Normally I'd say supercardioid, but glockenspiel is kind of wide so a conventional cardioid might be better. The polar pattern is important, but exploiting relative distances is a big part of isolation, as is using gobos (isolating panels).

I wish I had taken pictures tonight as I was just doing this. In the room was a piano, an upright bass and an electric guitar. We used DIs on guitar and bass, a U87 on the bass and a Royer R-101 on the piano. We placed two gobos between the piano on one side and the bass and guitar on the other. The Royer was positioned so that its front lobe pointed at the piano, the back lobe pointed at the treated ceiling and the null aimed at the other instruments. The U87 was about a foot away from the bass. The bleed was almost inaudible all around.

It helped that the room is large, somewhat treated and not at all square. In a small room with right angles it can be harder to isolate the sounds.

I have a number of shotguns, and they're pretty unpleasant sounding on instruments, because the narrowness is at the top end, leaving the lower end more cardioid. This creates a change in the sound. On a piano, for example, shotguns sound very 'unreal'.

The secret to multi-miking is knowing your mics, and using their quirks to your advantage. It's not really a case of looking for a mic that minimises spill, it's orienting the mics to take advantage of their nulls and lobes. Sticking one sound source in the less responsive areas of a mics polar pattern on another source is the key. Hypers can be quite useful here, having nulls in more useful places than a cardioid.

Lkj, post: 455905, member: 47827 wrote: OK, so I am building a setup in my home studio, which would allow me to record mainly percussive acoustic instruments like shakers, glockenspiel, and various types of ethnic or homemade instruments, in a live situation. Since I already use two condensers for vocals in the same setup and also play electric guitars "in the box", I want this mic to minimally pick up on all of these.

So I really need a valuable opinion.... is a "shotgun" type condenser the best choice for this setup?

We usually use overhead small condenser mics for small hand percussions live (on stage)
I just want to be shure of what you mean by live. Will it be on a stage with a big PA ? Will those mics be powered through the PA or just for recording ?
If it's only in studio, are you room treated or not ?
What's the level of quality and expectations ?

Paul is quite right - as a rule, shotguns are not good microphones for acoustic instruments.

Your post implies that the "percussive acoustic instruments" would be being recorded at the same time as the other instruments you mention, so minimising bleed is a priority. What you haven't said is anything about the space in which you are recording. What are your studio dimensions (L x W x H)? How are the walls treated (acoustically)? What is on the floor? How are the performers distributed in the room?

Having done a fair amount of both live and studio PA + recording of non-drum percussive instruments being performed at the same time as other instruments and vocals, I have found that there are several things that you can try in difficult acoustics: (1) stereo recording, (2) close miking, (3) inventive microphone positioning. Normally, you would put a couple of condensers on stands pointing down at the instrument, but it's not easy to close-mic a wide instrument such as a xylophone without getting an unbalanced coverage. However, placing a couple of hypercardioid condensers in an A-B spaced configuration underneath the instrument pointing upwards slanted towards the performer can minimise the bleed, while giving you a sound that punches through a mix without having to be loud to do so.

Crucial in the recording of percussive instruments is the type of pre-amp used for the microphone(s). The very principle of percussive instruments implies a lot of energy in the high frequency range released in a short period. If you were recording tingsha bells, for example, you would need mics and pre-amps that could handle audio frequencies up to 30KHz, and then use 96KHz sampling rates to go with that. You might well say that we can't hear anything above 20KHz, but we've had several discussions in these forums about the distinction between steady-state hearing and transient hearing. I can't hear much these days above 15KHz as measured by steady-state hearing tests, but I can certainly tell the difference between 44.1KHz and 96KHz recordings of tingsha bells recorded wth the same mics and pre-amps.