[eico1]

Well, if you have two stages, each producing Only 2nd and 3rd harmonics, the feedback signal also contains these harmonics. Now, when combined with the input signal, at the first stage input, one gets 2nd, 3rd, 4th, 6th, and 9th order harmonics thru the 1st stage. Of course, thru the 2nd stage is even worse.

If the level of the second harmonic is say 40 db down, and you are saying that negative feedback will then generate a 2nd harmonic from that again 40 dB lower, the 4th harmonic will be 80dB down and higher harmonics even lower. That wouldn't seems to me to be very audible. Maybe if negative feedback was in itself bad there is some other mechanism?

steve

[jules]

I have to differ with much of what is being said:

The volume of sound produced by an orchestra [and I use this as an example only because it makes a point] can multiply from 1 unit to more than 100,000 of the same unit [whatever you happen to choose]. An amp or a speaker is simply not capable of doing the same therefore on this ground alone something has to go missing. This missing information keeps being ignored here ... why?

It's also true that if you put an orchestra in a room the size of a large living room [and I've practiced in similar circumstances] what you get is a total cacophany of sound that is very unmusical. While it's impossible to measure [again due to the overwhelming amount of sound] I would say that distortion in the form of interefering sound waves is rife. So the point is, if you were truly able to reproduce the full output of an orchestra in your listening space the result would be the same ... chaos!

My own speakers lean slightly to the detailed end of the spectrum but the point here is that it's easier to reproduce a greater amount of information in the detailed zone than it is in the warm zone. A low Qts speaker can fail in sounding too light but a high Qts figure can cause sogginess. The sound of a bass drum contains a great deal of what I would call warmth but it is always going to more demanding to reproduce than the sound of a flute. [at least in terms of power and the amount of "space" available in the total sound output of a speaker].

It's unfortunate that there are two camps each denying the accuracy of the other. I'd like to see a system something that could achieve the qualities of a ribbon AND a silk dome with a low Qts mid AND a high Qts mid. Yes, I realize the difficulties but we seem to be very linear in our thinking at the moment.

jules

[timbley]

I have to differ with much of what is being said:

The volume of sound produced by an orchestra [and I use this as an example only because it makes a point] can multiply from 1 unit to more than 100,000 of the same unit [whatever you happen to choose]. An amp or a speaker is simply not capable of doing the same therefore on this ground alone something has to go missing. This missing information keeps being ignored here ... why?
...

My understanding is that 60db is a factor of about 1,000,000. Is that correct? What can an amp and speaker do?

[jules]

with all respect I think you have missed the point of what I was saying ... It is simply impossible to have the entire and complete sound output [or anything near it]of an orchestra in a space the size of a large living room!

The volume of soft passages involving half a dozen instruments would seem very loud and the volume of a loud passage becomes a total soup.

You can't simply produce a mini version by use of a volume knob ... It's a different thing entirely.

jules

[Steve]

Timbley, it should help with the nfb amp to use it at lower frequencies. Whether it is noticable will depend on the sonic quality of the amp. If pretty good, maybe not much.

The question also arises about the active crossover vs a passive. Alot will depend on the sonic qualities of the active crossover, such as parts quality, design etc. It is possible that you could have problems like hiss, esp if the amp is an integrated amp, with great sensitivity. Just like a mid priced or midfi preamp, an inexpensive active crossover may cause more harm than good.

-------------------

If the level of the second harmonic is say 40 db down, and you are saying that negative feedback will then generate a 2nd harmonic from that again 40 dB lower, the 4th harmonic will be 80dB down and higher harmonics even lower. That wouldn't seems to me to be very audible. Maybe if negative feedback was in itself bad there is some other mechanism?

True, but I think you are optomistic with the specs (plus there is another aspect below) as the best output tubes are usually only around 30db down for 2nd, and maybe 45db down for 3rd. We also have to account for all the stages, including the OPT in a tube amp.
Now for the bad news.

IM distortion (IMD) will generally be 3.2 times (maybe 10db worse specs) the HD value, % wise. That is really bad. IMD produces sums And differences between each and every harmonic produced as well as each harmonic and the primary signal.

A special case, if we have a 2 stage amp with 1khz fundamental, and 2nd, 3rd harmonics, we get sums and differences between 1khz, 2khz, and 3khz. For example:

2khz -1khz = 1khz, and 2khz + 1khz = 3khz. 3khz - 1khz = 2khz. 3khz + 1khz = 4khz. 3khz - 2khz = 1khz. 3khz + 2khz = 5khz. Etc. This example is a special case.
(Add a third stage and the combination of sums and differences gets worse.)

Notice, in this special case, the results come off as harmonics (1,2,3,4,5th order), but with millsecond phase and time differences. The wider the bandwidth of each stage, the less time and phase shifting that occurs.

But what if we have just two primary signals (could also be two harmonics), one at 800hz and one at 1khz, a different picture is painted.

1khz - 800 = 200hz. 1khz + 800 = 1800. Neither result is a harmonic of the two primary signals. It is garbage.  One can only magine, with an orchestra and many instruments, how many totally unrelated signals are produced from all the fundamental and harmonic tones the instruments produce.

---------------

>>"The volume of sound produced by an orchestra [and I use this as an example only because it makes a point] can multiply from 1 unit to more than 100,000 of the same unit [whatever you happen to choose]. An amp or a speaker is simply not capable of doing the same therefore on this ground alone something has to go missing. This missing information keeps being ignored here ... why?">

That is why progress has to be made in developing new components, and not stagnate. A live listener also misses info depending on the distance from the orchestra. So it probably isn't as bad as could be, using the best components.

It is possible to have an entire orchestra in a room and still distinguish the different instrumets. Granded, go "outside the walls" with the sound or get a larger room and it does get better.

Late now, so I better stop.

 

steve

[timbley]

with all respect I think you have missed the point of what I was saying ... It is simply impossible to have the entire and complete sound output [or anything near it]of an orchestra in a space the size of a large living room!


jules

I do see your point, and it's a good one. Really, we have to accept the fact that we're trying to do a simulation of something that gives us an accurate impression of the real thing without actually reproducing it in it's entirety. I've thought the same thing about movie screens and film. What if a movie could be as dynamic as real world lighting? We'd have to wear sunglasses and sunscreen to watch Lawrence of Arabia

[jules]

We agree   .

My view is that seeing sound reproduction from this point of view, maybe with the product being comparable with a fine painting, leads to different technical solutions.

to be totally mischievous in realtion to your film analogy ... the black and white medium can challenge our imaginations sometime where colour fails.

jules

[timbley]

... the black and white medium can challenge our imaginations sometime where colour fails.

jules

What's colour?
Oh never mind. My girlfriend's Australian, and she just explained it to me.

[_scotty_]

jules, it is true that even a single instrument such as a saxaphone will over drive the acoustic of the average living room due to the acoustic watts it can put into the air and its peak spl capability.  However the sound of an orchestra can be recorded and even its peak spl reproduced in your living room without over loading your acoustics.  The point is, it takes much less radiated acoustic  power to reproduce the orchestras dynamic range in your living room which is why it will fit. A 60dB range is no problem at all assuming the intrinsic noise floor of the room is 60dB below peak spl desired. What cannot be captured and reproduced with complete accuracy at this time is all of the information that an orchestra can produce. The other problem is the forced perspective that the recording gives us of the orchestra. Turning up the volume doesn't exactly correspond to moving from a seat in row 10 to a seat in row2. The apparent subtended angle occupied by the orchestra doesn't become wider as it would if we actually moved closer to the orchestra. There is probably only one correct playback volume which corresponds to the apparent width that the recorded orchestra occupies in our listening rooms.  A single instrument is a simpler subject and fares considerably better when subjected to the chain of custody that finally puts it in your living room. The realistic reproduction of a symphony orchestra in my listening room is the equivalent the Holy Grail in my opinion and I hope to hear the greatest improvement in this area from advances in the state of the art of loudspeaker drivers.  Intermodulation distortion in loudspeakers is probably the single biggest culprit responsible for less than realistic reproduction of symphony orchestras by stereo systems and it is a subject  that has not been discussed in any great detail by loudspeaker manufacturers.  Compared to SS amplifiers most speakers will measure like crap.
Scotty

[timbley]

So what do we need, Horns, more drivers and crossover points, or both?

[nathanm]

The main difference I see between real life and the recording is simply that there are only two vibrating elements reproducing hundreds of complex vibrating elements as is the case with an orchestra.  Still, it is amazing what two speakers can do.  

I think it would be interesting to do an experiment where each instrument\performer was given its own single speaker cab and the speakers were placed in a room analogous to where the real performers would sit.  The tracks of each instrument would be recorded in isolation in near-anechoic conditions.  Then feed all the speakers via digital multitrack recorder in a lively room and see if that better approximates the "scale" of real world sound.  Practical?  Hardly!  But it would be cool.  Somebody did it with singers once, I think there was a Stereophile article on it.  It'd be fun to try it with a rock band or a string quartet or something.  Now that's what I'd call surround sound!

[Steve]

>>"Intermodulation distortion in loudspeakers is probably the single biggest culprit responsible for less than realistic reproduction of symphony orchestras by stereo systems and it is a subject that has not been discussed in any great detail by loudspeaker manufacturers.">

The highest harmonic distortion (highest IMD of 15% or more) ampliifiers come in pretty high to, in my opinion.

[thepogue]

i need a drink...  


Pogue

[WEEZ]

So after all this discussion, it seems that 'transparency' is good. And 'resolution' is good. And 'detail' means something different to some than to others.

I maintain that 'detail' is the result of transparency and resolution. And I guess I assumed that distortion (actually, the lack of it ) was a given.

Steve S. and Frank V.A. should know about distortion. They spend their life designing components to get rid of it. So do some speaker manufacturers.

Sometimes though, in an effort to reduce distortion, some products sound "thin". They shall remain nameless to protect the guilty  

So, the trick is, how to remove the distortion; smooth the frequency response; achieve transparency; increase resolution; and still retain a rich, full-bodied soundstage.

(if I knew the answer to that; I'd be employed in the audio industry )

 

WEEZ

[Steve]

The answer is simple, but hard to implement. Design and parts quality is the simple answer. But it took me along time to test parts and alot of thinking and testing designs.
 

[jules]

At this point I'm inclined to agree with Pogue  

On the other hand, as a final post I submit a dissenting conclusion to that of WEEZ with a short summary:

Scotty, in most ways I agree with what you say in the sense that I agree that it is possible to reproduce the sound of an orchestra in your own home in a way that is both satisfying and in some ways better than the live experience. At the same time you said:

"What cannot be captured and reproduced with complete accuracy at this time is all of the information that an orchestra can produce"

and this is the point I have been making. It is not only impossible to reproduce "all the information that an orchestra can produce" it is not desirable. 100% accurate reproduction of all the sound coming out of an orchestra would be overwhelming and completely lacking in transparency and detail. What we achieve through our many ingenious system designs is a refined selection of essence of the sounds that appeal to our ears.

At the macro [orchestra] and micro [single instrument] level we are far from reality. boead made the point that  a trumpet can actually be fatiguing to listen to in the stridency of some of it's output. Scotty made the point that a sax challenges the acoustics of an average room. My own ears tell me quite clearly that much goes missing from all sorts of instruments at an individual level. When you think of the difference between the medium initiating the sound [eg the vibrating tapered brass tube and horn that makes a trumpet or the 24" piece of skin that  forms a large drum] then it's clear that a paper cone is not going to be able to mechanically replicate the physics of movement involved [I won't go into this in detail but please think about it]

When I look at audio circles I realize that most of the interest starts at the recorded information point and goes forward from there. A musicians circle or maybe a recording/miking circle might add a wider perspective here.

jules

[WEEZ]

Steve- from many reports, you have succeeded.

The amplifier/speaker interface is something that still baffles me. In my case, my speakers are highly damped. With tube amps they sound great. And with some solid state amps they sound great too. But with some solid state amps they sound thin. I guess it's an impedence (damping factor) relationship.  

I do know this, however: I'll take high resolution/high transparency components over cloudy, mushy ones anyday. A component that masks detail with too much so-called 'smoothness' might be fine for background listening; but for serious listening- components like that make listening too much like work. Like I said earlier- the ear/brain has to work hard to fill-in the missing information. Tiresome and fatiguing.

(of course- artificial etching or detailing: i.e.: distortion; is worse)

We could probably discuss various audiophile terms forever. But when one hears reproduced music that sounds 'natural'- you're at least half way there.  

WEEZ

[Steve]

Thanks Weez. One of the things I test is soundstaging. With amps, I use two or three soundstage mapping selections and see if the sound generally fits their description. Granted there is some leeway, but one can get a pretty good impression.

I agree. I don't like cloudy, messy music either.

>>"But with some solid state amps they sound thin. I guess it's an impedence (damping factor) relationship."

Doesn't surprise me Weez. Could be that, or parts, or design, HD or possibly even IM distortion problems.

As an example, I have two articles comparing resistors, (only one resistor is used in the test, not replacing all the resistors) and notes of how each resistor changed the sound. The ratings (based on 0 to 100, with descriptions) are as low as 70% or less, depending on the brand/type.

That tells me alot. Think of all the resistors used in a design. And that doesn't even count capacitors, topology, even the type of solder, the number of solder connections, or wire used etc.

As one can imagine, as a whole, carbons are the worst, and none ranked very high. They all had significant sonic signatures. But a few metal films were actually worse than the best carbons. Harsh, glare, thin etc were typical findings.

Jules,

     I think it is possible to obtain alot of detail from orchestras, even in a living room setting, without messing things up.
I agree that the recordings are a weak link, but that is understandable with the poor equipment most use, at least from what I have seen.
At CES, 2004, a recording setup had seven, $10,000 mics and $100k speakers. Sounds good so far.

Then mated were some cheap "25 foot, .25/ft" mic cables and a huge console that appeared to come from a horror movie.  

One could easily hear the difference between the live band playing and the recording played back minutes later. Go figure.

[doug s.]

...As one can imagine, as a whole, carbons are the worst, and none ranked very high. They all had significant sonic signatures. But a few metal films were actually worse than the best carbons. Harsh, glare etc were typical findings. ...

curious if you tried riken-ohms.  this is one particular resistor that i found to be wery benificial in my modded art di/o.  it was actually better than the holco (old style copper leads) metal films i had initially installed.  smoother w/no loss in resolution...

regards,

doug s.

[Steve]

Yes Doug, I think I still have some around in my parts box. Been a long time, so don't remember much. I do know there are some better ones, but the rikens are probably better than 85% or more of the metal films out there.
 
 Glad it improved the sound for you.