The lens after the knife is often a standard camera lens. If you have an optical table, we could look at your setup and determine the exact location the 50 mm lens needs to be placed. This will depends on the schlieren setup and sensor size.
To get a small luminous point, we use an LED (sometimes in form of a fiber light), a hive light, or other non-flickering light source. Note that an aperture can be added to make it a smaller point in all cases. For BOS schlieren, you can use highly luminous LED light panels such as the IES4438s or GS Vitec LEDS MX. The Nanlite may also be a good option for BOS.
Yes, the smaller the light source, the more refractive index sensitivity (RIS) you can achieve. You will be trading RIS for image brightness.
You could buy some color filter gels off Amazon.com that'll work fine, Edmund Optics has a book of over 200 Colored Filter swatches. Sometimes the plastic will soften or melt if too much light passes through, so colored glass filters are a good option too, like those from Thorlabs.com.
Yes, the smaller the light source, the more refractive index sensitivity (RIS) you can achieve. You will be trading RIS for image brightness.
In general, the recommended setup will depend of the constraints of test. Its possible a recommended liquid/liquid setup will differ from a gas/gas setup.
Depends on your application. Most applications can get away with the use of strong LED, but other applications that demand high-energy ultrashort exposure times, they may benefit from using a laser.
For BOS schlieren, you want to try your best to get the speckle pattern (of the background) and subject in focus. Sometimes this is difficult, so you may have to put up with soft edges of the subject. In order to be able to get good focus on both the background and subject, you will need to close the aperture down to extend the depth of field. This will decrease the brightness of your image, so additional backlights may be needed. Often times, for a 12ft test (from camera to backend), a pair of IES4438s work just fine. That is exactly what was used for the example where we showed our BOS - mask/coughing tests.
We have a case study on this topic. Coughing research through background oriented schlieren imaging.
Look into the academic literature by G Settles and M Hargather, they have published a series of articles of the past decade in this regard. One example is, "A comparison of three quantitative schlieren techniques."
You can certainly get a schlieren effect from transparent solids (since they serve as a refractive index gradient). When you have semi-transparent media, you will lose photons from absorption from the subject. While you probably could predict if your test will be successful/possible, I always encourage doing a test to see how things go!
I recommend you reach out to Vision Research applications team directly to discuss your exact setup/application.
Filters can be used to block specific spectral bands. You can look at Thorlabs or Edmund optics for application-specific filters.