jubilee

[sfogg]

"is it your position that this is how to determine the most accurate amplifier from a pile of amplifiers? "

No, but it is easy to see of the 'two' in this test which is more accurate in the domain presented. I didn't show the phase information but one had flat phase response the other changed in relation to the load. 3 guesses which was which.

"And what speaker is the load in the output shown here?"

NHT SuperZero.

"And what is the length and gauge of the speaker wire being used?"

I think it was around 10" of a fairly small gauge. Same wire for both scenarios.

Shawn

[jwc]

Welcome back 6foot8

I would love to see a picture of your room showing the three way jubs and the electronics.

jc

[sfogg]

"What's fun about these discussions is that nobody can talk about
measurements without being labelled as never listening to the music or
appreciating the art or whatever. The beauty of data is it's a way to
describe what you're hearing. "

That is for sure. Instead of randomly guessing what it is someone likes about something they can actually try and figure out the reality of what they like.

Shawn

[sfogg]

Mike,

Yes, I only posted the FR because it is something most people are more familiar with then looking at phase changes or whatever. The above test also showed phase changes of course. I originally had the FFT showing both FR+phase but it was harder to see the changes because of the smaller screen real estate for the measurements. Also, the posts were really to show that it is very possible to objectively test an amplifier using real music. It isn't all sines and squares....

"It's a bit more complicated than just that too. The output impedance of the amplifier will also have a reactive component that will also affect the transient behavior. The actual impedance of the speaker is always changing too. There is always the hypothetic perfect amplifier for driving the super complex load, or one can just simply lower the output impedance of the amplifier and keep the variability to a minimum."

Right, and the output impedance of the amp also effects the behavior of the speaker with regards to its back EMF which also changes transient behavior. Loose/flabby bass vs. 'tighter' (better damped) for example.

Shawn

[bodcaw boy]

Boy oh boy !!! what a bunch of loony squirts...Papa buys you the books sends you to school what do you do but eat the covers...You cant observe the accuracy of the amps without compairing it to the origional LIVE source...That is the ultimate test...And I do it at the studio with live musicions then objectively switch back to moniters and amps.....Do you want to know something interesting?? Nothing ,,I mean absolutely nothing has ever came close to that LIVE event.....And ive been doing this for over 50 years.... All the test equipment,, amps.. speakers,,, Anechoic chambers,,charts,, NONE have ever come close,,, Weather listening to a single flute or a full 70 piece orchestra...( I do this three times a week) And you guys are looking for accuracy and low distortion in an amp or speaker???? Jesus Christ !!! your not going to find it here on this planet. ROY don,t waite too long.

have absolutely no idea what you are talkinig about maron!! you need to expand your universe. and by the way, i will pray that you will be forgiven for taking my Savior's name in vain. that in itself shows the maturity of your thinking.....

in Christ, because of God's grace,

roy

[bodcaw boy]

Mark,

Nothing that isn't already well understood... amps with higher output impedance interact with a speakers impedance. Which in turn alters the amps FR to make it less accurate. A sort of uncontrollable EQ which makes system matching more important since the same amp won't sound the same on different speakers.

Happens with test signals, happens when you play the Rolling Stones too.

Shawn

Yes, speakers are not resistors. As you said, well understood. In light of the original question then, is it your position that this is how to determine the most accurate amplifier from a pile of amplifiers? If so, which one(s) did you test? I know someone years ago ran dozens of these (Moncrief?), but which did you personally test? And what speaker is the load in the output shown here? And what is the length and gauge of the speaker wire being used?

and also shawn,

what brand wire is it?[8-|]

in Christ, because of God's grace,

roy

[Coytee]

6foot8,

"It's the first clean 500 watts that really count..."

You da man!

[Y]

[bodcaw boy]

Hey Roy,

I don't know that I have ever thanked you for those wonderful speakers you and Chuck worked so hard to get me. So I would like to do that now...Thanks! I Talk to and email Chuck frequently ant still owe him pictures of the completed room and set up from our last discussion a few months ago when the KPT 884 showed up. I use a complement of 3 BGWs for the Jubliees and a QSC 4050HD for the sub. I remeber reading somewhere that PWK had an affinity for one particular amp that is near and dear to my heart, the BGW 100. It was back in the early 1980s that I started using those amps, and Klipsch speakers, in part because of each products reputation. That was a very important factor for me in selecting the amplifiers I used for the speakers you designed...PWK's opinion of the BGW 100. I am certain that there are some other amps (both Tube and SS) that would sound better on some recordings and not as good on others but I like what I have and it sounds good to me. For me I think it was the opinion PWK had of that amp and the nostalic experience of going back in time for me that was the two biggest determining factors of my slection...and I think it sounds very good none the less.

I have played around with settings and balances, EQ...etc etc for a year and a half or so and found the "tune up" that sounds very good to me without contantly changing bass, treble or the balance between the amps between each song. And I admit that although I break out an "electronic measuring device" occaisionally the over riding factor is whats sounds good to my ears. And while I am sure there are purchases that I could make that would enable me to travel a little farther down that yellow brick road to audio nirvana, I have come to that place in the road where I am fully content. I think anything else would be just "different"...not necessary any better overall and most certainly not worth the added expense.

Thanks again!

hey eric,

i am glad that you like the 3 way jubs. you were the first and only three way guy (that i know of) and now you have an 884 to boot (bill hendrix also recently bought an 884 for his jub ht; ask him about it)! most songs are not mixed flat (either bad engineers or bad monitors or both) and so you have to flavor the top and bottom end just a bit. i would like to see pictures of your setup if you don't mind.

in Christ, because of God's grace,

roy

[Deang]

"The klub amp will be producing far more "nasty" distortion, almost like a correlated noise signal riding on top of your music. Hence that grayed-out, crunchy, dry sound of a solid state amp at low power."

The amp(s) I have in here now don't sound like that. I was expecting them too, but they don't. I even had close to another $1000 hid away in case I had to send them back and get some "real" amps. I've quit pretending that I'm a techincal person so I'm not going to speculate as to why they don't sound anything like the sandpaper I remember from two decades ago. Something I always expect to hear, even with "good" solid state is some hardness to the sound of the notes, and the leading edge of the transients are usually a tad spitty -- a kind of brittle quality to the sound. O.K., well, I don't hear those things. The sound at lower volumes is different than tubes, but not dramatically different. The lack of grunge and clarity surprised me, but even more surprising was the sense of air and natural decay on the trailing edges of the notes. I wish I knew what was different about this stuff than twenty years ago -- the only thing I know about are the switching power supplies -- they only weigh 19 pounds each.

So the second Crown XTi 1000 shows up yesterday and I decide to take off of work early and get it done. I pulled the passives (which almost broke my heart), and started with the wiring. I stacked the amps and pulled out the laptop, which was kind of depressing because it seemed way too much like my real job. It sure feels wierd programming an amplifier. Not much to it really, you just have to pay attention so you don't do something stupid -- pretty much the same as building real networks. I get all of Roy's values into them and fire them up. A little noise, just a little -- I can just barely hear it from the chair. Bummer. OTOH, I've had tube gear which was much worse. I always start with the Moody Blues, always. I listened to most of it and finished about 3 in the morning. I really wanted to move on to something else after the first CD, but it was sounding really good and I knew if I was going to find an issue with the sound it was going to show up with this material -- it always does. I'll run a lot of different stuff though the system today, it's going to be fun.

Roy, I've stayed with the 48dB/octave LR alignment, it definitely sounds smoother, and yes -- I figured out the 'Q' thing.

[Deang]

BTW -- I don't care what kind of stuff you put in front of Heritage, it won't get within a light year of what I'm hearing right now.

[bhendrix]

AMEN, Dean!! I was up until 3:30am listening the my "Big Uglies" and "The Beast" (Jubilees with KPT-884) and karaoke amps (XTi's) and I have NEVER in my 56 years heard anything that comes close to what I was enjoying. The listening "sweet spot" is huge and the sound stage is expansive and detailed. Never before have I been able to position instruments or singers in the sound stage quite like I was hearing the other night.

You will be equally amazed when you add the 884, aka The Beast, to your Big Uglies to capture the deepest of the bottom octaves. The deep bass is clean and expansive even at very low listening levels. It is much better than any of the powered pistons in tiny boxes that I have heard.

[jacksonbart]

congrats on the set up.

[Deang]

"I'm not clear why you went through that exercise..."

Mark, you asked Shawn more than once for the measurements, and when he does them and shows you -- you question why he did it?

There is accurate sound, and then there is accurate reproduction of the signal -- I agree that the two don't always correlate. Clearly though, the amp with the low output impedance is faithful to the source signal, and the one with the high output impedance isn't. The end result in that context doesn't seem to be much different than applying EQ.

[Deang]

congrats on the set up.

Thanks, I appreciate that!

Bill, the 884 will have to wait until I can figure out where to put one!

[Guest " "]

measurements....measurements...measurements

I don't think anyone can measure what I hear anymore than anyone being able to measure what I can taste or feel.

Measuring equipment is still primative relative to the capabilities of the eye's, ears, tounge, and our sense of touch.

And certainly, these sense's interact and alter our perceptions.

Now...about those 3-way Jub's with the sub-woofer.......

[Deang]

I see, O.K., I get it -- thanks. I did enjoy those posts a few pages back. I basically agree, but in the end the only thing that matters is the actual sound. Amplifiers are just one part of the chain, and I've believed for some time that good ones, even great ones -- don't bring anything near the level of improvement to the sound as a better loudspeaker does. How many amplifier dollars/quality does it take to equal the performance/sonic gains found in better drivers and horns? A pCat/Peach setup approaches the cost of a pair of shipped Jubilees, and driven with even a modest amplifier -- sounds much more realistic and engaging then anything I ever heard pushing my Klipschorns -- both stock and modded. So, the amp debate seems kind of out of place to me in this thread. If you're sitting there with a pair of LaScalas or Klipschorns, and you're getting ready to spend close to $7000 on a new amp and preamp -- you might as well put the money towards a nice lobotomy.

[rigma]

Dean,

You just hit the nail square on the head!!!!!

rigma

[PrestonTom]

Not surprisingly, there are some things that have been said that will be confusing for many. Let me just tackle one: the issue of frequency response (as measured by a transfer function or impulse). This is a linear systems technique that is most useful for describing whether a signal is reproduced accurately in terms of its frequency spectrum (I will ignore the phase spectrum). For instance, a filter will attenuate some frequencies more than others and this is a perfectly linear operation (a phase shift can also be a perfectly linear operation). What Shawn is suggesting is to divide the input spectrum by the output spectrum (or using dB we can subtract these terms) to look at the transfer function. This is perfectly acceptable and can be quite descriptive of one aspect of "accuracy". The reason you would not want to use a Rolling Stones CD pertains to practical issues about whether there is sufficient energy across the spectrum to get reliable estimate of at a given frequency band for both the input and output (problems with noise floors contaiminating the measures etc). So there are better candidates for the test signal. Humans are quite sensitive to aberrations in the frequency spectrum, but guessing whether a specific aberration will be heard as "boomy", "forward", or whatever (even whether the difference can be detected at all) is difficult. I am ignoring changes in the phase spectrum, since we are less sensitive to those changes (not entirely, but I won't get into specifics).

Distortion is another issue. Most folks are familiar with the terms harmonic and intermodulation distortion. These can have many forms (especially IM) and they may be transient. This can make them less easily identified and they are frequently quite dependent on what frequencies are in the "test signal" and what the overall level is. It is important to note these are a result of a non-linearity. That is they can be level dependent and more importantly the number of frequency components at the output is greater than the components at the input. That is "extra" frequency components are being generated. This is also an inaccuracy

Now the important part and why you would not use a Rolling Stones CD to try and identify distortion (nonlinearities). One is that it would be difficult to see non steady-state distortions (eg, transient IM, which incidentally was a great concern to PWK, who among others, published on this). Second relates to the nature of a decibel scale. If the test signal was broad band (like a Rolling Stones CD), having many frequencies at high levels, any resulting distortion would be nearly impossible to see using a transfer function. If the "test signal" had 100dB or energy at 1000Hz and 100dB energy at 3000Hz (and frequencies in between and adjacent: i.e., a flat broad band signal) and the signal was somewhat steady state (ongoing) what would you see. Well, if the 3rd harmonic distortion was 5% (which is large and can be quite audible and aversive), the 3rd harmonic would be -26 dB down from the input (in our example it would be 74dB SPL and I will ignore its phase). Now the question is what is the level at 3000 Hz for the input (with no distortion): it is 100dB. What is the level at 3000Hz at the output (with the 5% distortion) it is 100 dB + 74 dB. This sum equals 100.1dB. The bad news is that it would be difficult to identify a 0.1 dB increment on the transfer function. So the issue of a test signal and what equipment is not a trivial one.

Since accuracy reflects changes in the frequency response and the phase response (both linear operations) as well as the introduction of added components or distortion (which can result from a number of processes), identifying them can be difficult. It is not a simple issue and can take some digging. It is also clear that using "music" would not be first signal to use when searching for an "inaccuracy". Although you might want to use this later when trying to determine whether the aberration is actually audible in "normal" circumstances (and whether it is necessarily aversive to any great extent).

Obviously there are many forms of distortion and there are certainly other aspects to "accuracy" of the reproduction (e.g., channel separation, noise floor, compression or the ability to reproduce dynamics - which may be hard to see with a spectrum analyzer). But these phenomena, and others, are obviously at the root of whether a device will "sound different". The relationship between the degree of aberration (physically measured) and the degree of audibility (subjective percept) can be elusive. BTW, for these kinds of auditory discrimination's, human listeners tend to be more similar than dis-similar.

I will underscore two things.

1) Dr Who (the blasphemous & young whipper snapper) pointed to the direction of some other "time domain" inspired techniques. These would certainly be of interest.

2) Shawn's comment (I believe it was him) that the real generator of distortion (and an altered frequency response) is going to be the speakers not the amplifier (unless maybe it is driving a difficult load - low efficiency and low impedance). As far as accuracy for a faithful frequency response, the room interactions will play a surprisingly substantial role. This is where the money & effort should be spent. IMO, you do not need to spend a fortune to get an accurate sounding amp. However, you do need to spend a fortune to get relatively accurate sounding speakers.

Now, can we all be friends again?

-Tom

[whatever55]

... so maybe Bose has something going.

The next step in audio revolution. Not only a cheap amp ... but cheap pistons also... and so many.

[DrWho]

Most were deeply puzzled by the lack of correlation between measuring and listening.

I would argue that if there is no correlation between the measurement and the listening, then you are incorrectly interpretting the measurement.