I recently removed an optical low pass filter from my digital camera. Do you have any suggestions on how to measure its Modulation Transfer Function (MTF)?

There are a few ways to measure Modulation Transfer Function (MTF), but it is very difficult to do so with any precision on a flat (plano) optic such as a window or filter. Measurements for MTF are generally done on systems by imaging a target of known contrast and known size at a known magnification and measuring the resultant contrast and size. In practice, this can be done by imaging a point, a bar target, a sinusoidal target, or any random target. There are many ways to test MTF but the reason none are appropriate for windows or filters is because they are all measurements of an image, whereas, windows and filters don't form images. One could determine the MTF of a window or filter by testing an optical system to use as a baseline, then inserting the window or filter into the optical path and re-measuring. The MTF of the window would then be the second result divided by the baseline result. The problem with doing this is that the MTF of the window or filter would almost certainly be within the uncertainty of the measurement. Even low quality windows and filters have very good MTF. Because MTF isn't very telling for windows or filters, their ability to transmit an image is usually given in terms of transmitted wavefront distortion. Rather than the error in contrast as measured in MTF, transmitted wavefront distortion measures the displacement of a theoretically perfect wavefront as it passes through an optic. Measuring this requires an interferometer or similar device. For example, a Schlieren System would help to visualize slight wavefront variations, but couldn't help measure them easily. Whatever test one does on a window or filter, there is very little chance that it could have an appreciable effect on any camera-based system's image.