The measurement floor can be seen in the multi graphs, nothing hidden there.
But you are right, there was no base measurment of the mike given.
I have some experience with very high resolution measurements, I have resolved 0.005dB drift in an AP analog generator for instance.
So, when I see those minute differences between the graphs shown in those ASR articles, those differences may well be explainable by the equipment drifts or temp changes during the (short?) time delays between successive measurements.
But, if there were really large differences between cap performance, they would have overwhelmed the small differences shown.
The very fact that there were differences at all make those results pretty convincing. YMMV of course.
Jan
But you are right, there was no base measurment of the mike given.
I have some experience with very high resolution measurements, I have resolved 0.005dB drift in an AP analog generator for instance.
So, when I see those minute differences between the graphs shown in those ASR articles, those differences may well be explainable by the equipment drifts or temp changes during the (short?) time delays between successive measurements.
But, if there were really large differences between cap performance, they would have overwhelmed the small differences shown.
The very fact that there were differences at all make those results pretty convincing. YMMV of course.
Jan
For that particular filter topology. A simple tweeter with a series C doesn't necessarily hold for all crossover topologies...
It's always good to have meaningful measurements. However, what me measure may sometimes not be relevant to what we hear. We can measure differences in harmonic distortion at ppm levels, but our ears don't care. Vice versa, we may hear differences from parts that measure the same.
That said, my experience with caps is that I don't hear a difference with DC blocking caps, where very low currents are involved. With loudspeaker xovers, it's a different story. See here: https://www.diyaudio.com/community/...-cap-or-mundorf-m-cap-evo.357122/post-6435538
That said, my experience with caps is that I don't hear a difference with DC blocking caps, where very low currents are involved. With loudspeaker xovers, it's a different story. See here: https://www.diyaudio.com/community/...-cap-or-mundorf-m-cap-evo.357122/post-6435538
Could you enlighten some new members like me with a list of posts of the last 6 month or so, i am curious to see how long the list will be.
I have done a search with the key words capacitor and measurement first and did not find the long list but may be i am doing something wrong, many thanks in advance for your patience and wisdom
Re frequency responses completely agree. Such tiny deviations are testament to a careful protocol.
Stefano, you can actually search and read thousands of opinions in the hundreds of mostly useless capacitor threads, just do your due diligence, do not start yet another.
Lots of chutzpah here !!!!!
Just USE the search function, and then READ what you find.
But I suspect you are not really asking for enlightenment, only some own preconceived idea confirmation.
You say:
Like betting on a horse after the race is over 😉
REALLY? 😱
Lots of chutzpah here !!!!!
Just USE the search function, and then READ what you find.
But I suspect you are not really asking for enlightenment, only some own preconceived idea confirmation.
You say:
... clearly "testing the waters" before posting your own, but of course you will wait and count how many match it.
Like betting on a horse after the race is over 😉
I do not understand fully your quite angry style of communication to be very honest with you. Tried again with the keyword "capacitor" "quality" "measurement" and could not narrow down enough the hits. It would really take a long, long time to get through all these posts. I really do not have an own preconceived idea; this is the reason why i am asking here for all sort of opinions. More specifically i have asked for results of seriously conducted A / B comparison with listening tests with crossovers build with different grades of capacitors. For this specific question to read your post is a waste of time - BTW what does chutzpah means? I'm not a native English speaker - i apologize for my ignorance.
P.S.: what works great to search is about the technical relevance of the posts of user JMFahey - i think it is just right to ignore JMFahey's comments at all
Last edited by a moderator:
Back in 2010, when I was in the re-design stages of restoring my lovely old RCA Victor console stereo, I wanted to go 'the extra mile' and get the most of the 6 chosen speakers that I decided to use (woofers/mids/tweeters)
My A/B testing of the midrange speakers was perhaps the most 'critical' part of tuning that system, since midrange contains most of the audio spectrum.
Now mind you, I'm not a 'nitpicker' as some seem to be on here, just picky enough to lean on the 'perfection' side of things.
A white-noise generator, an audio frequency generator, and a simple A/B switching setup provided me with a comparison of two capacitors in series with the midranges.
I already knew what value I wanted for the crossover, a common value 4.7uF cap.
switch loaded with both a 4.7 mylar and 4.7 electrolytic, I switched back and forth at various frequencies.... and listened to each cap's response.
I then used various musical selections - records, cd.
All this was done using the console's restored/modded tube amp of course.
And I'll be goll-damned if I could hear any difference, keeping my mind away from 'perception' of course.
I wound up using the mylar, simply because I knew it would outlast a 'lytic as far as stability and life.
So to me, a 4.7uF value is a 4.7uF value, regardless of what type it is.
And my trusty capacitor tester says the same darn thing - because it's not prone to "perception" issues like a human sometimes does. 😉
Bill, I dunno. That tweeter was just a vehicle to allow measurements of the cap. He could have done it just on a C-bridge, but by doing it on a tweeter he cunningly bypassed a lot of potential resistance ('we don't listen to C-bridges, we listen to tweeters').
You can dream up a lot of other situations in other crossovers, but do you really expect those caps all of a sudden start to show differences?
Not impossible, but I'd rather buy lottery tickets, more chance of success ;-)
Jan
As I READ the OPs question, I'm trying to read between the lines.
If the question is; Why do different capacitors SOUND different even though they measure very close?
The answer is; The same type of construction and materials will and different materials and types of construction won't.
Second; if you're going to test boutique capacitors, follow the manufactures suggestion for orientation and if they suggest 2 minutes or 1000 hours, and use some secrete way of mechanically joining the cap before using their super duper solder do what they say and USE the materials they suggest. Following the directions is part of the process.
I've pulled passive crossovers and replaced them with upgraded crossovers from the manufacture. The only
difference was the capacitor manufacture. Everything else was the same value. They not only sounded different. They took
2-300 hours before I could listen to them for long periods. They took time. TRT Teflons. Horrible things to listen to for 2-300 hours. I did it 8 different time with the same manufactures speakers. After the first time, I put the speakers face to face, blanketed them and let them play for a week. That was his suggestion. I went through a very simular experience with a
CJ preamp that was sold to me as junk AFTER an upgrade. I was patient, and did what the manufacture suggested.
If you did one speaker at a time the sound change was easy to hear. I had one as a center channel but it wasn't ready for pickup. I also use/used a Mcintosh C20/MC240 as the reference 2 channel system in my rooms. I don't always use them
but if I have the old brain fog, I go back and listen with that system. I was amazed at the difference in the center channel's sound vs the other two that were well over 350 hours by the time I picked the NEW center up. I hooked it up to the right and left the other speaker. Nice personal reference. I got away from AV or an analog center channel in the 90s.
Point being you can't install a cap and not wait. When a cap is as big as a soda can and you took out, one the size of a AA battery, you better be ready for some change. It's not in your head either. My rabbit won't stay in the room either.
If I'm not reading the question correctly, please enjoy your beverage, cheese and crackers, we will be landing shortly.
Regards
If the question is; Why do different capacitors SOUND different even though they measure very close?
The answer is; The same type of construction and materials will and different materials and types of construction won't.
Second; if you're going to test boutique capacitors, follow the manufactures suggestion for orientation and if they suggest 2 minutes or 1000 hours, and use some secrete way of mechanically joining the cap before using their super duper solder do what they say and USE the materials they suggest. Following the directions is part of the process.
I've pulled passive crossovers and replaced them with upgraded crossovers from the manufacture. The only
difference was the capacitor manufacture. Everything else was the same value. They not only sounded different. They took
2-300 hours before I could listen to them for long periods. They took time. TRT Teflons. Horrible things to listen to for 2-300 hours. I did it 8 different time with the same manufactures speakers. After the first time, I put the speakers face to face, blanketed them and let them play for a week. That was his suggestion. I went through a very simular experience with a
CJ preamp that was sold to me as junk AFTER an upgrade. I was patient, and did what the manufacture suggested.
If you did one speaker at a time the sound change was easy to hear. I had one as a center channel but it wasn't ready for pickup. I also use/used a Mcintosh C20/MC240 as the reference 2 channel system in my rooms. I don't always use them
but if I have the old brain fog, I go back and listen with that system. I was amazed at the difference in the center channel's sound vs the other two that were well over 350 hours by the time I picked the NEW center up. I hooked it up to the right and left the other speaker. Nice personal reference. I got away from AV or an analog center channel in the 90s.
Point being you can't install a cap and not wait. When a cap is as big as a soda can and you took out, one the size of a AA battery, you better be ready for some change. It's not in your head either. My rabbit won't stay in the room either.
If I'm not reading the question correctly, please enjoy your beverage, cheese and crackers, we will be landing shortly.
Regards
Anecdotes can be data. Perhaps consider trials of a new drug to treat some disease. Anecdotal reports of side-effects are tabulated and published because the information can be useful data for clinicians.
The first question to ask is: do they sound different if they measure very close?
It is very logical that people who spend $ 200 on a cap report that it sounds fantastic, but does it to anybody else who doesn't know the price or reputation?
Jan
My point was that, in this case the cap was almost acting as a DC blocker being a simple HPF and as Bateman showed, when there is little to no signal across the capacitor then distortions are almost immesurable whereas if there is a large signal across it then sensitive instruments can pick something up. Extending this I'd have been more impressed if he'd chucked a NPE cap in to compare.
But in general I expect boutique caps to measure worse than industrial grade...
Chutzpah is a wonderful Yiddish word meaning, roughly, "cheek" in English. Dreistigkeit , perhaps?
IIRC those measurements were also in the freq band below the roll off, so that's where there would be apprciable voltage across the cap.
Jan
I am with you that for a proper comparison values need to match closely to make sure you are not comparing two different corner frequencies. On the other hand all the e.g. Wima capacitors I have measured were much closer than 10%, even if rated as 10%. Compared to that "high end audio" capacitors often don't come with tolerance ratings at all or any datasheet.
Also there is a surprising spread in loudspeaker impedances from one unit to the next so these would have to be matched first, or the capacitor value matched to the driver's impedance and even then 5% difference doesn't have much of an effect.
If that effect is noticable by a relevant fraction of a test group in a DBT is unknown to me. Also there a re so many factors to that that it might be quite complex and confusing for most to state results.
Many moons ago I paid a lot of money for some highly regarded (by some...) Jensen paper in oil capacitors. I did a lot of "burn in" (Class D amp, pink noise, load resistor) but didn't like the sound. Going to polypropylene makes me a lot happier.
Different story, but still anecdotal.
this is a potential elephant in the room, were it not for the fact that the room has a much bigger effect on FR and is rarely symmetrical. But that is another debate for another time...
- Status
- Not open for further replies.