Classical speakers that do violins well??

Cymbals have a lot of room reverb effect. I agree they would be very difficult. I suggest MrTennis try the Shefield Labs Drum Track to test for realism. Why mess with making your own recordings when Doug Sax has already made a good one (without the usual compression on drums that let you know that it is not real). Note that you will need extremely high peak SPL capability in order to reproduce the dynamics on this recording. The surprise of uncompressed drums is that they do not sound louder.....they sound softer! Yes you get way more peak SPL's but the transients (from stick impact) are so brief that it does not register as loud.

"...The surprise of uncompressed drums is that they do not sound louder.....they sound softer!"

I find this to be true of symphonic music as well. I never find myself wishing the music were louder, even though you can usually hear any shuffling around or other quiet sounds by people near you. The transient peaks of instruments (I'm particularly thinking pizzicato violins, as well as percussion)is loud enough to catch your attention and be exciting. To get this same excitement at home out of a compressed recording, you have to turn the volume of the rest of the music up more than is natural.

you can't use a recording without a reference, if you want to assess the "accuracy" of a stereo system.

i still maintain that the difference between real and recording would be obvious to an experienced listener.

dave, would you be up for a wager ?

if you lived near by, i would identify the source, namely either your trumept playing or a recording of your trumpet played through your stereo system.

i would bet $100 that i could tell the difference.
i have two advantages, namely the dispersion pattern of the trumpet, vs the dispersion of the sound emanating from the speakers and the difference in spl, between the recording and your playing the trumpet.

i still maintain that one would want to find a speaker which minimizes errors in reproducing timbre. regardless of the instrument, certain driver technologies are more likely to be effective in this regard. cones are not the answer. i would look for a ;light and fast driver, such as a ribbon and electrostat. in fact, i am looking for a speaker myself.
i started my search yesterday with an audition of the analysis audio omega. i do not plan to audition cone designs, but may consider a hybrid, namely the piega tc 70x, which has a ribbon mid and tweeter and 2 8 inch drivers. the cones are crossed over at 200 hz.
i am a little concerned about coherence, but i am willing to listen, noetheless.

I never said that you couldn't detect the difference between a live instrument and a recorded instrument. I'd get that right 100% of the time in most rooms. The room cues are just too numerous. However, speakers can easily get the timbre right and you don't need an A-B comparison to judge that. I think this is where we started, talking about timbre and not room acoustics:

02-04-08: Mrtennis said:
"in order to do a definitive test of a speaker's ability to reproduce timbre accurately, it is necessary to record an instrument in one's living room and compare the recording to a musician's presentation of the same music."

Maybe it's because I'm a musician, but my aural memory is sufficient to know a clarinet when I hear it. The timbre is intact in my room, even if the clarinet was recorded in an isolation booth.

Of course my musician's perspective may be distorted from the average audiophile's perspective. I'll grant that. I've heard live clarinet literally thousands of times, up close and far away, through mics and acoustic, with soft reeds and hard reeds, etc., etc. I suppose that many audiophiles haven't really heard live clarinet, or only from the orchestra seats a few hundred times, or less. That's a lot of difference in exposure.

That's just thinking about clarinet. On trumpet I've heard world class players from 3-feet and 300-feet, playing every possible combination of trumpet type. Excluding hearing myself, I've heard live trumpet ten of thousands of times.

Somewhere up the thread I think I suggested to ask a musician over to listen for timbre. I still think that's a realistic option and much easier than your proposed test. Given the difficulty of recording an acoustic instrument in a space where we also have a system set up, it's just an impractical test of a system.

Still, this is an interesting discussion and I appreciate your view.

Dave

as much as a musician has exposure to an instrument, either playing it or listening to another musician playing, aural memory is the issue. it is not reliable.

so, you need a reference. a memory is not a reference. it is a collection of experiences stored in the brain.
i would not rely on someone's memory as to what an instrument sounds like. i would prefer to hear the instrument.

if the instrument is not available, preference and opinion takes over.