Generally spectrofluorometers use high intensity light sources to bombard a sample with as many photons as possible. This allows for the maximum number of molecules to be in the excited state at any one point in time. The light is either passed through a filter or Monochromator which allows you to select a wavelength of interest to use as the exciting light. The emission is collected at 90 degrees to the exciting light. The emission too is either passed through a filter or a monochromater before being detected by a PMT, photodiode, or CCD detector. The signal can either be processed as a digital or analog output. Systems vary greatly and a few things must be considered when choosing. First is signal to noise, there are many ways to look at the signal to noise of a given system but the accepted standard is by using water Raman. It has been said many times however tell me what signal to noise ratio you want and I can get it for you. This is true you can cheat on your signal to noise ratio. People do this routinely by using filters and various slit widths and never publish these settings. Make sure you compare apples with apples on this, ask for a specific band pass and no filters with 1 second integration time and use the same wavelength to compare too. Sensitivity or detection limit is another spec to look at. How little light can they measure, this is important because you may not want to waste a large amount of a specific expensive sample. Well if you can get data with very very little in the solution then great. So the standard for this would be fluorescein in NaOH, typical values for a high end instrument are in the femtomolar range. Stray light is another big issue in these instruments. Straylight is basically how good is the monochromater of a system performing. You will get some stray light bouncing around in side the monochromator compartment and it comes through to the sample too. This matters when you have a highly scattering sample, however one can always use an excitation wavelength further away from the emission band to negate this issue or use a laser or interference filter. Stray light plays a big role in the signal to noise ratio and detection limit thus some people may have what seems to be bad stray light characteristics but they have better signal to noise ratio and detection limit, this is all in the design of the system. Thus stray light is a much less important measure than the others because if you cannot see the signal for the noise, cannot detect the sample, or have to use too much it will cost you more to buy one that you can use in the end.
These systems come with many options now here are a few:
Polarizers Peltiers Cryostats Cold Finger Dewars Lasers for lifetime measurements LED's for lifetimes Filter holders Adjustable optics (very important) Solid sample holders Slide holders Integrating spheres NIR detetors Ruled gratings Blazed gratings Bilateral slits Manual slits Computer controlled slits Fast switching monochromaters Filter Wheels The list goes on forever