Sound Quality Vs. Measurements

Status
Not open for further replies.
while I know our guru's prejudices go the other way it is a easily demonstrated that high global loop gain reduces distortion components at frequencies where the loop gain is high (duh) - high loop gain in the audio band reduces errors measured by the feedback network, diff input that have audio frequency components - even IMD difference products from ultrasonic signals – even if the test signal is injected into the output stage from the speaker terminals

making sure that the input diff stage is linear with the test signal, expected source - local degeneration/fet input diff pair, 200KHz low pass in front of a audio power amp seems like good ideas

DSD does need heoric filtering to remove the rising with frequency shaped noise - should exceed 5 th order, Sony had to reduce the original rec of 100 KHz fc dow to 50 KHz afeter some audiophile amps had problems with the high frequency noise content
 
Last edited:
Hi,

while I know our guru's prejudices go the other way it is a easily demonstrated that high global loop gain reduces distortion components at frequencies where the loop gain is high (duh)

Actually, I never disputed this "greater distortion reduction where NFB amounts are high". Further, it is so brutally obvious, anyone to whom this is not immediately obvious should retake EE201, at a minimum.

However, would you mind elucidating us what this has got to do with the price of Jasmine Tea in China?

It has been known for the last 6+ decades, that "THD" is meaningless.

So what is your point?

Ciao T
 
For life like realism, good buddy, you also need to have a front end capable of it, and a back end (speakers) capable of reproducing something like 114 dB SPL in peaks.

And the size and furnishing of your room are also highly relevant factors.

Personally, I just made it. I have a small room, some 14*12 feet, fairly full of everything. My speakers do 92 dB/2.83V/1 m, and my Karan amp is rated at 180/250W into 8/4 Ohms. 180W/8 Ohms works out to 22 dBW, plus a single speaker's efficiency of 92 dB equals 114 dB. Two speakers in tandem will of course increase the actual SPL, but the room will soak up some of it, so let's assume I'm even on that point.

My point is, I know from first hand experience excatly what you are saying. But I must now dissapoint you - in real life, peaks of 114 dB are EXTREMELY rare, and I do believe 99.9% of all audiophiles have system which don't even get near that, yet play music very well. Also, both the HK 680 integrated (rated at 85/130 W into 8/4 Ohms) and the Marantz 170 DC power amp (rated at 85/8 Ohms) are extremely convincing in ANY reproduction.

Ah, if it all boiled down just to power ...

Size D, not just SPL, Lifelike , comes not from just SPL, but also from size. Small amplifiers and speakers can play loud, but the sound is hi-fi ish and there is no growth...

Listen to a single FR driver at 95db and compare it to a large 4 way at 95db, big difference ..
 
Size D, not just SPL, Lifelike , comes not from just SPL, but also from size. Small amplifiers and speakers can play loud, but the sound is hi-fi ish and there is no growth...

Listen to a single FR driver at 95db and compare it to a large 4 way at 95db, big difference ..

It's a bit lame to say that it's not the size but what you do with it that matters, but it is true nonetheless. The problem with small speakers is mainly one of distortion at higher SPL. Driver technology is improving though, and in a moderately sized room, a 10 liter enclosure should get you a long way. No way in a 1-way though, 2-way minimum.
 
OK A, just where does one get a full range driver? There are a lot of drivers with wider than the usual decade range, but none I have ever seen is full range. Even large ES panels, just can't get down to where , oh let's pick a simple multi-way, 2CE or good old B&W 801's can.

TVR,

What I said was a single FR driver cannot sound like a big 3 or 4 way speaker, even when playing at the same SPL. The major difference being dynamic compression, which relates to image size and Percussive energy.
Hit a tom tom and you will understand how speakers have difficulty with the acoustic power necessary for realism.

It's a bit lame to say that it's not the size but what you do with it that matters, but it is true nonetheless. The problem with small speakers is mainly one of distortion at higher SPL. Driver technology is improving though, and in a moderately sized room, a 10 liter enclosure should get you a long way. No way in a 1-way though, 2-way minimum.

Small speakers lack percussive energy, of course small room , small speaker, there is no way around that, speaker to room ratio has to be optimized for proper reproduction ...
 
a,
We agree on the result, but not at all on the reason. If you are thinking some 4", xmax of .5mm driver like is popular in the single driver-low power club, of course it can't produce the same levels. ( A dozen in a line array can) My point was no single driver is anywhere close to full range with respect to frequency, even large ESL's that do have the capability to provide dynamics within their bandwidth, they can't reach into the real bass.

Now put in a very limited bandwidth signal, like a 1K test tone. The 4" and the three way will probably sound identical. Why? the mid range IS basically the same as the FR. If they don't sound the same, at least one of them is wrong.

Yes, there is a reason I use two 12" woofers.
 
Hi,

where did my post say anything about THD?

Okay, you did not say HD, however, the result invariably is low THD with extended high order harmonics.

All the various distortion mechanisms in Amplifiers tend to increase with rising frequency. Open loop gain tends to fall with rising frequency. So distortion rises with frequency often quite dramatically.

So it matters very little what happens where inherent (pre NFB) distortion is low and where loop gain is high, especially if you have sources with some ultrasonic signal content or noise (LP, Reel2Reel, SACD, Non-OS DAC's, > 48KHz sample rate PCM).

Using multiple distorting stages in series will create additional high order products not present in individual stages distortion. Using "local" methods (degeneration, cascoding etc.) will linearise individual stages so we do not get as much "distortion of distortion". Maximising the input stage linearity reduces "re-entrant distortion".

All this means relatively low levels of looped NFB with a very wide bandwidth. However if care is taken low enough HD can be attained this way.

Ciao T
 
Hi,

OK A, just where does one get a full range driver?

That depends on your definition of "full range driver". If you define Full Range as 20Hz-20KHz +/-1dB there is no such thing and there is also no such multi-way speaker, if we do not apply massive averaging.

However it is possible for example, to make a driver that can handle 40Hz - 16KHz in a 40 Liter vented enclosure, with a reasonably flat response to qualify for use as Studio Monitor in smaller rooms and at civilised levels.

In this case we are talking about the East German Z131/L40 Monitor which used the Schulze KSP-215 8" Fullrange Driver, rated at 12.5VA continuous power, introduced in 1966 and still in wide service in East Germany in 1989...

Ciao T
 
I remember bi-amped BEAG studeo monitors, size of refrigerator, they were gorgeous devices!

sh_020_745545.jpg
 
Hi,

I remember bi-amped BEAG studeo monitors, size of refrigerator, they were gorgeous devices!

sh_020_745545.jpg

These look like influenced by the Telefunken O85. Old Mono Systems with very wide dispersion. With Stereo transmissions they had to be replaced.

The L40 looked rather more pedestrian. If matched with an Amplifier (original EL84 push-Pull, later transistorised) that equalised the response (instead of the Passive EQ in the Box) it became the Z132 Monitor System...

L40_Front.jpg


There where many versions using this full range driver, from the 20 Liter sealed "Kugelboxen" K20 (spherical speakers of which some 60,000 where made and which hence still occasionally show up on e-bay) to the 80 Liter Reflex "L80".

More products from Heli in East Germany, both domestic and Studio may be found here:

www.HELIRADIO.de (Produkte)

Following the "Anschluss" in 1990 like many especially smaller companies in East Germany Heli was allowed the slow death of a thousand cuts, until it had to close.

About the only East German Studio/Pro Gear manufacturer that survived is justly famous MEG (Musikelektronik Geithain), oh yes, Vermona, maker of what I derived the "Zen-EQ"from also still cling on to life, in the day we used to call their gear "Zerrmona" (a wordgame substituting one of the german words for distortion into the company name)...

Ciao T
 
Thorsten, I must take issue with the idea that FM tuners have poor pilot tone (19 kHz) and subcarrier (38 kHz) rejection. That was a problem resolved even in the late 70ies.

Yes, some did have bad breakthrough, but we've moved a LONG way since then. But most had the problem more or less resolved. And I am not referring to my own reVox B760, rated at -70 dB out to 19...300 kHz (including all harmonics), that would be unreasonable given that its price made it very much a High End tuner in its day.

My Sony 3950 tuner did fairly well itself, rated at -50 dB 19 kHz and -70 dB 38 kHz. And that was no High End tuner, just areasonably good representative of its genre.

Ultimately, one could (and should!) question the logic of using a say DM 2,000 amp with a DM 300 tuner, which might have some problems.

The point being that I think MOST (but certainly not all) tuners would not be suspect in that arena.

SACD is another matter. Frankly, I know much less about it, I tried a few, even bought one (Denon 3910) of those "universal players", but the sofware came in greatly lacking, so eventually I gave it up as a moribund standard, a victim of MP3. A pity, I feel, but there it is. Anyway, again a sample or two could have problems with outgoing ultrasonic trash, which could upset the amp. But here again, one questions the logic of using a high quality amp with a cheap, run-off-the-mill Chinese made DVD/BD/SACD device.

That said, I still think reVox' logic is a good hint - make the amp capable of doing 300 kHz or better, and then install a first order filter at its input with a cutoff frequency of 200 kHz. It's simply a precaution well worth taking, that's all.

Brings into play the logic of a "tailored bandwidth", as not infrequently used in UK audio products. Quite simply, a high pass and a low pass second order filters, with cutoff frequencies of say 3 Hz and 80 kHz, usually executed with two op amps. Well, they used to do that, don't know how it goes these days.

Some of the DIY legends, such as Prof W. Marshall Leach's amps, used to have second order filters at their inputs, I forgot the cutoff frequency, but it was kinda low(ish). Just in case.
 
Hi,

Thorsten, I must take issue with the idea that FM tuners have poor pilot tone (19 kHz) and subcarrier (38 kHz) rejection. That was a problem resolved even in the late 70ies.

When I was younger the Pilot Tone drove me nuts (like ringing in your ears), despite the Tuner being fitted with a Pilot tone Filter. I ended up building a 19KHz T-Network filter that added another 40dB or so and even fitted it to my Bathroom portable (mono) radio... Maybe I'm more sensitive than most.

The Revox should do okay.

Ciao T
 
Wayne,

You are wasting your time knocking on an open wide door, arguing that small speakers lack the wherewithall for a lifelike sound, at least in my case.

Historically, I got over that problem in 1972, when I exchanged a pair of small Uher speakers, which my dad purchased with their Royal de Luxe tape recorder (with built-in amps 2x10Wrms), for a pair of AR5 speakers. I'd have to be stone deaf not to notice the difference.

Ever since then, I have owned nothing but larger speakers, and I never once looked back.

As a general rule, small speakers image better than most big speakers, they tend to portray a wider sound stage, but dismally fall on drive impetus and sheer size presentation. Balls of glass, you might say.

I prefer balls of brass, even if I do pay for it for a little less imaging. Remember, I used to bang on the drums in my time, and anything that fails to do drums properly hasn't got a chance in hell with me. This assumes a bass driver no smaller than 10 inches (26 cm), and it has to be a 3 way speaker, because by the time one graduates to 3 way, one is able to optimize each driver for its near optimum reproduction range (assuming a good initial choice of drivers).

But let also say this - if you want to play the power game, then you have in all probability lost it before you even started unless you opt for well made (not just any, ol' ...) pair of active speakers. Consider - active speakers do not use passive crossovers, which rob you of a lot of input power. If their own built-in amps are well judged, the sheer feeling of a bottomless pit of power will be there, that subjective feeling that it has limitless power on tap - a trait I treasure greatly in all instances.

And, in addition, you can actually optimize your power amps for a specific driver. You do NOT need, under any cirumstances, more that say 50 Watts for a midrange driver, unless it is uncommonly inefficient, and your tweeter doesn't need more than say 30 Wrms - but these could well be pure class A watts, and NOTHING beats that.

Trust me on this, I have heard active sšpeakers from Philips, Grundig, Klen & Hummel (we had them in the studio of the TV editing room), ATC and several other less well known makers. And in all cases, the sheer dynamic range of them left like 90% of classic passive speakers standing still, even when driven by real high class amplification, which invariably ran your bill way higher than when using active right from bat.

Small monitors, even active ones, make sense only in ENG vehicles, where space is at a premium. Hence models like the BBC LS monitors, all the rage in the UK in the 70ies. Nice speakers, but they managed to lose bass altogether, they simply had none. That might be acceptable in an ENG vehicle, but certainly not im my room. I want and need the slam of it, and I'm not going to get any real slam with a 6.5 inch "bass/mid" driver.

And it generally works the other way as well - sticking in a 12 inch (30 cm) bass driver into as small a box as you can make it is not going to give that driver much of a chance. The only excpetion to this rule that I know of are some (not all) models made by a German company called Canton, such as model 530 and so forth - they had 12 inch drivers in smallish boxes which still somehow managed very decent bass (these were acoustic suspension models). They don't go below around 50 Hz, but what they do, they do well.
 
Hi,

When I was younger the Pilot Tone drove me nuts (like ringing in your ears), despite the Tuner being fitted with a Pilot tone Filter. I ended up building a 19KHz T-Network filter that added another 40dB or so and even fitted it to my Bathroom portable (mono) radio... Maybe I'm more sensitive than most.

The Revox should do okay.

Ciao T

Yeah, I know, that chirping sound, like you had birds inside the tuner. I've heard it on some occasions, and I believe you when you say it drove you nuts, it would drive me nuts too if I had to listen to it all day long.

My own situation was even more complicated. Being a city slicker, I also needed tuners which killed the pilot tone and subcarrier dead, but also did AM suppression well. Not easy to find, and not cheap either (at least in those days).

But again, the B760 does this uncommonly well, rated at -70 dB.

Actually, the REAL miracle here is that I still have 3 or 4 even FM stations which do NOT use PC driven MP3 for their programme, rather they stick to the true 16-bit CD format and pay attention to their sound quality. Not infrequently, you don't know whether you're listening to the tuner or the CD player, which is saying something.

But I suspect high compression DAB will end all that soon enough.
 
Status
Not open for further replies.