Syno8, you just don't understand the REAL WORLD. I was THERE when Prof. Ashley admitted to Walt Jung that he rejected Walt's et al SID paper for the JAES due to discussion on the Volterra Series in the paper. Walt and I happened to be giving separate papers for the IEEE in Tulsa, OK for an IEEE Audio conference at the time, 30 years ago. This intimidated the other two authors, one who worked for Sound Technology, and the other who worked for Dolby Labs at the time, to withdraw from the paper entirely. There was nothing wrong with the math, in my opinion, although it was a little 'over the top' for such a paper.
Syno8, I do NOT appreciate your deriding of my experience in these matters. If you can PROVE that I am:bs: I will stand corrected, but unless you know for sure, you had better watch your mouth.
Syno8, I do NOT appreciate your deriding of my experience in these matters. If you can PROVE that I am:bs: I will stand corrected, but unless you know for sure, you had better watch your mouth.
john curl said:
Well, we are talking about different things here (conference papers vs. Journal papers, see my comment above) but honestly I fail to understand why you need to defend Walt's work; I'm sure he can do it by himself if he thinks it's worth of. I'd rather believe there was a problem with the paper rather than a problem with prof. Ashley. FYI, I was (15 years ago) a regular reviewer for IEEE Transactions of Electron Devices and member in the Selection/Program Comitees for several important IEEE conferences and I'm pretty familiar with the IEEE procedures. It happens that even for conferences the IEEE policy is still towards two stages (short and extended abstracts) multiple (usually 5) anonymous reviews. Most likely, prof. Ashley was voicing a collective decision. BTW, I had my own rejections and never bothered to blame anybody for that.
On the same note, it would certaily help the dialog if you could use the same standards you are suggesting above, for your own contributions here.
BTW, I don't have to prove anything to you. I am speaking from my own experience and, with all due respect for your age and experience, you are certainly not the one called to tell me to watch my mouth. Thank you.
It is not a funny story, AES / IEEE,
but this is how my mind re-acts:
Anybody tells he is or was in any way associated with audio organisations, non-public groups,
with exclusive memberships for people with similar background
... will make me associate him with some poor 'wanna-be-something by belonging'
... a sort of audio super jetset, if you will
And in my eyes this is in itself bit suspect.
As a man doing good audio works, should be good enough on his own.
Not need to be judged by who is reading and reviewing his stuff.
But his stuff should be good on its own merits.
It is just the same as when I read amplifier reviews/testimonies
on a website, as a promotion for audio products.
So, in my eyes, AES or IEEE, on the CV of any of diyaudio.com member is very repellent.
Some may be sorry I tell this. But this is the truth for me.
In this way I am sure I am a real prejudgist
Lineup
but this is how my mind re-acts:
Anybody tells he is or was in any way associated with audio organisations, non-public groups,
with exclusive memberships for people with similar background
... will make me associate him with some poor 'wanna-be-something by belonging'
... a sort of audio super jetset, if you will
And in my eyes this is in itself bit suspect.
As a man doing good audio works, should be good enough on his own.
Not need to be judged by who is reading and reviewing his stuff.
But his stuff should be good on its own merits.
It is just the same as when I read amplifier reviews/testimonies
on a website, as a promotion for audio products.
So, in my eyes, AES or IEEE, on the CV of any of diyaudio.com member is very repellent.
Some may be sorry I tell this. But this is the truth for me.
In this way I am sure I am a real prejudgist
Lineup
We again' have a failure to communicate'. Walt and I, or even Matti Otala are not against joining organizations and learning from their internal communications. In fact, I am a life member of the IEEE at this time, after 40 years of continuous membership and subscription to many of the subgroups within. However, it is rare that the IEEE, and even the AES in the last 25 years has really helped me design better audio products. Both organizations are just too esoteric and narrow, to really help real world audio designers.
For example, I still belong to the IEEE 'Circuits and Systems' 'Instrumentation and Measurement' and 'Consumer Electronics' groups, but I just get 1 or 2 real breakthroughs (for me) in a year from these groups, and most of the articles are just about unreadable, at least to a non-specialist.
The reason that the AES was formed initially was to give real design engineers a forum for exchanging ideas. The IRE (the former IEEE) already had an audio section, but I suspect that it was a bit too 'lofty' for the majority of audio designers, even back in the '50's. The AES for several decades was readable, and useful for real audio design engineers. Membership was not automatic, as you had to qualify, but you could always be an 'associate' and still get the info, and go to the conventions. About 30 years ago, this started to change, and both Walt Jung and I started to separate from the AES, because it rejected all the interesting stuff, and only put in the 'insider' stuff. Internally it was more like a 'Royal Order of Raccoons' convention rather than a serious society, back 30 years ago, which also put us off.
It is important that real audio designers get real information, rather than the ultra technical, impractical inputs, from publish-or-perish professors, and the like.
For example, I still belong to the IEEE 'Circuits and Systems' 'Instrumentation and Measurement' and 'Consumer Electronics' groups, but I just get 1 or 2 real breakthroughs (for me) in a year from these groups, and most of the articles are just about unreadable, at least to a non-specialist.
The reason that the AES was formed initially was to give real design engineers a forum for exchanging ideas. The IRE (the former IEEE) already had an audio section, but I suspect that it was a bit too 'lofty' for the majority of audio designers, even back in the '50's. The AES for several decades was readable, and useful for real audio design engineers. Membership was not automatic, as you had to qualify, but you could always be an 'associate' and still get the info, and go to the conventions. About 30 years ago, this started to change, and both Walt Jung and I started to separate from the AES, because it rejected all the interesting stuff, and only put in the 'insider' stuff. Internally it was more like a 'Royal Order of Raccoons' convention rather than a serious society, back 30 years ago, which also put us off.
It is important that real audio designers get real information, rather than the ultra technical, impractical inputs, from publish-or-perish professors, and the like.
PMA said:
Amen, PMA. Many people do not realize that the same nonlinearity causes all of these different nonlinear distortions. The tests are just ways to reveal the symptoms of nonlinearity differently. Some tests are more sensitive in some ways than others, but they are all exercising the same nonlinearity.
Cheers,
Bob
john curl said:
The main advantage of the three-tone Multitone InterModulation MIM test that I developed many years ago was that the triple beat nature of it brought the odd-order IM products down into the 1 kHz area of the audio band. It also made for very simple instrumentation. Back then, a good audio spectrum analyzer was very expensive. A limitation of the two-tone CCIF IM tests is that the odd-order IM products lie at high frequencies near the two test tones.
Cheers,
Bob
syn08 said:
Which pretty much shows you don't have the foggiest idea on how the peer review process is running, at least for important scientific journals.
It's not a "professor" that accepts or rejects a paper, but the paper is sent to 3-5 individuals that are usually selected based on their previous (on topic) contributions for the journal. The reviewers have first to choose if the paper is on the journal topic. Then the content evaluation form that each individual has to fill in has usually 4 grades "Publish immediately", "Publish after a minor revisions", "Publish after a major revision", "Reject". Only if the peers do not agree on a verdict (e.g. one says "Publish immediately" and another one says "Reject" the paper goes to a member of the scientific board, a "professor" but this happens once every decade, I guess. The evaluation reports with the grade (which also contain a detailed description of the objections) are then sent back to the author(s) and it is up to them to revise the paper according to the objections (or debate them) and resubmit. It is important to mention that the reviewers are completely anonymous for the author(s) all along the process.
So the claims that a certain "professor" rejected a particular paper submitted to a scientific journal are just :bs:, unless the journal is a piece of junk by itself, not worth of any scientific classification. Of course errors happen in the process, even famous scientists had rejected papers, there are ideas that were way ahead their time, etc... it's a fact of life. But, in general, the system is as good as democracy. Not perfect, but the best of all worlds. Because of this, a brilliant mind won't be turned off by a first rejection; it would rather understand the objections and review his work in an attempt to improve the communication, add details, further explain and justify his conclusions, etc...
If you can imagine a better process, you are free to contact the journals and speak up.
Sorry for the OT, let's go back to voodoo
As a member of the AES review board, I can confirm that this is an accurate depiction of the process.
Although there can be occasional mistakes, I believe that the reviewers as a whole try very hard to be fair, and often help authors to make their papers more clear and easy to read. I don't think there is a bias against unconventionality, but there is certainly a preference to papers wherein the assertions are reasonably well-grounded in theory, fact, measurement or experience.
Cheers,
Bob
Triple beat distortion due to 3 or more tones has been known for many decades. I learned about it in about 1970 in class. However, it was put forth by Bob Cordell as a substitute for the Sine-Square test that we presented in 1977. The 3 tone test is not a bad test, BUT it cannot replace the Sine-Square test because it does not create a consistently fast rise-time that triggers TIM distortion, and that is the point of developing the Sine-Square test, in the first place.
Bob Cordell said:
Thanks for the remark, Bob. We both and many others for sure agree on this, and this is the reality. I guess that some would see a non-linearity problem easier from multiple tone spectrum, probably for the reason they might be unaware of correlation of THD and CCIF IMD, e.g. The relationship between even and odd THD harmonics, and difference tone spectral line and "skirts" around freq. twins might be a good example or lecture.
Regards,
Pavel
All sorts of distortion measurements were investigated by Otala et al, in the mid 1970's. Many comparative graphs are available from AES preprints from that time period. It has been found that it takes at least 2 separate tests to evaluate a complex transfer function, such as a feedback amplifier. One that emphasizes rate of change, and one that emphasizes normal static nonlinearity. No one test does everything well within the confines of a 20-30 KHz bandwidth.
PMA, it is difficult for me to determine what test you are doing, but if you look at our AES Journal paper, April 1977 p.172, you will see a comparison between a simple multi-tone test and a Sine-Square multi-tone test in Figures 3 and 4. The only significant difference is the rise-time.
John, it is the standard 2-tone CCIF IMD distortion test, 13+14kHz in this case. You know, this signal resembles 100% amplitude modulated sine, and slope (derivative) at zero crossing changes during period.
I know the paper you speak about, I am just trying to mention that different methods reveal the same problem, as there is only just one non-linear transfer function of the non-linear system; and this function covers slew rate as well; the case is to make a choice of the 'right' test signal, that quickly reveals issues.
I know the paper you speak about, I am just trying to mention that different methods reveal the same problem, as there is only just one non-linear transfer function of the non-linear system; and this function covers slew rate as well; the case is to make a choice of the 'right' test signal, that quickly reveals issues.
- Status
- Not open for further replies.