itisprobably true, in one box, the ground is shared so both power supply float, there are currents in the ground and can affect , the power transformer will also modulate on one another, my amps have separate power supply enclosures, 4 boxes
It does makes a subtle difference, probably...
As for subjective appreciation, you have good sound when you convince yourself it is good , pursuing this logic, the owner of his new twice powerful amp is now convinced he is listening to improved dynamics, sense of extended bass and control , weather is true or not isnt important, It is his conviction that he found the solution to the riddle.
Many change caps , wires, isolators, special metal contacts, dsp xo, etc and succeed at convincing themselves that this is the way, the THD club is like that and the sellers and engineers of amplifiers know that and exploit it. I will not say if I approve / like or / disapprove Mr Pass, it is not important, what is important is his appreciation. And consideration of analysis of what exactly happens to music beyond THD figures.
The Thd is like the size of the engine, it says nothing else
It’s a great discussion.
My piano analogy definitely isn’t perfect and I totally agree the harmonics present in a piano note aren’t literal distortion but as a thought experiment it illustrates that harmonics (or at least, non-linearities) exist anywhere in nature that a wave travels through a medium. Sometimes desirable, other times not.
Really the only place we differ is on notions about tube amps. Where you see a limitation of the technology, I see an aesthetic choice. (Not sure if it’s still the case, but at one point Schiit offered two different headphone amp models, one with vanishingly low THD and another with…well, with tubes in it. The tube amp’s product literature literally bragged about the distortion and directed anyone with a problem to their other amp 😂).
That said, recently I heard a class D setup that absolutely blew my mind. Flew in the face of everything I thought I knew about how a “good” playback system should work. More on that later… it’s late here.
More on this, too! I have many thoughts on this subject that I’d like to talk with an engineer about.
Not correct. Anything about the sound wave entering our ears can be measured to more detail than human perception is capable. What cannot be measured is how your ears and brains perceive the sound as that is subjective. So our tools are measuring as expected but your brains are measuring through the lens of your timbral perception which is quite likely different than mine or anybody elses.
I attended an amp comparison, many years ago:
* same speakers,
* same CD source
* just a different amp.
With one amp ... the singer appeared to be in line with the plane of the front of the spkrs; with the other amp ... the singer jumped forward about a metre.
What would you measure to be able to explain why this happened?
Were the levels matched to 0.1dB? Higher mismatch can have audible impact even in soundstage. Also I assume this was a sighted listening session.
Nothing wrong with the usual measurements. On top of that you can record the amp outputs with a high quality ADC and analyze and compare the recordings in detail with e.g. DeltaWave (https://deltaw.org/). Before making claims of "unmeasurable audible impact" you should first make an effort to find the differences with the forementioned methods.
Nothing wrong with the usual measurements. On top of that you can record the amp outputs with a high quality ADC and analyze and compare the recordings in detail with e.g. DeltaWave (https://deltaw.org/). Before making claims of "unmeasurable audible impact" you should first make an effort to find the differences with the forementioned methods.
Your input, I'm afraid, is completely valueless.
I asked you to define what should be measured, in order to explain the difference we heard.
If you in fact do know what you are talking about ... IMO, you should be able to say: "Measure this"! "Measure that"! The fact that you have not provided this simple answer suggests to me that whilst you "talk the talk" ... you absolutely don't know how to "walk the walk".
No problem - it simply reinforces my belief that "Not everything worthwhile can be measured, and not everything that can be measured is worthwhile". 😀
Assuming that question was addressed to all 🙂 I would check several things like:
- THD + harmonics profile and phase of the second harmonic.
- Resulting amplifier + loudspeakers acoustic frequency response.
Now, I’m nowhere near golden ears club but have determined (using very precise balance controls) that 1 dB shift in channel balance is a very noticeable soundstage shift for me, while below 0.5 dB I can’t tell.
In the same way, 1 dB change in the important vocal FR range can shift perceived virtual source position of a singer.
There are certainly other possibilities, but I would look first there.
Thank you, @tombo56 . 👍 Some real measurements that could explain the difference in presentation.
The first amp (and remember, I said this test was done quite a while ago) was a Bryston - so low overall THD and very powerful.
The second amp was a SE ss amp of about 40w (into 8 ohms). Being SE, I would expect it had a much higher H2 component - though whether this was '+' or '-' ... I have no idea. Though from what you wrote ... it would have to be '+'; I will ask the designer. 👍
The first amp (and remember, I said this test was done quite a while ago) was a Bryston - so low overall THD and very powerful.
The second amp was a SE ss amp of about 40w (into 8 ohms). Being SE, I would expect it had a much higher H2 component - though whether this was '+' or '-' ... I have no idea. Though from what you wrote ... it would have to be '+'; I will ask the designer. 👍
It is valueless if you don't know anything about measurements. As I said before sound perception is subjective so it is fairly difficult to know what to measure based on your subjective statement.
Why don't you start with the level matching. Most sighted listening sessions fail even at this first step.
Your post simply reinforces Brandolini's law.
With the SE amplifier, probably more influence was by its higher output impedance, compared to the Bryston, and resulting FR change, especially if loudspeakers were a heavy load.
I've always thought this to be the case, and I've fiddled with it using the H2 generator.
The phase of an even component HD with relation to perceived imaging being "front or back" has always made sense to me, but when things seem so simple, my understanding is usually wrong. But ... here goes nothin' to see if my understanding of "why" in the most basic terms is aligned with reality.
If the (let's say) 2nd harmonic is "positive" (in the vernacular being used commonly around the forum), then the distortion product (voltage) is 'added' to the fundamental at their relative peak voltages. Edited to add - I haven't done the math to see if the "overall net" across the entire waveform is higher or lower. Thus... that frequency would be very slightly louder. Thus bringing it forward in the image.
If it is "negative" then the distortion product is 'subtracted' from the fundamental and the resulting voltage is slightly lower...
If an amplifier has a consistent distortion behavior across the frequency range and within the power band used to play the music... then the "imaging" or placement of instruments / voices / room effects may likely be more consistent / less "muddy" perhaps.
If an amplifier varies the distortion characteristics (and many many that I have measured do vary) both across the frequency spectrum and within the power band ... then the soundstage can perhaps be more "muddy" or less defined.
If there is little to no meaningful distortion, then one could likely assume that the soundstage from the recording may be more "accurate". Is it more pleasing... who knows?
Since we know how our ears / brains detect spatial placement (see below) ... it seems plausible, but I'm not willing to conduct a definitive study.
"It includes differences in the precise time at which the sound reaches each ear (interaural time difference, ITD) and the intensity of the sound reaching each ear (interaural level difference, ILD). The ITDs are more effective for localizing low frequencies (lower than 1500 Hz), and the ILDs are more effective for localizing high frequencies (higher than 1500 Hz). Frequencies in the range of 2000–4000 Hz are poorly localized. In addition, spectral changes that occur when the sound encounters body parts, such as the torso, head, and pinnae, also provide information on the location of the sound source".
Excerpt from - Measuring “Where”: A Comparative Analysis of Methods Measuring Spatial Perception - Fostick and Fink
Edited to add one example of a "negative phase 2nd". I can't seem to quickly find an example of "positive phase 2nd" from my measurements.
Last edited:
Attached picture is the correct example of what is called ‘negative phase H2’. I’m not sure if effect on virtual source perception is created only by H2 affecting amplitude or by something else.
We may notice that ‘negative phase H2’ is suppressing rising edge of the fundamental while boosting the falling edge and flattening the top. With distortion being 0.1 – 1%, amplitude deformation is 1/100 to 1/1000 part.
Our sound perception is unreliable but seems that ears are the pretty incredible ‘instruments’.
We may notice that ‘negative phase H2’ is suppressing rising edge of the fundamental while boosting the falling edge and flattening the top. With distortion being 0.1 – 1%, amplitude deformation is 1/100 to 1/1000 part.
Our sound perception is unreliable but seems that ears are the pretty incredible ‘instruments’.
Anecdotally, my experience with lower noise has been similar. I have yet to understand how to measure the output impedance of a PSU nor try and do it, so I can't add anything to that piece of the puzzle.
My oversimplified take on it is since we take our spatial cues from differences in timing and amplitude of the waves, then having the lowest noise to allow for the greatest delta between "no placement / no sound" and "the music" would allow the best representation.
In my vernacular... with a very low noise PSU, the sound just seems to 'emerge'. It's odd... and I'm probably choosing the wrong word, but it's all I can muster.
I'll dig around to see if anyone out there has documented relative amplitudes necessary for spatial cues. I'd have to believe someone has done it.
Great thought starter! Thanks!
That can be measured. There is no need for uncertainty. If it's set up properly... it should be in perfect phase out of the driver.
It's certainly fun to try and figure some of it out over a coffee and some music. Greater minds than mine have tried. I just try to read up on as much of it as I can.
I'll leave the 'real' work to the engineers, the scientists, and the mystics. I'll enjoy the tunes.
Cheers!
IMO this is the crux. Soundstage is an illusion and not really measurable. Also it is close to impossible to know what was the original soundstage in the recording. Still it should be clear that a device with low distortions, low noise (including close-in phase noise), flat frequency response, good channel balance (phase and magnitude) etc. will have a better chance at reproducing the soundstage of the recording. Whether or not everybody likes it is another matter and as with anything subjective there is no correct opinion.
Agreed. Poor phrasing on my part. Apologies.
The information from the recording conveying what we'd perceive as the "soundstage" would be more accurate.
That information can be "added" by a recording engineer for effect, or the engineers' goals may be to convey accurately the acoustics of a "real" space.
That's the sticky wicket... when people say that a system "images well" or has a "great soundstage", they really have no true north toward which to point.
I simply read it to mean, "I like it".
It may save us all from the some unnecessary repeated debate 🙂, as a lot was said here:
https://www.diyaudio.com/community/threads/measuring-the-imaginary.409662/
bohrok2610 said:
Also it is close to impossible to know what was the original soundstage in the recording.
it is very easy, just listen...
Some have mix of close mic + stereo , some just stereo, some are just 'pan' forced separation,
You can watch many performances on youtube where you see microphones placement, everything.
Or better, purchase recordings of concerts that you attended. talk with the recording engineer, I learned very much that way.
Also it is close to impossible to know what was the original soundstage in the recording.
it is very easy, just listen...
Some have mix of close mic + stereo , some just stereo, some are just 'pan' forced separation,
You can watch many performances on youtube where you see microphones placement, everything.
Or better, purchase recordings of concerts that you attended. talk with the recording engineer, I learned very much that way.