Very true. I was just responding to a PM where I mentioned more or less the same type of concern: An AP machine doesn't test devices under real world conditions. Thus, to sell products to people who only look at AP numbers, no manufacturing cost needs to be spent on things like EMI/RFI noise immunity, etc., and customers don't want to pay for it anyway, since there is no "proof" it is needed. For some people, Value = SINAD/$
Set up a test environment and test with a transmitter with varying frequency/output power. Too bad I don't have access to the measuring equipment anymore but I have done such spectrum measurements. A simple smart phone test can also be pretty convincing. Another simple test is to use Powerline devices with audio devices. You won't have any numbers but you have noticed at least something. Visit someone that lives in an apartment where large cell phone transmitters are installed on the roof and bring your own audio devices etc. etc.
Is there really a need for some person somewhere to be the definitive answer on a global scale?
Is there really a need for some person somewhere to be the definitive answer on a global scale?
Last edited:
We are dealing with multiple problems. Some things are mostly impractical to measure (at least with today's tech), some other things are impractical to usefully generalize. Humans sometimes hear things that aren't real, other times they don't hear things that are real. ABX isn't as simple as it seems like it should be, as evidenced by the now well-known false-negative bias (at least its well-known by today's perceptual scientists, just as it is now known how to overcome it). The list goes on and on.
Unfortunately there are no practical solutions for some of these things. If there were, then the list would be that much shorter.
Yeah, but an awful lot of things really are practical to measure, but aren't. For example, look back to post #29 on this very thread. (I'm pointing to a post I made, so that nobody else gets caught in the crossfire. I am hardly the only one to make a measurement like that.)
Just why does nobody measure CMRR? Or much of anything else beyond what was being measured back three quarters of a century ago? Heck, you can measure CMRR with a soundcard. The measurement I presented was made with about $700 US of equipment, not including the computer. You would think that audio equipment manufacturers and audio reviewers could make measurements at least as sophisticated as that. But, they do not. At least I haven't seen them. (Again, somebody tell me where I can find them if they do exist.)
I am certain that there is a good answer to all of that. I just don't know what it could be.
Just why does nobody measure CMRR? Or much of anything else beyond what was being measured back three quarters of a century ago? Heck, you can measure CMRR with a soundcard. The measurement I presented was made with about $700 US of equipment, not including the computer. You would think that audio equipment manufacturers and audio reviewers could make measurements at least as sophisticated as that. But, they do not. At least I haven't seen them. (Again, somebody tell me where I can find them if they do exist.)
I am certain that there is a good answer to all of that. I just don't know what it could be.
True.
Seems kind of like there are "standard" measurements which may sometimes be compared like baseball card stats.
Also, the results of some measurements have not been studied for correlation with audibility. So, why measure something when there is no "proof" it could have an audible effect? And even if it could potentially have an audible effect, at what point should it be considered a problem?
In addition, sometimes people find reasons not to bother considering some types of possible problems. Don't know how many times I have read that GHz frequencies cannot possibly affect audio because they are so far above 20kHz. Similarly, have read to the effect that, "shielded power cords cannot have any audible effect because they are only a tiny part of the total resistance going all the way back to the power company generator." Sometimes the people who say things like that get "likes" for saying it. What can you do when the more experienced and knowledgeable forum members don't want to bother responding to such claims, which can then end up as commonly believed audio forum memes?
If an average electronics DIYer concludes something is going on (like stuff sounding different when the day passes) but lacks the knowledge/equipment to stick a measured number to it mitigating the effects by experimenting does not hurt at all. This as compared to doing nothing. When I had access I brought the RF test equipment home a few times to check my environment of then. It resulted in a separate filtered audio mains group. Also there was RF stray in on loudspeaker cables due to a nearby transmitter reduced (but not 100% solved) with clip on ferrite rings. I can definitely not measure such today anymore but one can keep to basic RF/EMI reducing methods by construction/experience/empirical tests.
No one needs an academic title to do so or to explain stuff fully nor to grasp and analyse every situation from all theoretical angles. Maybe the improvement was not the maximum that could be achieved but it was an improvement anyway. It is a hobby to most, not a lab setup.
Final results of simply being passive are worse.
No one needs an academic title to do so or to explain stuff fully nor to grasp and analyse every situation from all theoretical angles. Maybe the improvement was not the maximum that could be achieved but it was an improvement anyway. It is a hobby to most, not a lab setup.
Final results of simply being passive are worse.
Last edited:
"A policeman sees a drunk man searching for something under a streetlight and asks what the drunk has lost. He says he lost his keys and they both look under the streetlight together. After a few minutes the policeman asks if he is sure he lost them here, and the drunk replies, no, and that he lost them in the park. The policeman asks why he is searching here, and the drunk replies, "this is where the light is". https://en.wikipedia.org/wiki/Streetlight_effect
Some things are easier to measure e.g. distortion and SNR, however anyone who believes an SNR of 125 dB will sound better than 120 dB has no understanding of the reality. The thermal noise floor of air at room temperature probably limits the SNR to -120 dB SPL (@ 20C fun useless fact). And typical quiet ambient noise levels are closer to 25 dB SPL (A wtd).
Pretty much all our measurements are steady state which music is not. And perception vs. measurements should take in everything from mood to ambient temperature. It has been reported that Gold grilles on Quad ESL's sound warmer than black grills. The sound will be identical but the reaction of the listener will reflect aspects that acoustic measurements cannot measure.
However arguments over shielded power cords (which can be measurable) become tedious especially when neither side understands the technology. It would be interesting to measure the impedance of the power line across the usable spectrum. At power line frequencies its quite low, on the order of 0.01 Ohm. However at harmonics and higher its probably much higher. I have had what seems to be good results terminating power cords with a 100 Ohm resistor in series with a small cap. This should prevent reflections and reduce unwanted conducted EMI from radiating from the cable. A ferrite clamp should also help. However I have had audiophiles claim ferrites ruin the sound. Back to perception versus measured reality.
Some things are easier to measure e.g. distortion and SNR, however anyone who believes an SNR of 125 dB will sound better than 120 dB has no understanding of the reality. The thermal noise floor of air at room temperature probably limits the SNR to -120 dB SPL (@ 20C fun useless fact). And typical quiet ambient noise levels are closer to 25 dB SPL (A wtd).
Pretty much all our measurements are steady state which music is not. And perception vs. measurements should take in everything from mood to ambient temperature. It has been reported that Gold grilles on Quad ESL's sound warmer than black grills. The sound will be identical but the reaction of the listener will reflect aspects that acoustic measurements cannot measure.
However arguments over shielded power cords (which can be measurable) become tedious especially when neither side understands the technology. It would be interesting to measure the impedance of the power line across the usable spectrum. At power line frequencies its quite low, on the order of 0.01 Ohm. However at harmonics and higher its probably much higher. I have had what seems to be good results terminating power cords with a 100 Ohm resistor in series with a small cap. This should prevent reflections and reduce unwanted conducted EMI from radiating from the cable. A ferrite clamp should also help. However I have had audiophiles claim ferrites ruin the sound. Back to perception versus measured reality.
There was a time when manufacturers competed on specs and needed to publish measurements to prove it.
Nowadays, why would you publish measurements that may cause discussions, when you can sell with flowery prose and shiny pictures and movies?
So they may still do measurements, but they don't publish them.
Jan
Nowadays, why would you publish measurements that may cause discussions, when you can sell with flowery prose and shiny pictures and movies?
So they may still do measurements, but they don't publish them.
Jan
Don't know about for other equipment, but in some cases ferrites do seem to cause problems with clock phase noise in dacs (although the effect does not show up very well in correlation-based phase noise measurements). OTOH, ferrites in Monster HTPS 7000 common mode chokes (which I understand you were involved in the design of) are found to be helpful.
Point is, maybe better not to over generalize about ferrites always being good or always being bad. IME, it depends.
Also, thanks for the tip about RF terminating power cords at their approximate characteristic impedance 🙂
Ask and you shall receive! At least, this time.
This is from Henry Ott's book Electromagnetic Compatibility Engineering, page 493. "Excerpts" from the book can be found online...
So, when people suggest that a good median impedance for an AC mains line at high frequencies is 50 Ohms, this is the data that it comes from. A LISN is designed to present this impedance for testing for EMI/RFI compliance In simulations in LTspice, I use a balanced source with a 25 uH inductor in both legs in parallel with 25 Ohms to simulate the AC mains. The inductors also have a series resistance of 0.1 Ohms, which I figure is about what the wiring might be - that defines the low frequency impedance. +/- an Ohm or two.
So, your 100 Ohm resistor idea isn't far off.
Besides, there's ferrites and there's ferrites. It's a whole science and different materials have vastly different characteristics over frequency, with wildly different resistive and inductive components. Most have body resonances, too.
Just clamping some random ferrite piece onto a cable will give, umm, random results. Not always positive.
Books have been written about this, too. Lots online.
The problem is that the manufacturers only published a very limited set of measurements, and these were made with test gear that has been improved upon quite a bit. It seems to me that when Audio Precision came out with their System One, the amount of published measurements actually decreased rather than increased.
But, I get your point. They can't win any of those arguments. Why try?
One thing about ferrites that I don't see in most of the literature so far is their tendency to produce hysteresis noise/distortion. Sure, they can be lossy too, which is often the main focus of technical literature.
We tried applying the literature but found the ferrites to be more effective.
Ferrites properly installed don't cause problems!
All the conductors in the cable need to go thru the ferrite.
All the conductors in the cable need to go thru the ferrite.
For all ferrites including ferrite beads "properly installed" on PCBs, or only for ferrite cable clamps?
If all conductors in a cable are passed through a ferrite cable clamp, and if all the cable currents and their associated magnetic fluxes cancel out, then the ferrite does nothing at all, neither good nor bad!
In that particular case the ferrite isn't creating a problem?
If all conductors in a cable are passed through a ferrite cable clamp, and if all the cable currents and their associated magnetic fluxes cancel out, then the ferrite does nothing at all, neither good nor bad!
In that particular case the ferrite isn't creating a problem?