I think PSRR may be an issue
One should measure how amp specs change when supply is polluted with mains noise product
My Mirand is the first amp where I cannot hear any change whenever other electronics is turned on...might be important...well, its not completely imune to mains noise as it does sound better late at nights
One should measure how amp specs change when supply is polluted with mains noise product
My Mirand is the first amp where I cannot hear any change whenever other electronics is turned on...might be important...well, its not completely imune to mains noise as it does sound better late at nights
Re: Re: Re: Do amplifiers sound?
I shall, if I can find it.
Direct, please. You had mentioned other types of tests earlier in the thread. Is this what you are alluding to?
Yeah, I know. But I was trying to be brief. Harmonic structure might, if THD or IMD is high enough. Or strange FR or phase. Damping factor comes to mind.
But yes, point taken.
SY said:
I shall, if I can find it.
Direct, please. You had mentioned other types of tests earlier in the thread. Is this what you are alluding to?
AJinFLA said:
Yeah, I know. But I was trying to be brief. Harmonic structure might, if THD or IMD is high enough. Or strange FR or phase. Damping factor comes to mind.
But yes, point taken.
Re: Do amplifiers sound?
I doesn't even to be that complex/extensive/expensive.
Speakers come in pairs (some matched much better than others, I know). Connect two amps in question to the two speakers. Run the same source into two amps under test (level match of course), then observe differential voltage at, say, + side of both speakers while playing your favourite audiophile CD/record/tape. Swap the speakers and repeat and see if differences (if any observed) are the same type/sign (meaning it was the speaker impedance imbalance causing them, not difference between amps).
No need to listen, just measure at speaker end. After all, if voltage input is the same at speaker terminals (on real load), there can't be any audible diffrerence, can it ?
The real crux of the matter is this - we know that we will detect the differences (voltage is easily measured well into nanovolts and below). How small the difference is inaudible ?
And this is where science ends, I'm afraid.
Nikolas Ojala said:
I doesn't even to be that complex/extensive/expensive.
Speakers come in pairs (some matched much better than others, I know). Connect two amps in question to the two speakers. Run the same source into two amps under test (level match of course), then observe differential voltage at, say, + side of both speakers while playing your favourite audiophile CD/record/tape. Swap the speakers and repeat and see if differences (if any observed) are the same type/sign (meaning it was the speaker impedance imbalance causing them, not difference between amps).
No need to listen, just measure at speaker end. After all, if voltage input is the same at speaker terminals (on real load), there can't be any audible diffrerence, can it ?
The real crux of the matter is this - we know that we will detect the differences (voltage is easily measured well into nanovolts and below). How small the difference is inaudible ?
And this is where science ends, I'm afraid.
I must admit I have a hard time accepting that all amps sound the same for several reasons arising from my own experiences. The most relevant of these may be a blind 'speaker cable test I participated in 15 odd years ago run by the local audio club. the listeners had to identify whether the B sample cable was different to the A sample cables. Some tests had sample A twice. The comparisons were conducted in 20 samples without a break.
My score was 19 correct out of 20.
Of course what I may have been hearing was whether the amp was more or less stable driving cable A to cable B. I can't remember the amp after all this time but I was assured it was a belt and braces design with output Zobel; no Naim or Spectral type of issues with driving speaker cables.
It does make you wonder though.
One observation I can make is that I was exhausted by the end of the testing, the one I got wrong was sample 18 or 19.
Rob
My score was 19 correct out of 20.
Of course what I may have been hearing was whether the amp was more or less stable driving cable A to cable B. I can't remember the amp after all this time but I was assured it was a belt and braces design with output Zobel; no Naim or Spectral type of issues with driving speaker cables.
It does make you wonder though.
One observation I can make is that I was exhausted by the end of the testing, the one I got wrong was sample 18 or 19.
Rob
I've tried. Nothing directly applicable for Olive, though one enticing paper I forgot confirming the notion of expert listeners. Clark, one hit:
One for Lipshitz & Vanderkooy:
Two hits a quarter century old. While there can be no argument it's where the answer currently rests as far as the JAES is concerned my searches haven't unearthed an overwhelming body of evidence. Instead I get the impression the question has all the contemporary appeal of finally settling the optimum profile for a smithy's hammer.
One for Lipshitz & Vanderkooy:
Two hits a quarter century old. While there can be no argument it's where the answer currently rests as far as the JAES is concerned my searches haven't unearthed an overwhelming body of evidence. Instead I get the impression the question has all the contemporary appeal of finally settling the optimum profile for a smithy's hammer.
Re: Re: Do amplifiers sound?
And I'm sure you did a blind comparison between daytime and nighttime sound of the amptinitus said:
No it doesn't; just the methodology changes from measurement to blind testing. Neurology and psychology are sciences too, not just physics and engineering.Bratislav said:
Correct assumptions
My assumption is that the compared amplifiers are not identical.
What if the measured voltage signals were a little different?
What would you do after that?
The question is not about what does it do to the voltage at the speaker terminals. The question is about what does it do to the sound coming from the speaker. Yes, I know that there is a very big correlation between them, but still knowing that correlation is not a reason good enough to leave the question unanswered.
The chosen amplifier affects to the loudspeaker and vice versa. When components are connected with a cable, they are parts of one system. We should measure the output of the system, because that is the important thing. The next logical question is "How" and I tried to answer to that question in my earlier post.
"Hey where is that mic when we need it?"
Do we need exactly an expensive anechoic room? Not necessarily anechoic, but a some kind of room built for measurements. The room should be strictly controlled and standardized, so that someone could later repeat tests and get identical results.
That is a great "if" and I would not make that assumption.Bratislav said:
My assumption is that the compared amplifiers are not identical.
What if the measured voltage signals were a little different?
What would you do after that?
The question is not about what does it do to the voltage at the speaker terminals. The question is about what does it do to the sound coming from the speaker. Yes, I know that there is a very big correlation between them, but still knowing that correlation is not a reason good enough to leave the question unanswered.
The chosen amplifier affects to the loudspeaker and vice versa. When components are connected with a cable, they are parts of one system. We should measure the output of the system, because that is the important thing. The next logical question is "How" and I tried to answer to that question in my earlier post.
"Hey where is that mic when we need it?"
Do we need exactly an expensive anechoic room? Not necessarily anechoic, but a some kind of room built for measurements. The room should be strictly controlled and standardized, so that someone could later repeat tests and get identical results.
Hey thanks rdf! I've been diggin thru the JAES catalog too, but am not sure what's worth buying or not.
If anyone has recommendations, please let us know.
I will try to get a list of the articles together, I am ready to buy* and study whatever is needed.
*Too bad my SMPTE membership doesn't get me an AES discount. Time to join AES, I guess. Only need to download 6 articles to pay for it.
If anyone has recommendations, please let us know.
I will try to get a list of the articles together, I am ready to buy* and study whatever is needed.
*Too bad my SMPTE membership doesn't get me an AES discount. Time to join AES, I guess. Only need to download 6 articles to pay for it.
rdf, yes, you will not find anything very new, just like you won't find measurements of the fine structure constant or the e/m ratio in physics journals of the past 30 or 40 years- it's a settled issue.
Unless someone comes along with peer-reviewable data to the contrary. Here's your chance.
Unless someone comes along with peer-reviewable data to the contrary. Here's your chance.
SY said:
I don't think that is correct. I know the "Swedish Audio Society" (LTS) use to blind test amps that should be transparent according to your claims, however most of the amps tested do color the sound. Think they have tested a hundred amps or something like that and I think two amps have ended up being non detectable in the path.
The same goes for AD>DA converters. Preamps are relatively often transparent though as I understand it. Pre amps also normally has lower distortion and flatter Fr. and typically has an easier job driving a high resistive load.
I've never heard two amps that sound the samt and even though I did not have means to measure them all they do fit within your standards according to manufacturers own info... GamuT, Rowland, Yamaha, LC audio, Pass and many more. Oh, I should mention I'm talkin about way below clipping.. as a couple of watts of output only or less than that.
/Peter
Subjective evaluation……..
Back when I has in high school our Tech Studies teacher showed the class how to fusion weld two pieces of flat mild steel together using the oxyacetylene kit, then left us to it.
On my first attempt I pulled off a ~10cm long weld that was better than the teachers demo.
I proudly took it up to the teacher’s desk for evaluation and to my dismay the prick gave it a lousy “C”, handed it back to me and told me to try again.
So I wander off to a secluded corner of the workshop, sit down and eat a bag of chips. Fifteen minutes or so later I front up at his desk, wipe my brow and plonk the same bit of metal back on his desk for evaluation. He examines it closely, comments positively on my progress and awards me a “B”.
I exclaim that I will try for an “A”. He lets me take my bit of metal back, I go through the same routine once again (I really can’t remember what I ate or did the second time for 15 minutes though) and I get my hard earned “A”.
Back when I has in high school our Tech Studies teacher showed the class how to fusion weld two pieces of flat mild steel together using the oxyacetylene kit, then left us to it.
On my first attempt I pulled off a ~10cm long weld that was better than the teachers demo.
I proudly took it up to the teacher’s desk for evaluation and to my dismay the prick gave it a lousy “C”, handed it back to me and told me to try again.
So I wander off to a secluded corner of the workshop, sit down and eat a bag of chips. Fifteen minutes or so later I front up at his desk, wipe my brow and plonk the same bit of metal back on his desk for evaluation. He examines it closely, comments positively on my progress and awards me a “B”.
I exclaim that I will try for an “A”. He lets me take my bit of metal back, I go through the same routine once again (I really can’t remember what I ate or did the second time for 15 minutes though) and I get my hard earned “A”.
As SY pointed out in the last sentence, there are examples where higher numbers didn´t lead to rejection of the nill hypothesis in a double blind test; see the great stereophile amplifier dbt and the older HN/RR-DBT.
But that brings up the question if the methodology choosen was appropriate; there are things to argue because these were discrimination tests and afair no controls were included.
I could post some more related papers/articles from AES-Resources and links to other related discussions later on if there is some interest.
@ Nikolas Ojala,
it is an interesting experience to listen to the recorded output of a loudspeaker driven by different amplifiers. At least it could make the comparison a bit easier but you have to be careful about additional varaibles, just to name possible influence of power compression.
@ G.Kleinschmidt,
Yep, subjective evaluation can lead to ridiculous decisions.
But, on the other hand that holds true for blind test evaluation as well.
Wishes
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.