Sound Quality Vs. Measurements - Page 68 - diyAudio
Go Back   Home > Forums > Member Areas > The Lounge

The Lounge A place to talk about almost anything but politics and religion.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 9th January 2012, 11:18 AM   #671
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
A 10 cycle pulse of 4kHz, starting from zero, has a lot of transient high frequency components - not just 4kHz. Reproduction will depend to some extent on the HF rolloff of the amp, so it is possible that this is some or most of what you are seeing. If this correlates with perceived audio quality on music then it could be interesting.

Also, HF rolloff may have two components: voltage gain, and output impedance. To what extent is your tweeter sensitive to driving impedance?
  Reply With Quote
Old 9th January 2012, 11:26 AM   #672
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by Wavebourn View Post
There were tons of proposals similar to what DF96 proposed, but no agreement about what and how to weight. People want something simple, like 2x2=4, so any marketing department guru can understand, any lawyer agrees.
As mentioned in my previous post, I think trying to find a weighting to provide a single figure result, even if we knew which measurements to weigh, is futile.
Quote:
I've found one measurement: if subconscious reaction on sounds happen before people realize that it's reproduction, it is good quality. My usual test: close eyes and try to imagine that the sound is real. The easier it is to imagine, the better is quality.
Bingo. That's what I do too

Let's call it "willing suspension of disbelief". Your rational mind knows the sound is not real, your eyes confirm its not (clearly you're in a room with speakers, not somewhere else) and yet when you close your eyes, you have to work hard to convince yourself that it's not real. That's when you know you're getting close.

Sometimes the set-up that produces this result measures "well" with conventional measurements, sometimes it doesn't. This alone suggests that we still don't know everything that matters in the measurements, or we misunderstand something about our hearing perceptual mechanisms.
Quote:
However, such measurements require some calibration. Untrained people eager to judge experienced designers can't use them.
I think its one of those things that can happen as a Eureka moment. Not many speakers will convince you of the realness of the sound, until you have heard some that do, you will always assign "speaker sound" to speakers, and have no frame of reference to what is possible beyond that.

As soon as you hear a system that is convincingly real with your eyes closed you have a "Eureka!" moment where you see what is possible, what you previously thought was impossible, and you then spend the next xx years trying to understand how it happened, and define (unsuccessfully) which measurements led to that result...

Once heard its hard to un-hear it as well, which leads to being disappointed with anything less than perfection, a problem no doubt many of us face when we're now disappointed in the sound of a system that many years ago we might have thought was pretty good, simply because our expectations have been raised so high.
__________________
- Simon

Last edited by DBMandrake; 9th January 2012 at 11:33 AM.
  Reply With Quote
Old 9th January 2012, 12:41 PM   #673
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Quote:
Originally Posted by DBMandrake
I think trying to find a weighting to provide a single figure result, even if we knew which measurements to weigh, is futile.
Who is after a single figure? My guess is that as a minimum you would need four or five figures just to cover the basics of what we know now: harmonics, IM (two measures of non-linearity, perhaps with both as function of frequency), LF/HF rolloff (or, equivalently, bandwidth and LF/HF balance), noise. Let's assume we can choose a set of figures which capture what matters - we would still find that different people attach different importance to each of them, partly through taste and partly through environment. That is not a reason not to look for the figures, though.

People like to knock THD etc. but they need to remember that originally nobody knew what mattered, so careful experiments were conducted on distortion, frequency response etc. We need better experiments, not attempts at 'proof by assertion' from either side of the debate.
  Reply With Quote
Old 9th January 2012, 01:21 PM   #674
diyAudio Member
 
DBMandrake's Avatar
 
Join Date: Nov 2010
Location: Glasgow, UK
Quote:
Originally Posted by DF96 View Post
Who is after a single figure?
Pano, with his "Sound Quality Index" ?

Toole, with his speaker ranking studies ?

Note: not a single measurement, but a single aggregate figure of merit derived from a group of measurements. That's what I'm saying is futile, because I think you simply can't boil down so many competing and largely orthogonal factors into a simple scalar figure of merit, especially if personal preference is indeed a factor.

And there's not much reason to, unless your goal is to replace controlled listening tests with measurements to see how speakers rank against each other in listener preference. A single boiled down figure of merit gives no guidance into what needs improving in each speaker, so doesn't help the designer.

Quote:
My guess is that as a minimum you would need four or five figures just to cover the basics of what we know now: harmonics, IM (two measures of non-linearity, perhaps with both as function of frequency), LF/HF rolloff (or, equivalently, bandwidth and LF/HF balance), noise. Let's assume we can choose a set of figures which capture what matters - we would still find that different people attach different importance to each of them, partly through taste and partly through environment. That is not a reason not to look for the figures, though.
I'm 100% with you there, you're arguing the same side of the argument as I did in my longer post...? I want to find those things that most matter too, however I don't see any point in a single figure of merit for ranking purposes. Keep the handful of important measurements (whatever they are) separate.
Quote:
People like to knock THD etc. but they need to remember that originally nobody knew what mattered, so careful experiments were conducted on distortion, frequency response etc. We need better experiments, not attempts at 'proof by assertion' from either side of the debate.
People knock THD because it doesn't correspond well, if at all, with what we hear. Yes gross distortion like clipping or a voice coil going out of its gap is audible but below gross levels its very unclear what's going on.

I used to believe strongly in trying to get vanishingly low levels of distortion at mid/high frequencies in particular, but in recent times I've been having a "crisis of faith" in this area where experimental measurements and listening tests keep pointing in the direction that once below a modest threshold, the audible difference between low and ultra low distortion on music is nill. (At least for low orders)

Of course Earl Geddes has been beating this drum for years, somewhat on his own, and I have reluctantly come to the conclusion that he may just have a point after all, although I perhaps wouldn't go quite as far as he does.

Rather than trying to achieve ultra-low distortion at modest SPL it seems to be more important to maintain modest distortion levels to very high SPL levels - eg a large well specced driver sounds good not because distortion may be ultra low at modest levels, but because its distortion is still only moderate at much higher than necessary levels.

THD also doesn't take into account what order products are present or in what proportion, and its clear that this relationship does hugely affect how audible distortion is. For this reason alone THD is meaningless. At the very least we need to talk in terms of how far down each individual distortion product is, preferably in dB rather than percent.

Distortion matters, but probably not as much as we think and not in the ways we've traditionally thought about it, and THD as a figure of merit is pretty much meaningless due to the lack of information contained within it.
__________________
- Simon

Last edited by DBMandrake; 9th January 2012 at 01:23 PM.
  Reply With Quote
Old 9th January 2012, 01:36 PM   #675
diyAudio Member
 
Join Date: Dec 2011
Quote:
Originally Posted by DBMandrake View Post
I
As an analogy, how would you go about defining a "Picture Quality Index" for the picture on Televisions ? Unlike hearing, what we can see is a lot better defined and understood, and a lot less subjective than hearing.

Convergence errors, geometric distortions (barrel distortion etc) grey scale tracking, black level, banding on gradients, colour balance, (highlights and lowlights) update rate/ghosting, resolution and sharpness, motion compensation processing, noise, halo effects, the list is nearly endless. Some screen technologies suffer from some problems, some from others, just like certain speaker topologies tend to suffer more from certain problems.

All these things can be measured fairly easily with the right tools, most can be seen individually with the naked eye with the right test patterns, they are all well understood, but could you measure them all numerically (without any visual inspection) and calculate a single quality index figure that would agree with the subjective ranking of TV sets by individuals ? Maybe, maybe not, but should we even be trying ?
Hi Simon,
Actually I was the Technical Lead and electron optic designer for the 32V HDTV real flat CRT for Philips in 1998-2000. We did have two quality metrics QK(color) and Qr(rastor) In simulations ( a 3D Greens function of the electron magnetic field and its interaction with the RGB electron streams) we spent months balancing Qk against Qr we then moved these design to prototypes for another 4 months of development . We had minimum accptable values for both Q's and in the end without simulation's and metrics the dipole through 18 pole analysis would have been impossible.
Yes othere are other metrics as you mentioned, BTW we also had a non quantifiable element, we called it "sparkle" similar to "live" in audio , but I see nothing wrong with a couple of q's e.g
Q speaker & source load ( ability to handle reactive loads without delay or distortion and handle phono->Digital loads)
Ql(live) ability to amplfy and "maintain" or "enhance" the live experience, QLoud and Q finesse ( perfect RIAA, dynamic Loudness)
and finally most important QWAF (wife/partner acceptance factor)

Lou
  Reply With Quote
Old 9th January 2012, 01:49 PM   #676
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Folks,

One issue is that even with weighting applied and all we can mostly only make negative determinations.

For example we can make an aggregate distortion metric that allows us to class the effect of the distortion as "no effect", "small effect, not objectionable", "small effect objectionable" and so on onto major effects.

So we can determine when distortion will influence what is heard and when it will become objectionable.

Note, this is not a positive determination as in "number so and so on this index guarantees good sound".

As much of the research (naturally) focuses on determining what kind of degradation is tolerable (perceptual coding research is a goldmine of information on that topic) it does not actively pursue anything that is deliberately positive.

I know for example that using very minor manipulations of frequency response I can make a system sound ridiculously spacious, but distant, equally I can get an effect that creates more "presence" and many others, by smal frequency response manipulations.

I do know that a lot of 2nd HD (usually more than I allow my commercial circuits) can also enhance the sense of spaciousness and of detail (likely the "lost fundamental" effect comes to play).

So, if I can take some modest liberties (within the boundaries of what is not perceived as objectionable) with traditional measurements I can create a sound that is immensely appealing and enjoyable. The "UltraFi"crowd (read Horn Speakers, SE Amp's, Vinyl etc.) is majorly on this kick. Much of their gear measures revoltingly bad by traditional standards, yet it makes musik come a-live like little else in the mainstream of HiFi and High End can...

So, where does this leave us?

Is the extant approach to characterising audio equipment by certain measurements even defensible or sensible? Or is it just Captain Slater (Witchlocator) continuing to mumble "But we have this here ducking stool...".

Ciao T
  Reply With Quote
Old 9th January 2012, 01:56 PM   #677
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Comparison with TV also brings in another issue: given a free choice, some people opt for distortion. For example, many people set up a TV with colours which are brighter (e.g. more saturated) than reality. They then get used to this, so a correctly set up TV looks pale and bland to them.
  Reply With Quote
Old 9th January 2012, 02:47 PM   #678
M Gregg is offline M Gregg  United Kingdom
diyAudio Member
 
M Gregg's Avatar
 
Join Date: Jun 2010
Location: UK
Quote:
Originally Posted by DF96 View Post
Comparison with TV also brings in another issue: given a free choice, some people opt for distortion. For example, many people set up a TV with colours which are brighter (e.g. more saturated) than reality. They then get used to this, so a correctly set up TV looks pale and bland to them.
Here is a thought..

Can a TV produce a picture that cannot be distinguished from reality...

Can a HIFI produce a sound stage that cannot be.....

What are the limiting factors....

The limiting factors are distortion....<<<<not true to reality

How do you measure the difference....forget everything else...

If it can be measured it can be controlled?

Regards
M. Gregg
__________________
What is the sound of one hand clapping?

Last edited by M Gregg; 9th January 2012 at 02:51 PM.
  Reply With Quote
Old 9th January 2012, 03:01 PM   #679
M Gregg is offline M Gregg  United Kingdom
diyAudio Member
 
M Gregg's Avatar
 
Join Date: Jun 2010
Location: UK
here is another thought..

If someone plays a trumpet in a room it will sound like a trumpet..

If they play in a different room it will sound like a trumpet

Take a HIFI to a different room it sounds like a HIFI..is the soundstage the same?

What makes a trumpet sound like a trumpet in any location?

Regards
M. Gregg
__________________
What is the sound of one hand clapping?
  Reply With Quote
Old 9th January 2012, 03:02 PM   #680
DF96 is offline DF96  England
diyAudio Member
 
Join Date: May 2007
Metrics get more complicated when the possibility of making something 'better' than reality is included, whether brighter-than-life TV pictures or 'appealing and enjoyable' sound. If 0 is perfection, 1 is barely discernible difference, 2 is tolerable, and 3 is horrid some people may prefer 1.2 - especially if they rarely hear live music. For example, some brass instruments can make quite an unpleasant noise when played loudly - hear it at home and you suspect a tweeter fault, but in the concert hall it is reality.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Quality Control differences = variations in sound quality? KT Class D 3 4th June 2014 12:02 AM
Sound Card for Measurements Marik Solid State 2 2nd January 2012 08:59 PM
Sound Card Recommendations (For Audio Measurements) dchisholm Equipment & Tools 5 16th July 2011 09:40 AM
How to protect sound card during amp measurements? okapi Everything Else 13 2nd September 2008 03:06 PM
Sound cards - test and measurements jackinnj Everything Else 2 5th July 2003 03:02 PM


New To Site? Need Help?

All times are GMT. The time now is 03:46 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2