So shorting with the wrong type of expensive cap dulls the sound? Show me...
While I
really hate delving into this one, I tried this way back in the late 70's/early 80's with very mixed results. My boss heard of some brand of shorting caps and wanted to see if they were worth adding to the store's product line. He actually wanted to test the idea first.
I did not have various fancy caps to try so I used regular RCA plugs I shorted by filling with solder, using a small wire, or braided copper. I did not measure nor hear a difference among the various techniques, but there was a difference with or without them on the inputs in some cases. I tried both phono and line inputs; phono was usually more sensitive, natch, since gain is so much higher. I measured at tape outputs but generally went with the line output with the gain cranked and no signal present (input shorted or open; open was more sensitive, natch, but picked up noise with or without the neighbors shorted). In all cases I was looking for any difference in the output of the "victim" from the shorted or open "aggressor". I measured the pre-amp output using a spectrum analyzer (audio/RF) and 'scope. I did a listening test with small bookshelf speakers (LS3/5a's, I think). It was me and my boss, no listening panel, no DBT. I had the advantage of being able to test near a high-power AM/FM transmitter tower, and in a screen room. I do not remember all the components but it included several SS preamps and a couple of tube preamps.
IIRC I never tried shorting an output. Terminating an output in 10k ohms whilst driving it full-scale was an exercise in crosstalk measurement and thus uninteresting for this test. Outputs are low-impedance and I would not expect any change.
I never tried non-shorting dust caps. I have had them now and then but have never measured nor noticed any difference.
I have no record except my memory so don't ask for such. You are free to replicate the results yourself.
- It was very dependent upon the component. In-store/screen-room most showed virtually no change on the outputs using a spectrum analyzer, audio or RF, nor could any difference be heard. On some the output noise floor was noticeably lower, measured and heard. However, I don't think any of them was audible a few inches away. I suspect differences in layout and decoupling were at play.
- The difference heard varied and was well-correlated to the output spectrum. Some exhibited higher hiss, some more LF rumble. In some cases I suspect some ground issues; at high gain some motorboating (LF rumbling) was evident.
- In-store and in-screen room results were essentially the same. Near the transmitting tower more components exhibited measurable differences (don't recall doing audible tests there). The worse ones before were mostly much worse in a high-power RF field, but in a couple of cases I recall discovering poor RF rejection. However, a big caveat was that the victim (measured reference channel) in most all cases had much higher noise itself in this situation, rendering any improvement from shorting caps on the other inputs moot.
- One or two components actually did not like a dead short. I built a couple with 10-ohm resistors and in those cases they worked better. Again I suspect ground issues with those designs.
- The change was in those certain cases audible and measurable, but as soon as a signal was applied it became virtually impossible to determine any difference in output with or without shorting caps on the unused inputs, again with the exception of a couple of components iin the high-RF environment.
For the majority of components there no significant difference, measured or audible, but for a few there was a drop in the noise floor that was significant (3 - 10 dB). Many benefited from shorting inputs when placed in a high-RF field, not a huge surprise. However, the "live" input still suffered in that scenario so the benefit of shorting unused inputs was negligible in a practical sense.
Would I use them? Hmmm... I have not felt the need, though since I had them I did use them on my old preamp for a while. When I moved and lost them I never felt the need to find them again. In real life, any audio signal is going to mask (by orders of magnitude) any coupling from unused inputs so I find it very hard to justify on the basis of audible improvements. Changing the noise floor from -90 dB to -95 dB is not going to matter much to most... If I had them I would use them, if I thought I had a real problem I would get some (though I find that scenario unlikely), and no way would I pay much for them or for special metallurgy or whatever.
FWIWFM - Don