While I don't qualify as a speaker-tweaking expert, I have designed and built two studio monitor systems from scratch I've been quite pleased with; I gave one pair to a musician friend for his home studio and he and his fellow geetar pickers were impressed enough to request two more pairs. My main interest, however, is rebuilding and upgrading older commercial systems with modernized drivers and crossovers--vintage Klipsch, Polk, Celestion and Advents to date.
There's a ton of data out there regarding the benefits of bypassing large-value crossover caps (I mostly use Solens, though I plan to go with Black Gate bi-polar 'lytics in my next project) with small-value audiophiliac metallized film and film/foil caps. There are certainly many manufacturers willing to sell high-priced caps specifically for this purpose, and reams of published data that demonstrates their superior performance, at least on paper.
Being a trusting and gullible type by nature, for my last two projects--rebuilding the crossovers for my Klipsch KG4's and Polk RTA11-TL's--I bought a supply of Audiocap PPT Theta film/foil caps to bypass, using values equivalent to 1% of the base caps. The results were a mixed bag; the Klipsch's were completely transformed (I also replaced the original phenolic dome/voice coil assemblies in the horns with Klipsch's newer and better polymer domes). I heard detail I had never heard before and at first was blown away--However I learned that the benefits were dependent on the source material. With CD's such as Alison Krauss' "Union Station Live" the Klipsch's really excelled--her soprano voice was transfixing and the speakers were smooth and comfortable throughout the range. However when I switched to other material, such as Lynard Skynnard and Stevie Ray Vaughn, the material sounded harsher and somewhat irritating. These are AAD CD's, remastered from original analog tapes, and too much high-end detail is not a good thing. I removed the bypass caps and after further listening decided to leave 'em out of the circuit. Although I miss the marvelous detail in some recordings the benefits seemed directly related to the quality of the source material.
For my tests I used CD's for comparison, although probably half my listening is on vinyl. This is probably a whole different subject, since the dynamic range and frequency response is quite different from CD.
The Polk RTA's noticeably suffered from the bypasses. This is partly due to the SL-3000 metal-dome tweeters, which I'll soon be replacing with Polk's newer silk domes, but I concluded that added detail notwithstanding the bypasses gave me a listening headache.
I've also got a pair of Celestion SL-12's, which were my primaries before I modded the Klipsch's, and in A/B listening they sound rather muddy and lacking in HF detail. With this particular speaker I think bypassing may make a positive difference.
Anyhow, I've concluded that the benefits of tweeter crossover bypassing is largely dependent on the specific speaker's characteristics and also the source material.
I wanted to see what advice you experts have to offer; I'm also interested in recommendations regarding bypassing the mid/bass crossover sections. Bypass caps are used in amplifier circuits specifically for their ability to transmit high frequency signals--cheaper caps (especially electrolytics) have lousy high frequency response. Is it worth bypassing the large woofer/midrange caps, or are the audio frequencies involved largely invisible to these drivers anyway? By adding bypasses in these positions, will I also be passing on the sonic glare I wish to avoid to the HF section of the crossover?
Thanks for any opinions y'all have to offer.
B&K ST-1400 Series II power amp
Parasound P/HP-850 Pre
Adcom GCD-575 CD player
Linn Axis turntable w/Rega RB-300 Cardas-wired tonearm
Goldring 1042 MM cartridge
Klipsch KG4's, Celestion SL-12's, Polk RTA-11TL's