This is something a lot of people try to do from time to time.
This is a good one.
I also built a mount for a fairly large HD camcorder - pic here. That one used two "double cup" mounts which I got on Amazon, called "Delkin Fat Gecko Camera Mount". The current URL is here.
These suckers are not exactly dirt cheap but they are very good.
The problem with mounting on the airframe, and particularly on the front window, is vibration. This will - at best - cause a blurred image and is likely to play havoc with the hand motion stabilisation of the camera which is better to turn off if possible.
As can be seen in the Photobucket images, I used some silicone rubber spacers. They came from here. At about £15 each they are not cheap but they are very effective. In fact the image quality is fully HD and limited only by the condition of the window.
There are professional solutions of course but they are bulky because they support the camera in a "frame" which is attached (via anti-vib mounts) both top and bottom, which stops any twisting motion, and any motion which is purely translation side to side will barely show up on the image anyway.
I should add that it is a good idea to add a piece of cord or a lanyard, which will catch the whole thing should it fall off the window, otherwise it could smash onto something important 
I found a camera mount with a large butterfly type grip for £7 in Maplin, UK. It grips nicely to the top of the instrument panel and holds my Canon DSLR still enough to capture smooth take off and landing videos facing out towards the nose. Although its a sting grip, it's about 3 inches wide and doesn't seem to dig in and leave any marks. I can't find a pic on the Maplin website but for £7 it was a bargain, and sucker mounts are too risky for a large DSLR
If I may offer a bit of advice, the vibration dynamics of mounting imaging devices (or anything else for that matter), does need a little thought to avoid the shake problem. A rule of thumb for minimum shake is small/light camera on a stiff mounting (and that includes the surface its mounted to).
Essentially the device on its mounting will have resonant modes of vibration in six degrees of freedom (3 translations and 3 rotations) - which will occur at a frequency controlled by the Sq Root of its mass over its mounting stiffness.
These 'modes' are then excited by vibration sources such as the engine and prop rotational speeds and any other vibratory modes of the airframe its is attached to.
And the worst case occurs when this frequency is the same as the shutter rate - i.e. one frame opens with the camera viewing once scene then closes and opens again with the camera looking at a slightly different scene because the camera has moved [e.g to one side] - resulting in image distortion.
Further effects take place to do with frequency differences between e.g blade passing and shutter speed . . .
Would be pleased to offer further mitigation steps if interested
A rule of thumb for minimum shake is small/light camera on a stiff mounting (and that includes the surface its mounted to).
I think that depends on what the vibration is, and the camera. I agree the above statement holds if one wants to avoid shake, but it is the worst scenario for transmission of vibration, the removal of which needs the opposite i.e. a heavy camera on soft supports.
In my plane, the vibration (analysed here) consists of some fast stuff and some slow stuff. Interestingly there is a lot of stuff at much less than the crank speed, presumably because there are 3 combustion events per revolution.
The front window is also getting a lot of aggressive prop airstream buffeting which makes it an additional challenge to mount anything on it. The instrument panels are also rubber-mounted and they vibrate along happily at some lower frequencies of a few Hz...
It gets damn hard to get rid of the slow stuff. That's why I used the extra brass weights on my camcorder mount (and the camcorder is heavy too) in conjunction with very soft silicone supports.
As far as I've managed to find out, the general trend in professional anti-vibration mounts is to suspend the camera on a "framework" made up of a thick elastic band, inside an outer frame. At the top end of the market they have gyro stabilisation as well.
I can imagine that some of the little cameras (the Go Pro sort) can work despite being mounted on a badly vibrating surface - because they have no moving parts. The "shutter" is electronic, there is no iris, and there is no motion sensing or stabilisation (or it can be turned off). But camcorders seem to suffer very badly from almost any vibration so one has to really just get rid of virtually all of it...
I started with a Fuji FInepix in cine mode, suspended from the centre bar where the cockpit clamshell doors meet. I moved on to a Canon HG10, which was a disaster - I eventually got it vibration free enough for the HD not to shut down, and switched off image stabilisation, but there were other problems. I 'm now using a Panasonic Lumix FT2 on cine mode, mounted on a tripod, on a frame, fixed to my flight bag, which is secured by the straps to the RH seat. I tried foam and lead in various mounts, but vibration has increased on our Jodel DR1050, and I gave up on them. I viewed Peter's mount with interest. The Fuji wouldn't work at the low winter cockpit temperatures - when taken out from a warm inside pocket it quickly stopped. The Lumix is cold resistant. (We've checked the prop - that doesn't seem to be the vibration source. Engine coweling is suspect.)
We've checked the prop - that doesn't seem to be the vibration source. Engine cowling is suspect
A lot of work has been done on this by Socata.
Very little has been documented but IIRC one thing they found was that engine vibration could be transmitted to the aircraft via engine parts (including the exhaust and hoses) touching the cowlings.
Also the type of baffle material is important. I have just re-done a part of mine using this stuff (their video is worth watching) and will re-do the rest during the Annual in January. I got a good CHT drop but an improvement in vibration would have to be measured with another vibration test flight...
But I think piston singles just vibrate badly anyway The only piston plane I have ever flown in which didn't vibrate was a Cessna 421C, which was smooth as a train.
"I think that depends on what the vibration is, and the camera. I agree the above statement holds if one wants to avoid shake, but it is the worst scenario for transmission of vibration, the removal of which needs the opposite i.e. a heavy camera on soft supports."
I agree that mass-damping and energy loss through 'working' elastomeric mounts can be beneficial, -my 'rule of thumb' statement is highly generic. More damping has the effect of reducing the level of the peak vibration but at the expense of spreading the frequency range over which a resonance occurs. My thoughts on inflight cameras are to generally try and use the smallest possible camera for both minimum vibration and airflow disturbance reasons. Every camera/mounting/airframe combination is certainly going to be unique and the deployment of specialised equipment to troubleshoot is very sensible, its just not a simple problem.
"In my plane, the vibration (analysed here) consists of some fast stuff and some slow stuff. Interestingly there is a lot of stuff at much less than the crank speed, presumably because there are 3 combustion events per revolution."
You must have a six-cylinder engine then to see strong 3rd-order vibration - which as you say is due to three firing pulses per crank rotation. A single blade out-of-balance statically, will produce first-order vibration (assuming no reduction gearing from crank speed) and a further 3rd order contribution will come from the blade-passing pressure pulses - if you have a three-bladed prop. This would be 4th order if you have 4 blades, 2nd order if you have 2.
The half orders are probably due to differences in the actual pressure levels that each cylinder sees during detonation - which are often subtly different. Diesel engines are usually much worse, though active anti-knock controls on modern (e.g. car) engines usually suppress this effect these days.
PS: the most successful camera arrangement I've used so far is to mount my Muvi-Pro onto my headset. The human head seems to be good at basic stabilisation plus the interest factor of seeing a changing view every few seconds does seem to make the resulting movie a bit more realistic.
airflow disturbance reasons.
Sure; I would never consider using that anti-vib mount outside the cockpit
A small camera, securely mounted far out on a wing, should be relatively vibration-free.
"A small camera, securely mounted far out on a wing, should be relatively vibration-free." Agree.
My previous post: " . . . which will occur at a frequency controlled by the Sq Root of its mass over its mounting stiffness."
Oops - I should have said Sq root of the stiffness divided by the mass in motion of course . . . and to be completely correct the resulting frequency is expressed in radians per second instead of Hz - a simple 2xPi conversion to get Hz though.
A small camera, securely mounted far out on a wing, should be relatively vibration-free.
I have been toying with an idea to replace one of the underwing inspection hatches, installing one with a glass window and a remotely controlled camera behind it. It won't affect the airfoil, so no flight testing should be required.
