Jon Risch distortion test signal

[J Risch]

Quite a flurry of activity, but all noise and no substance.
Since YOU continue to rail against me, and post the same tired old trash about the Phi Spectral, I need to point out the flaws and errors in your posts, as usual.

In post #127,
Jon Risch distortion test signal
you said:
" Doesn't change the fact that you cannot generate the signal easily with NBS traceable equipment. According to your paper, you're using a CD."

You aren't really serious about this kind of stance, are you? I provided a very easy way for others to duplicate what I did with a minimum of equipment and investment. This is not a crime or a problem.

As I noted in another post here, the signal can easily be generated via a full duplex soundcard by playing back a previously constructed wave file, and running the sound card with an FFT spectrum analysis program on the input. Other test/measurement equipment can generate and spectrum analyze the Phi Spectral signal, such as the AP System Two, one of the Rhoades & Schwartz units, the SYSid system, etc. One is NOT limited to a CD player as a signal source.

As for NBS tracability, this is not exactly that common among modern computer based THD analyzers or for spectrum analyzers, yet Cabot and many other well respected audio engineers can recommend and use a spectrum analyzer to measue THD and to measure IM, despite a lack of "NBS tracability". You are the one who seems stuck in the past, in love with old General Radio mechanically linked analyzers and THD meters. Just another red herring from you, trying to create another artificial hurdle to jump. Not too many people in the audio industry hang there hat on, or live and die by "NBS tracability".

"Doesn't change the fact that it cannot be interpreted with anything less than lots of computing horsepower."

I addressed this earlier, with regard to just how much computing power was needed.

The use of a simple Excel (or other brand of) spreadsheet allows one to hone in on as many distortion products as one wishes to check, including spcifically for HD, THD, IM, etc. Last time I heard, such a spread sheet will run on just about any computer that is still working!

I also called out the simple expedient of a visual comparison bwtween two DUT's, and to examine the overall level of the visibly displayed distortion products. This requires no more computing power than to run the spectrum analyzer software used with a soundcard, etc. Such programs have become very common and inexpensive, and readily available over the internet.

As for not having a dedicated program, or dedicated hardware or software to specifically provide these things automatically whe making measurements with the Phi Spectral, I am an individual, with limited resources, what I can afford to do or to provide all by myself is not necessarily an inherent limitation of the test method, it should be quite obvious by now that it would not take a programing genius to create such a software program, nor would it take any extraordinarily expensive hardware to implement it. A studio grade (full duplex) sound card should suffice for the hardware, and a simple variation on one of the existing spectrum analyzer packages availabel could easily be made to do the computational dirty work, including automatic distortion product analysis .

As I noted in the paper, this test signal was presented as a concept, one to be explored and to be worked with, and it would not be too difficult to create the necessary standard configuration, create a standard by which it could be made transportable, etc.
You are definitely in the minority if you feel that spectrum analysis can not be made to be acurate and repeatable.

In your post #130:
Jon Risch distortion test signal

You said:
"Again, I urged you to go look at Steve Eddys comments on this at AA, have you???"

Yes, he should go look, because then he would discover that what Steve Eddy was posting about was primarily his concen over wire movement, and whether or not it would be sufficient to generate significant levels of IM. In typical fashion, he went all over the map, got off on side issues, and generally ignored the content of my replies (much like you attempt to do here), YET NOT ONCE DO I RECALL THAT HE CRITICIZED THE ACTUAL TEST SIGNAL ITSELF, OR THE METHOD BEHIND IT. He was arguing about ONE of my tentative interpretations of the data with regard to ONE of the specific measurements I did using the Phi Spectral test signal.

So put the Steve Eddy train to bed, it is not at all like you are trying to portray it here.

You said:
" My argument isn't that you can't get some results, but whether they are useable, transportable, or can be standardized. And to now, the answer is no."

I have addresed some of the issues with regard to standardization and transportability. These are not insurmountable issues, and could easily be dealt with.

You said:
" Do they reveal anything that can be garnered by more elegant and simple means, no."

I have already addressed this at some length, and shown you to be wrong on this count.
You can post it all you want, but that won't change the facts. In point of fact, the new signal has been used to discover a form of CDP digital filter clipping distortion that occurs on transient signals, or signals with a lot of different frequency components, it has been used to provide a measure of loudspeaker sound quality that was NOT obvious from the traditional HD measurements, and has been used to detect VERY subtle and hard to spot forms of distortion that are all but invisible with the traditional HD and IM methods.

"Go do a Google groups search on Jon Risch and see how he was challenged there by his peers. He no longer posts there, because his theories could not stand up to the scrutiny of his peers."

This is another challenge you hope will go unanswered, because if they look, they will see that who I was challenged by were some of the very worst of the hardcore naysayers, folks who make FLZapped look like a pleasant and reasonable man. Arnie Krueger is notorious for being one of the most illogical, word twisting, unscientific naysayers on the newsgroups.

I quit posting there, because due to SHEER volume of posts (Arnie must do nothing else), the naysayers used enough bandwidth to asure that their POV would always have the last word, always have the last word twist, always take a last pot shot. Kind of reminds me of how this place looked not too long ago, where anyone who posted about cable sonics in the Cable forum would get shouted down by a horde of naysayers.

Let me show you how Jon gets erroneous results.....

First look here:
http://www.geocities.com/jonrisch/page7.htm
Please note Figure X - in particular, the curve labeld "Tweeter" which is green.
Now continue to:
http://www.geocities.com/jonrisch/page8.htm
Jon says this:

"This graph shows the reduction in IM distortion products when using two separate speaker cables to the separated crossover sections of a speaker, otherwise known as bi-wiring, compared to using a single speaker cable. - - - - Only the single cable vs. the tweeter cable is shown here for clarity."
I
So he has his test signal on a "single cable" system(black) is overlayed with the "tweeter cable" signal in a bi-wire set-up(magenta). He claims a reducton in IM, however, if you go back to the first curve, and overlay it with the results curve, what you are getting is NOT a reduction in IM, but just the natural action of that half of the system crossover. You would get the same results on a "single wire" system by going inside the box and connecting to the tweeter side of the crossover.

I explain this at my website, specifically for those who could not grasp that the IM products were NOT following exactly the simple curve of the crossover impedance slope.
See the Analysis section down at the bottom of page 11:
http://www.geocities.com/jonrisch/page11.htm

Besides, if you were to connect the "single wire" to ONLY the tweeter portion of the crossover, you would then be bi-wiring! If you literally meant a connection to the OUTPUT of the crossover, then it has become even clearer that you do not understand what has been written about, and what has been measured.

Please also note that you cannot clearly see the difference between any system generated noise and any generated intermodulation product.(Do you remember what I said about that earlier??)

This is more a problem with your ability to look at and understand what is seen with these graphs, than any inherent limitation or problem with the test signal or the measurements.

On these graphs, ANYTHING on Fig Z that is not a primary tone (which are the primary tones are pretty obvious, in this measurement, they are grouped together in two bands, as shown at page 10 in Fig Y) that is above approx. - 95 dB is definitely a result of a distortion product being displayed by the spectrum analyzer. Due to the use of the current probe, which has a 20 dB signal level penalty compared to a voltage measurement (the output of the current probve is 0.1 volt per 1 amp), I was close to the actual noise floor with these measurements. The measurement system noise was ALL well below -95 dB, but I am being conservative in stating that anything above -95 is quite certainly a distortion product. As with ANY spectrum analysis that is looking at such a wide dynamic range, one must always establish the noise floor baseline, and be sue they are looking at something other than noise. This is true of ANY sort of spectrum analysis, and not just of my test signal or my measurements.

Once again, I clearly point the facts out at my website, and explain why what Bruce says is totally incorrect and once you read the explaination, it becomes obvious where Bruce has gone wrong.

Steve Eddy pretty much made these same observations, including errors in Jon's methodology as he is more familiar with the test equipment being used.

No, actually, he did no such thing. I want to see the exact post referenced where he said this.
He did NOT make these same observations, nor did he point out any errors with my methodology. He argued about wire movement and IM, rather than the actual test signal itself.

These types of fundemental errors is why Jon gets the hell beat out of his theories all the time by his peers.

Bruce, it is these kinds of slanted and untrue statements such as you have made in the last couple of posts, that show why you and so many of the naysayers are more interested in the personalities, rather than the real science involved.

There ARE no "fundamental errors" for me to get beat up over, rather, you and a few others (both now and in the past) have shown that, in fact, it is your failure to comprehend fully that is resulting in YOUR fundamental errors of interpretation. I find it ironic, that you then ascribe these errors on your part, to me. You have done this repeatedly throughout this thread, and yet deny it has happened, instead, acting as if I were the one who was so grossly in error.

It is unfortunate that what is being discussed is as esoteric as it is, because it would be immediately clear to someone familiar with these kinds of audio measurements, that what I am saying is correct, and that your comments are the ones off in left field.

Jon Risch