Thank you very much indeed
I think I can say that I have solved some important doubts
There are statements like the measurements mean nothing or at least traditional ones are of no use that have confused me.
Now I am more convinced that an audio equipment should have very good performance at least with traditional tests (i.e. noise, distortion, square wave) to be considered faithful to the original sound
Then I understand that also matching with the other elements of the audio chain and the listening room is of fundamental importance
Just for curiosity I downloaded this nice free software
LF Generator
and performed a sweep
In the bass range I felt at least two remarkable variations of level, sign of some nasty ambient reflections I suppose
And I am pretty sure that a sound level meter could show other bad things in the freq response
To end I would like to thank you all to take me down again on the rational planet
I promise ... no more silly question ... a part one on caps reforming but in another place of course
Have a nice day
Kind regards,
gino
Last edited:
Excuse me, when you say better results you mean also better measurements don't you ?
I am trying to fix in my mind this important belief, better measurements=better sound
There is no reason to say differently but any confirmation is welcome
Kind regards,
gino
Just one last thing
Many years ago I had an email exchange with a person much more expert than me (not a big deal actually )
Anyway he was so kind to simulate a 2 bjts line stage for me
He sent me some amazing distortion spectra
I mean, once we have established that good measurements are the aim they can be obtained even with simple, if not minimalist, topologies
Moreover I have seen discrete op-amps extremely complex and other much simpler
It would be only weird that the simpler ones would measure better in the end
Maybe, and I stress maybe, increasing the voltage rails leads to better measurements instead of complicating the circuit
Not too simple but not too complex, especially if this added complexity does not translate in better measurable performances.
Thanks again and kind regards,
gino
Many years ago I had an email exchange with a person much more expert than me (not a big deal actually
)Anyway he was so kind to simulate a 2 bjts line stage for me
He sent me some amazing distortion spectra
I mean, once we have established that good measurements are the aim they can be obtained even with simple, if not minimalist, topologies
Moreover I have seen discrete op-amps extremely complex and other much simpler
It would be only weird that the simpler ones would measure better in the end
Maybe, and I stress maybe, increasing the voltage rails leads to better measurements instead of complicating the circuit
Not too simple but not too complex, especially if this added complexity does not translate in better measurable performances.
Thanks again and kind regards,
gino
Last edited:
this may apply to honest manufacturers who know what they are doing and to some DIYers too.Just one last thing
Many years ago I had an email exchange with a person much more expert than me (not a big deal actually
Anyway he was so kind to simulate a 2 bjts line stage for me
He sent me some amazing distortion spectra
I mean, once we have established that good measurements are the aim they can be obtained even with simple, if not minimalist, topologies
Moreover I have seen discrete op-amps extremely complex and other much simpler
It would be only weird that the simpler ones would measure better in the end
Maybe, and I stress maybe, increasing the voltage rails leads to better measurements instead of complicating the circuit
Not too simple but not too complex, especially if this added complexity does not translate in better measurable performances.
Thanks again and kind regards,
gino
but it's not always the case. some people do the discrete thing just because. and I mean just because.
moreover, I suspect that there are manufacturers which do not own and/or use a good scope or any at all. and maybe they even think they understand electrons, but they don't.
there can't be any other explanation for the fact that some discrete circuits measure worse even in the most basic ways compared to cheaper integrated versions. there are examples out there.
one other thing is that there's not much point in trying to get your head around that cognitive dissonance. I'm sure there are preconceived ideas even in math. do you wonder about audio?
don't you agree that the fact that preference for a certain circuit/topology can be explained by subjective preference and system synergy? I mean, I'm sure there are situations where some amount of distortion is desirable, even if only for the fact that it masks other problems or it acts in consonance with them.
the fact that a large number of people think this or that to be true doesn't say much. 9 out of 10 people just repeat stuff they hear or read. even the best of us can be manipulated.
and I repeat. I am sure there are people who implement simpler circuits for a reason, be it even financial. but I'm also sure other (many?) just do it because they follow some dogma.
what I'm trying to say in a convoluted way is that IMO you're asking a very general question which has many answers.
it is possible to bend/skew the test, just to make worse product win against better one
remember report where hummer wins in enviromental against hybrids ?
yeah right.... 10mpg is good enough
it is possible to bend/skew the test, just to make worse product win against better one
remember report where hummer wins in enviromental against hybrids ?
yeah right.... 10mpg is good enough
I am pretty sure that if a listening test is done correctly if could be found a sort of correlation between listening experience and measurements
For instance I remember a comparison at a friend's house of two preamps one tube and one solid state in the same rig
With the solid state one the system gained one octave in the low
I am pretty sure that the tube preamp was not driving properly the power amp
Something that the right set of measurements could have predicted I guess
Now I am again rational ...
With you kind and valuable help
Thank you !
Kind regards,
gin
it's not as esoteric as it may seem
http://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1116-1-199710-I!!PDF-E.pdf
Are we sure that everyone even has remotely similar criteria by which they judge an audio system? To me it seems obvious that a good audio system is going to have to reproduce cleanly the sound of an orchestra at something like realistic levels. It's going to include tinkly high stuff, and huge bass 'whoomps'! If it can do that, it can also do rock, and middle of the road jazz etc. Only after establishing that would I begin to worry about the nebulous, probably imaginary, stuff.
But is this how most people judge a system? There seems to be an awful lot of Dave Brubeck's Take Five being played at audio shows, and that would sound the same on my superb (of course!) speakers or any old valve radiogram from 60 years ago. Or there are people sitting in reverential silence listening to the tinkliest piano music barely emerging from huge horns. Or listening to compressed AOR. None of it, it seems to me, is a particularly good test of an audio system. It makes me wonder if we are all talking at cross purposes, and when people say how wonderful minimal amplifiers and speakers are, does it really apply to the sort of music I listen to.
But is this how most people judge a system? There seems to be an awful lot of Dave Brubeck's Take Five being played at audio shows, and that would sound the same on my superb (of course!) speakers or any old valve radiogram from 60 years ago. Or there are people sitting in reverential silence listening to the tinkliest piano music barely emerging from huge horns. Or listening to compressed AOR. None of it, it seems to me, is a particularly good test of an audio system. It makes me wonder if we are all talking at cross purposes, and when people say how wonderful minimal amplifiers and speakers are, does it really apply to the sort of music I listen to.
I love you.
LOL
I COMPLETELY agree that the typical crap found on the Chesky test CDs is completely irrelevant for evaluation. there are many "usual" (as in not specifically made for testing) CDs that are way better. crowded passages, transients, loud passages, silent passages, all in the same song. AND musically engaging.
simple music sounds good even on PC speakers.
I laughed out loud when a guy from ESS said during a RMAF presentation that "he has come to love the audiophile favorite Spanish Harlem". what kind of people listen to that type of music? if I were an audio salesman, seeing a guy bring a Chesky CD to the showroom would mean "audio snob alert".
@mr_push_pull
I have had a very sheltered upbringing and I didn't even know about Chesky CDs! I see what you mean, though. There's clearly an entire sub-culture out there, of middle-of-the-road anodyne music that is offensive in its inoffensiveness.
It's a quandary as to what music you should use to 'test' a system, I think. As you are going to be hearing the same music over and over again on good and bad systems, it would seem wasteful to use music that you actually like - you'll end up sick of it, which would be a shame. Perhaps the best stuff would be something you've already listened to death, so you know it inside out, but are not so influenced by the music itself any more.
I always find that the first two seconds of this at high volume tells you what you're going to get (non-minimalist systems vs. minimalist, or large speakers vs. small) but the whole track is thrilling:
Stravinsky: The Firebird (L'oiseau de feu) - Infernal Dance of All Kashchei's Subjects by Valery Gergiev on Spotify
I have had a very sheltered upbringing and I didn't even know about Chesky CDs! I see what you mean, though. There's clearly an entire sub-culture out there, of middle-of-the-road anodyne music that is offensive in its inoffensiveness.
It's a quandary as to what music you should use to 'test' a system, I think. As you are going to be hearing the same music over and over again on good and bad systems, it would seem wasteful to use music that you actually like - you'll end up sick of it, which would be a shame. Perhaps the best stuff would be something you've already listened to death, so you know it inside out, but are not so influenced by the music itself any more.
I always find that the first two seconds of this at high volume tells you what you're going to get (non-minimalist systems vs. minimalist, or large speakers vs. small) but the whole track is thrilling:
Stravinsky: The Firebird (L'oiseau de feu) - Infernal Dance of All Kashchei's Subjects by Valery Gergiev on Spotify
Logically, shouldn't the arguments in favour of minimalism in audio, also apply to optics..?
Sure enough, everything else being equal, it would be better to look through a thin window than a thick one. But what other similarities are there?
We could base a working camera (or projector) on a single pinhole. It would work after a fashion, but not be very sensitive (or powerful).
We could use a single lens which would be OK in the middle of the picture, but out of focus at the edges, with some colour fringing, rather like a child's toy camera.
But, as designers have known for decades, the key to real quality from the centre of the image to the edges is to make a compound lens from many elements, using different materials with different refractive indexes.
Why is there no movement to get back to minimal optics? Why do people not proclaim the pinhole camera as the ultimate for true quality? Have you ever looked through a pair of binoculars or a camera with complex compund lens and thought "This is OK, but it lacks the 'opticality' of my first toy camera when I was a child. If I had some spherical lenses made out of high quality glass, and mounted them in a hand-made ebonite box with knurled brass knobs, it would produce the ultimate images. It's all about the materials. Why, oh why, have the evil corporations taken us to this place where lenses have a dozen different elements? If only they had my insights." etc. ?
Sure enough, everything else being equal, it would be better to look through a thin window than a thick one. But what other similarities are there?
We could base a working camera (or projector) on a single pinhole. It would work after a fashion, but not be very sensitive (or powerful).
We could use a single lens which would be OK in the middle of the picture, but out of focus at the edges, with some colour fringing, rather like a child's toy camera.
But, as designers have known for decades, the key to real quality from the centre of the image to the edges is to make a compound lens from many elements, using different materials with different refractive indexes.
Why is there no movement to get back to minimal optics? Why do people not proclaim the pinhole camera as the ultimate for true quality? Have you ever looked through a pair of binoculars or a camera with complex compund lens and thought "This is OK, but it lacks the 'opticality' of my first toy camera when I was a child. If I had some spherical lenses made out of high quality glass, and mounted them in a hand-made ebonite box with knurled brass knobs, it would produce the ultimate images. It's all about the materials. Why, oh why, have the evil corporations taken us to this place where lenses have a dozen different elements? If only they had my insights." etc. ?
Going ahead with this very interesting comparison i wonder if a test analogous (on principle) to this one
MTF - Modulation Transfer Function
could be developed also for audio circuits.
It is widely accepted for lenses testing i suppose.
At least for "transparency" in sound ....
Regards,
gino
MTF - Modulation Transfer Function
could be developed also for audio circuits.
It is widely accepted for lenses testing i suppose.
At least for "transparency" in sound ....
Regards,
gino
Last edited:
Sorry the last provocative question
This is the schema of a well reviewed headphone amp
As with low impedance headphones (i used this type) even a good buffer without voltage gain could be enough, it may be possible that a 4 bjts diamond buffer would sound better than this one ?
If the answer is yes i call this perversion ... not only overdesign
The idea to go with very complex topologies without having explore all the possibilities of simpler ones for me is insane
Only when i am sure that top performances cannot be obtained with simpler topologies i would search for more
I do not know what is the drive sometimes
Is not still valid the KISS principle ?
I swear ... no more ramblings
Good day !
gino
This is the schema of a well reviewed headphone amp
As with low impedance headphones (i used this type) even a good buffer without voltage gain could be enough, it may be possible that a 4 bjts diamond buffer would sound better than this one ?
If the answer is yes i call this perversion ... not only overdesign
The idea to go with very complex topologies without having explore all the possibilities of simpler ones for me is insane
Only when i am sure that top performances cannot be obtained with simpler topologies i would search for more
I do not know what is the drive sometimes
Is not still valid the KISS principle ?
I swear ... no more ramblings
Good day !
gino
Last edited:
I've tried discrete ss headamp circuits and they rarel end up minimalist, always a nasty compensation cap to prevent oscillation or a dc servo which imo is nothing more than a electronic capacitor.
I swear by minimalist. For phones like Grados and Senns all one need is a spud SET. A 6c45pi series fed output transformer, led bias, SSHV B+, separate chassis. The signal current loop as Lyn Olsen refers is just nothing no caps, and it even measures good +-1db 20-20k. Even 1khz 2H distortion is below audibility at normal listening levels. It is as close to a wire with gain as you can get. Quite a few have built around this topology.
I swear by minimalist. For phones like Grados and Senns all one need is a spud SET. A 6c45pi series fed output transformer, led bias, SSHV B+, separate chassis. The signal current loop as Lyn Olsen refers is just nothing no caps, and it even measures good +-1db 20-20k. Even 1khz 2H distortion is below audibility at normal listening levels. It is as close to a wire with gain as you can get. Quite a few have built around this topology.
Hi ! thank you for the valuable reply but let me elaborate a little
I think that the correct way is to start from the purpose of a circuit
I am not sure but i think that some line preamps could drive properly high impedance headphones.
My point is that on this basis a line preamp is the unit more similar to a headphone amp
So why do not take a nice and simple line preamp and beef it up ?
Take an elegant schema like this one below (i am referring to the output amplifier module)
change output devices for something stronger, maybe use +/- 24V ...
and it is done isn't it ?
What if a schema like this sounds better than the other hugely more complex ? wouldn't be it weird ?
A line preamp, or a very low power amp, are the closest thing to an headphone amp
So let's start from some nice examples of this kind of equipment
An exceptionally simple and good units of this kind do exist
The problem is to fine tune them maybe
I do not know ... i am confused by this love for complexity... really
When you say
is there anything, apart Stax, left out ?
I understand that you like tubes. I like more minimalism than tubes.
Because i am sure that something good and simple can be done even with SS
But yes, tube circuits are fundamentally minimalist and this i like very much
But again between minimalism and ultra complex a reasonble complexity can be accepted easily
Sometime in some headphone amps i see a huge number of tiny output transistors paralleled .... why not just a serious pair ?
Or again ... why instead of a driver and a output bjt not using a darlington ?
One instead of two
Speaking of heapdhone amps the excellent Grace M901 was built around only one op-amp
And it was a professional device able to drive anything
Again ,,, i do not understand this love for elaborated units
It is like those cuisines extremely elaborate where maybe the raw materials are ****
The very good meat is sublime even raw
Thanks again
Kind regards,
gino
Last edited:
well its a misconception that headphone amps are similar to line amps, a dynamic driver of 32 ohms has nothing in common to an amp input of 10k. I am a fan of tubes as well as transformer coupling. When the audible distortion and noise is below audibility with 98dB/w phones its hard to argue that any thing is more minimalist than a single active component of the most linearity device ever invented (triode) in common cathode stage, you should try it. These amps measure better than you can believe. Granted there is help in SS with the power supplies, but they are uniquely out of the signal current loop.
I agree with you
A line preamp must have a low voltage gain and be able to drive 10k with some 500pF (the load of a power amp)
And headphone amp should have a more complex task because headphones can go from 30 to lets say 300 ohm with different sensibilities
I dont say to use a line preamp to drive headphones
I say to take a line preamp schematic and mod it to drive headphones
Then everyone is free to choose what he likes
If he likes a line buffer with 150 bjts free to buy and use it
I would prefer a diamond buffer ... even only from an aesthetic point of view
It is so much more elegant
Regards,
gino
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.