Sound Quality Vs. Measurements

Status
Not open for further replies.
Whether Frank got the gist of your posts is another matter, but Frank is saying something I deeply agree with.

A truly good amplifier, never mind how powerful or not, what it's made of, etc, needs to maintain the same tonal balance from top to bottom. Actullay, from bottom to top is better, more of them will sound fine at lower levels, but not half as many will manage to keep composed once you start cranking them up.

If its overall sound balance comes unstuck at higher levels, then it will lie to you with each and every transient that comes along.

While this requirement seems logical and almost natural, in real life it is very rarely met.

Because both my Marantz 170 DC and Citation 24 power amps DO manage to do this right is the key reason why I insist on them. Accuphase E-206 integrated, for example, did not manage to pullthis off, despite its hefty price. But the Karan integrated did.

And remember, I have exceptionally well behaved speakers, very easy to drive. Whne I notice changes even on them, then that amp has serious problems in my book. Taking it to another room and using AR 94 speakers, this effect is usually made much more prominent, as they are not very forgiving.

In my experience, this has little to do with power output, although truth be told, more powerful and more expensive amps tend to suffer less than the samll guys. Presumably more work was invested into them to justify the higher price.

Surprises are possible. Some time ago, I had an old Toshiba SB-620 integrated amp with me, datng back to 1977 or so, fully refreshed. It has a SEPP output and is rated at 62/75W into 8/4 Ohms, yet it managed to sound way bigger and more substantial than its specs indicate. As does my 5 or 6 years younger Toshiba SB-640, which I own only to prove a point - you do not need megawatts for quality sound, and small and cheap sometimes can surprise very pleasantly.
 
I believe, right or wrong, that this has everything to do with internal overload margins of an amp. Two xamples.

The VAS currenty Demian and I were on about not too long ago, which Wayne spoofed and suggested that we drive the speakers with the VAS. Let it ride on good current, say 10-15 mA per trannies, and it will become very linear and if some local degeneration is present, very low distortion. The chances of it being overloaded, or aksed to deliver more current than they actually posses, are greatly reduced, although of course much also depends on what follows and how.

Or take the usual FET input differential stage. FETs are gluttons for current, it's as if they will eat up just about anything you throw at them. Not quite so, of course, but generally I find they are often underbiased, and are treated almost like they were standard BJTs. The resut is some loss of clarity and focus, with somewhat increased distortion figures (all depending on how underbiased they are). This is NOT to say that you can dump currents into them without reserve, but try hard enough and you will find a bias point at which they will surprise you, and it will almost by default be above what they were set at in the factory.

Recently, Demian suggested that I should crank up the bias of my input FET stage. It was originally set at around 1.5 mA per trannie, typical BJT thinking on my part. I admit I was a bit surprised at his suggestion, but I did try, and boy oh boy, was I suprised, or what! When I reached a point of 4.13 mA, things started to happen, distortion dropped suprprisingly much, for example my 20 kHz THD dropped from 0.046% to below 0.02 %, it literally halved. This may sound academical, but when a small change halves your distortionm, you can bet your blessed cotton socks that other things begin to change as well. To name but one, my voltage slew rate went up from approximately 100 to over 200 V/uS.

I don't think a living soul on this world knows exactly how a higher slew rate impacts the sound, but as both John and Demian said, each in his way, somehow high speed amps tend to sound better than low speed ones. And it can be done fairly easily, but you have to purpsefully go for it. It can't hurt, that's for sure.
 
article refs? reseachers names? - just where is this loop?

a very unique claim given the realities of biological systems information usually being transferred by diffusion of neurotransmitters and most neurons saturating at 4-8k pulses per second

many mis/overinterpret the interaural time delay resolution of few us to imply 100s of kHz "hearing" - but the test signals are strictly band limited with no > 20kHz components

the precise mechanism may still be a subject of research - but from a signal processing perspective a correlation process doesn't require any such supersonic hearing

using correlation you can trade S/N, observation time for raw time resolution in precisely measuring time delays of similar signals

You won't get much because these guys get threatened by animal rights . It got ugly . The most interesting research was on hamsters . They simulate us better . Some of what I was told will not be published as it may have been on themselves and not easy to do the control tests . Here is the last remnant I can find . When I got the parts they never said animals . They hinted mice were useless analogues .

These guys had nothing to prove in telling me . They just wanted me to understand the picture is confused by so many assumptions . Measurements are fine if you know where it is going . Experiment is the fast short cut .

http://www.cell.com/cms/attachment/2007952222/2030510559/mmc1.pdf
 
Disabled Account
Joined 2012
------- somehow high speed amps tend to sound better than low speed ones. And it can be done fairly easily, but you have to purpsefully go for it. It can't hurt, that's for sure.


If you/others want details on this -- see the DIYAudio forum on CFA design and what it is, And then the forum called SlewMaster CFA vs VFA and the designs there as well as from dadod. There you will find a lot of info and real designs -- SIM and built -- on this subject which you are so surprised to learn about such things from JC and DM.

THx-RNMarsh
 
Last edited:
EX7tKoU.jpg


At last I have some data that might give joy to many ( pocket wise ) . The 5 inch Scan bass units are now replaced by the Eminence Beta 12 Lta on an open baffle ( OB ) . If one ignores the 10 dB output difference they are not so different . Eminence graph is very honest . Before you judge look to see it is 2 dB variation typically . The old Quad ESL 57 would confound the measuring equipment as badly as this . Qts is similar as is the low bass end .

The advantages of the 12 Lta is it can stand the loudness ( 50 watts ) even though using + 16 dB at 30 Hz EQ .No bottoming which can not be said for the 5 inch unit . Other than that 70 % over lap in the good and the bad .


The 12 Lta's are a riot . I think like some girls I have loved , I suspect it will be a summer romance . First thing to say is in many ways these are the least compromised OB speakers I have heard . Not as good as Quad ESL 63 and the new re-branding version . In my opinion 63's need to have the cloth removed . This is risky . If you ever had a belt off of 63's you won't want to do it again . It really hurts , that is if you survive .

There is plenty to criticize in these speakers . Like the Quad ESL 57 and my Magneplanars the off axis is a joke . The Scan units are wonderful and better than anything I ever heard off axis . The graphs show that . 30 degrees is a linear response to 10 kHz . I am surprised the Scan works so well as the baffle that is 2 feet wide . The one thing to say strongly is unlike Harbeth speakers that have an excellent graph these OB to me suggest real music . I heard some big Harbeth I liked so not saying all Harbeth .

The 12 Lta is almost a full range unit . The very best news is that very standard tone controls work well on this baffle . Makes me look as if I knew what I was doing , it was pure luck . + 12 dB 100 Hz and + 4 dB 10 kHz seems fine . Unlike box speakers tone control not only work they work nicely . The old Kenwood ( Trio ) amplifier my dad owns would be interesting . It is a great sounding amp with the tone controls wrapped around the power amps ( not inline with it ) .

The 78's I like are 70 % as good as I want them . Ironically the ribbon tweeter on order might solve that . A homeopathy idea if ever I saw one . What I hear if asking is the room the people are in . It must be that the signal is fast but can not be resolved as harmonics . In other words the stamper had the information and the granular shellac couldn't give the detail .

I work towards a " play everything " solution these days . My Magneplanars will do that . The 5 inch jobs mostly can and mostly more real sounding than the Maggie's . Quad ELS 57 and 63's equally flawed .

To point out something important . I am using 2 inch x 2 foot x 4 foot expanded polystyrene blocks usually used as insulation . It is nearly impossible to use as an anchor . The theory being minimum energy storage . Bass is excellent and I supect the reduction in wood sound is useful . I felt this might be important if I was to take on the mighty speakers .

Two outstanding moments .

Film , The Talented Mr Ripley . No 1950's sound. No boom . Voices very natural . Music also excellent . Never fatiguing . Space and reality above average .

Captain Sensible , Happy Talk . It came up by accident . I was mesmerized . Better than any band stuff I ever heard . The sound is everywhere in the house and contrary to the usual sounds no less real with distance .

I pressed into service some Motorola piezo clones . They start at 10 kHz and are not loud enough to meet the 12 Lta . All the same they are useful for now .

The drive unit is central . This means if your life style allows you can build a speaker to rival many for very little money . If you get fed up with them eBay people will want them . A fatal flaw it is how quickly they die off axis . However they are far more intelligible when dialog than the Maggie's when off axis . They are a bit coloured which I will judge when they have the 50 kHz added . Remember if you don't use polystyrene you might not get such a good result . They don't like 1950's stuff much as they fail in the same ways as the material . They don't hurt my ears as lets say real opera does . I suspect the tweeter will help that . They do not have the micro detail of the Scan , nothing does .

I have to say I take my hat off to Eminence . They are so much better than they should be . I couldn't brake them try as I might .

The Scan's where not for this project . As they are smaller I thought why not as the baffle was not damaged by doing that . I think I will return to them and use my 15 inch bass to stop them bottoming .
 
If you/others want details on this -- see the DIYAudio forum on CFA design and what it is, And then the forum called SlewMaster CFA vs VFA and the designs there as well as from dadod. There you will find a lot of info and real designs -- SIM and built -- on this subject which you are so surprised to learn about such things from JC and DM.

THx-RNMarsh

I will take a look, thank you, Richard.

As for fast amps, I have had that personal impression ever since the second half of the 70ies, when fast amps started cropping up all over the place, from Sansui, Kenwood, Pioneer, etc.

Too bad they were so hell bent on slew rates and the rest fainted if it encountered a speaker with an impedance of 6 Ohms or less. I drained the life out of most of them with my AR 5 speakers I had at the time, and remember, AR 5 was the good brother of the family, with a nominal impedance of 8 Ohms, but with a 10" bass unit rather than 12" as in AR 3A Improved, 10, and 11.

Also, at the time, I still had my home brewed version of the Otala & Lohstroh amp and enjoyed it. One of the stupidest things I ever did in my life was to sell it off some years later.
 
Last edited:
Sorry Wayne I missed the ref, no idea where this refers to.

Jan

No problem Jan.... :)



@Dvv,

You may or may not disagree with Frank , not the issue, the issue is when discussing paint and he flies in discussing rollers... :rolleyes:


As to amplfier and system spectral balance from high to low , thats academic and necessary for any system to sound correct on quite and loud passages, unfortunately thats not what i/we were discussing , what Frank does is cloud with his useless PC system reference...


:2c:
 
If you/others want details on this -- see the DIYAudio forum on CFA design and what it is, And then the forum called SlewMaster CFA vs VFA and the designs there as well as from dadod. There you will find a lot of info and real designs -- SIM and built -- on this subject which you are so surprised to learn about such things from JC and DM.

THx-RNMarsh

Looking into trying slewmaster CFA ......
 
On the subject of CFA vs. VFA, I do not believe either type is better by default. Both have their pros and cons, my belief is that other aspects, present in both types, are far more important that the type of feedback.

For example, a while ago I audiotioned some of the latest crop from Marantz, and frankly, they sounded like El Cheapo devices desipite their much touted CFA. I am, of course, not saying that CFA is junk by deafult, just that in my view Marantz failed to get it right.

Like many others failed to get VFA right.
 
You won't get much because these guys get threatened by animal rights . It got ugly . The most interesting research was on hamsters . They simulate us better . Some of what I was told will not be published as it may have been on themselves and not easy to do the control tests . Here is the last remnant I can find . When I got the parts they never said animals . They hinted mice were useless analogues .

These guys had nothing to prove in telling me . They just wanted me to understand the picture is confused by so many assumptions . Measurements are fine if you know where it is going . Experiment is the fast short cut .

http://www.cell.com/cms/attachment/2007952222/2030510559/mmc1.pdf

thanx - that does look like they are working in the general field of iteraural time difference resolution
the article doesn't have the info - but it wouldn't surprise if ferrets with >2x human upper frequency limit and ~10x smaller ear-to-ear distance can resolve to below 1 us time shift in signals presented to both ears

but turning that 1/t number into MHz really is misleading

the IATD numbers as I mentioned don't say anything about hearing bandwidth - or even the neural detector system's speed - because neural circuits are often massively parallel and can do correlation/pattern matching even with the slow bottlenecks of individual neuron few kHz max firing rate and delays from transmitter chemical diffusion between synapse


for humans the IATD numbers in textbooks is 5-10 us, with a few reports of ~2us out there - when listening to clik tones in headphones

really doesn't say much of anything interesting for even moderately competent audio electronics - matching R,L channel delays/phase to that level is standard consequence of flat frequency response
except possibly the low frequency time constant with AC coupling or feedback DC blocking - and there speakers and room modal response dominates by several orders of magnitude
 
BTW, what exactly are we discussing just now?

Just wonderin' ...
Very specifically, dvv, what differentiates "expensive" gear from "cheap" gear, if they're running within their limitations. Let's pull out the BT methodology: put a sheet up at the end of the room, have one of a.wayne's drool-worthy setups there behind it, as well as one my tweaked el cheapos, match levels, get rid of rumble box ... will a.wayne be able to pick out his 'baby' ...?
 
For example, a while ago I audiotioned some of the latest crop from Marantz, and frankly, they sounded like El Cheapo devices desipite their much touted CFA. I am, of course, not saying that CFA is junk by deafult, just that in my view Marantz failed to get it right.
Interestingly, same experience at the Sydney audio show. They had Mr Marantz there, Ken, looking rather sad and forlorn, hovering over a Marantz unit pumping out slightly better than boombox sound - he looked as if that was the last place he wanted to be ...
 
I think at least as interesting would be to have YOU there, too, Frank


They could have an end each - behind acoustically transparant curtains. I'm sure that Frank would not be able tell which system was playing; his 'puter speakers + whatever he drives them with, or Wayne's exotica! [To start the comparisons an audiologist would need to determine the aural state of the contestants; the results may well expose some anomalies - which would explain a few things which have been posted! .;)]
 
Status
Not open for further replies.