I think that 1/f noise is everywhere, but so what? The human ear removes most of it from being detected, maybe it is designed that way. Carefully done input stages may have MEASURABLE 1/f noise added to the Johnson Noise, BUT it will go away with A weighting, we do this all the time with audio design, especially phono and analog tape circuits.
No, you're just lying about what they've actually said. Or can't understand it. Take your pick.
John,
I am not arguing that price really has anything to do with how a wine tastes. That is a marketing and supply situation, not necessarily a quality one for me. It is just the fact that the study results are misleading and that it is a poorly stated premise. It is very easy to do testing and make false assumptions based on a poor test methodology. We learn this early in science, if you ask the wrong questions or leading questions you can get a false positive or null resultant. We could easily take this study and point at your own design of the Blowtorch. We could make a test where many people would chose a $200.00 Sony off the shelf product come out on top. If you incorrectly set up a blind test it does not make it a fact or even a valid subjective test. This happens all the time, not only in these subjects but in medicine. Proper testing is not an easy thing to do when it comes to subjective testing.
I am not arguing that price really has anything to do with how a wine tastes. That is a marketing and supply situation, not necessarily a quality one for me. It is just the fact that the study results are misleading and that it is a poorly stated premise. It is very easy to do testing and make false assumptions based on a poor test methodology. We learn this early in science, if you ask the wrong questions or leading questions you can get a false positive or null resultant. We could easily take this study and point at your own design of the Blowtorch. We could make a test where many people would chose a $200.00 Sony off the shelf product come out on top. If you incorrectly set up a blind test it does not make it a fact or even a valid subjective test. This happens all the time, not only in these subjects but in medicine. Proper testing is not an easy thing to do when it comes to subjective testing.
Two noise sources are fundamental in the sense of being unavoidable, calculable and understood. Thermal/ Johnson/ Nyquist noise is a function of temperature, resistance and bandwidth. Shot noise is a function of the number of electrons jumping a barrier and bandwidth. Both are "colorless", frequency-independent. ("Shot noise" was also once used to mean the thermal noise inside vacuum valves, which confuses us old-timers no end. Hopefully this will finally end some day, when RDH4 goes out of print. Hah.)
All other noises are called "excess", and get colloquial names like Flicker, popcorn, 1/f, etc.
None of these are fundamental, all are process dependent, unique to a particular manufacturing process to the batch or even device level. They can be estimated based on past history of a process but cannot be calculated in a true engineering sense. Some people may understand the details of the excess mechanisms (maybe), but mostly it's all ad hoc and French curves.
Thanks,
Chris
All other noises are called "excess", and get colloquial names like Flicker, popcorn, 1/f, etc.
None of these are fundamental, all are process dependent, unique to a particular manufacturing process to the batch or even device level. They can be estimated based on past history of a process but cannot be calculated in a true engineering sense. Some people may understand the details of the excess mechanisms (maybe), but mostly it's all ad hoc and French curves.
Thanks,
Chris
Point out the problems in the 'scientific' double blind wine tasting tests to me, Kindhornman, and then we can both agree. Then, I might point out the problems in the 'scientific' double blind audio testing that I have found over the decades, and you just might come to agree with me on this subject. '-) Do your homework, Kindhornman, before criticizing me about what I have brought forth.
There may not be any. The issue isn't their test, it is your repeated mis-characterization of what the test was and what the results mean.
John,
The subjective question was only could the entire sample size tell which wine was the most expensive and rate the wines by price alone. The problem with the result, not that anyone could identify the most expensive wine, but the conclusion therefore that we should never buy an expensive bottle of wine since they are all the same in taste. That was not what was tested, it was a test only to determine a selling price. Two very different questions and two different answers. The author made the conclusion based on the subjects ability to rank price only.
John,
You were going to go out on your birthday with Scott to an expensive restaurant. I assume that the food is good and that is why they could charge high prices. But if I did an experiment with people who didn't like that type of food and they said they would not pay a high price for the food does that mean that you should never go to an expensive restaurant because all food must taste the same, price has no effect on the quality of the food served? If he offered to take you to In&Out Burger would you have been excited about that. Price and quality are often not connected, but not all the time. The same with rating a wine based only on price. I worked in a medical lab when I was young. One thing we "had" to do was to run a control sample before and after the testing. This was not something we took lightly. If the result from the pre and post controls did not match the tests had to be done over, even if you thought that the results were correct. There was no issue with this. No wiggle room was allowed.
The subjective question was only could the entire sample size tell which wine was the most expensive and rate the wines by price alone. The problem with the result, not that anyone could identify the most expensive wine, but the conclusion therefore that we should never buy an expensive bottle of wine since they are all the same in taste. That was not what was tested, it was a test only to determine a selling price. Two very different questions and two different answers. The author made the conclusion based on the subjects ability to rank price only.
John,
You were going to go out on your birthday with Scott to an expensive restaurant. I assume that the food is good and that is why they could charge high prices. But if I did an experiment with people who didn't like that type of food and they said they would not pay a high price for the food does that mean that you should never go to an expensive restaurant because all food must taste the same, price has no effect on the quality of the food served? If he offered to take you to In&Out Burger would you have been excited about that. Price and quality are often not connected, but not all the time. The same with rating a wine based only on price. I worked in a medical lab when I was young. One thing we "had" to do was to run a control sample before and after the testing. This was not something we took lightly. If the result from the pre and post controls did not match the tests had to be done over, even if you thought that the results were correct. There was no issue with this. No wiggle room was allowed.
I explained it, very carefully, several times. I did the same the last several times you pulled out that press release. If you still don't get it (or more likely, won't get it), I can't help that, it's between you and your conscience. Your shot at Jan which kicked this off was not only cheap, but aimed the wrong way.
Just two questions, John:
-It is giving an insult to say that quoting somebody else as a proof of something is not a scientific approach ? don't you feel the difference with a personal remark ?
- What is a scientist ?
Yes, 3 questions, in fact, i'm not very scientific
Last edited:
John,
What does the sample size have to do with confusing the conclusion with the original question. I have taken statistics, I hated it but go through it. It has nothing to do with the mean, medium, outlier, or anything else if you ask a leading question and use that to get a result that had nothing to do with the original question. Why do you believe the conclusion, it was irrelevant to the study that was done.
John,
We could easily set up a test such as this to prove that high end audio is a fraud. Since all the devices are capacitors, resistors, inductors and even Opamps that are used in all audio products they are therefor all the same. There is no real difference between the general parts used. I could set up that type of test and use 1000 people in the study and word the question however I would like. I could get almost the entire test sample to agree that all of your products are a fraud. Does that mean it is true or I believe that, no, but a statistical test does not tell lies. It only gives you some information but you can easily skew a result in statistics and that is one of the lessons we had to learn in two semesters of that horrible math.
What does the sample size have to do with confusing the conclusion with the original question. I have taken statistics, I hated it but go through it. It has nothing to do with the mean, medium, outlier, or anything else if you ask a leading question and use that to get a result that had nothing to do with the original question. Why do you believe the conclusion, it was irrelevant to the study that was done.
John,
We could easily set up a test such as this to prove that high end audio is a fraud. Since all the devices are capacitors, resistors, inductors and even Opamps that are used in all audio products they are therefor all the same. There is no real difference between the general parts used. I could set up that type of test and use 1000 people in the study and word the question however I would like. I could get almost the entire test sample to agree that all of your products are a fraud. Does that mean it is true or I believe that, no, but a statistical test does not tell lies. It only gives you some information but you can easily skew a result in statistics and that is one of the lessons we had to learn in two semesters of that horrible math.
Last edited:
- Status
- Not open for further replies.