest.

First, you need good quality speakers to judge the ability of an amp to get them to sing in such a way as to give you goosebumps. Don't bother looking for an exciting amp if you have drearily dull speakers.

With good speakers, borrow or buy a bunch of amps of Class A, AB, D, SET, OTL, or whatever and spend a few days listening to one and then another. You'll soon realise that not all amps sound the same, despite similar measurements.

An extreme point to illustrate this. I bought a Benchmark AHB2 amp on a rave recommendation and its exceptional measurements to replace SETs I'd been previously using. Instead of listening to music, I found myself listening to Muzak such that I wanted to turn down the volume - kiss of death surely. I bought or borrowed numerous ss amps and listened to each for between a week and a year to establish their worthiness in my system. Some were dire while others excited the soul and neither group was influenced by price.

 

[audio_tony]

How was the test conducted? Was the amp driving a dummy load, and the output taken from the dummy load into the ADC?

 

[Chr1]

What a sheltered life you've led. Far from all amps sounding the same, there's a massive difference despite what readings might suggest.

First, you need good quality speakers to judge the ability of an amp to get them to sing in such a way as to give you goosebumps. Don't bother looking for an exciting amp if you have drearily dull speakers.

With good speakers, borrow or buy a bunch of amps of Class A, AB, D, SET, OTL, or whatever and spend a few days listening to one and then another. You'll soon realise that not all amps sound the same, despite similar measurements.

An extreme point to illustrate this. I bought a Benchmark AHB2 amp on a rave recommendation and its exceptional measurements to replace SETs I'd been previously using. Instead of listening to music, I found myself listening to Muzak such that I wanted to turn down the volume - kiss of death surely. I bought or borrowed numerous ss amps and listened to each for between a week and a year to establish their worthiness in my system. Some were dire while others excited the soul and neither group was influenced by price.

Doh. Don't think you actually understood the points that I was trying to make.

PS. I'll have your AHB2 if you don't like it. Ta.
Lol.

 

[MAB]

To adjust the sound to taste. As with phono preamps, graphic EQs etc. Definitely old school, but a still valid reason to buy a "voiced" amplifier imho.

DSP is obviously far superior, but many folks, including pro producers, musicians still use such things in both production and playback.
Usually optional and obviously, not strictly as "hifi" but not necessarily "bad" either.
My 2c.

Can you post a link that shows effective use of a tube amp to EQ a speaker, with the frequency response deviation, or some measure of a “voiced amplifier” actually being used to actually improve the response of a speaker in a room? I’ve never seen one example that is even dimly useful.

I now believe that all amps sound the same unless they are either clipping or otherwise distorting.

The "distortion" bit can either be: deliberate/by design, as with many valve and Class A amplifiers. ("Warmth")
Or otherwise (ie poor design/implementation)

I can’t figure out what you are saying, your comments are mutually exclusive.

The output impedance of many tube amps is high. This will change the sound, audibly with some speakers. Not in a useful way, unless you are into Rube Goldberg EQ.