Sorry that I wasn't able to respond earlier.
If you wish to estimate the brightness of a star as measured by a CCD or CMOS camera, you need to calculate the number of photons from the star which will be collected by your telescope and detector. Let's leave the Earth's atmosphere out of the problem for a moment. The steps I describe below are also listed, with illustrations, at
http://spiff.rit.edu...arris.html#gory
First, you need to find the spectrum of the star in question with absolute calibration. There aren't many, so it's likely you won't find one for your star. A good place to find some is in the CALSPEC library:
https://www.stsci.ed...atalogs/calspec
These spectra have a long list of values of (wavelength, flux at that wavelength), where the flux is ergs per sq.cm. per second per Angstrom. I suggest choosing Vega (alpha Lyr) as a starting example.
Next, you need the passband for your instrument; that is, the efficiency with which it turns photons into counts in the detector, as a function of wavelength. In theory, this should be a combination of the reflectivity of your optics, the transmission of the filter, and the QE of the detector, but many people just grab the transmission of the filter and call it "good enough".
Now it's time to convolve the stellar spectrum with the passband. As shown in the URL I provided at the beginning of this post, the idea is to break the spectrum into a series of narrow chunks, multiply the spectrum's value in each chunk by the passband's efficiency in the chunk, and so calculate the flux in energy, inside this chunk, that should reach your detector. Then, convert that energy into a number of photons, using the energy per photon inside this chunk. Add up all those photons over all the little chunks over all wavelengths.
The result will be a flux in photons per second per collecting area of your telescope, for that model spectrum. Write this down. If you were to observe that star through a telescope far above the Earth's atmosphere, this is the number of photons per second per sq.cm. you should record. For the example of Vega, using a Bessell V-band passband, the result is about 884,000 photons per second per sq.cm. If you were to use a 6-inch telescope, which has a collecting area of 182 sq.cm., you'd collect about 160 million photons per second.
But what if you're not observing Vega? What if you're planning to observe, say, HD 20291, which is the same spectral type (A0), but has a magnitude of V=7.0 instead of V=0.0? Use the definition of magnitudes to determine that a star of mag 7 should have a brightness which is only 10^(-0.4*7.0) = 1/630 = 0.00159 times that of a star of mag 0. So, instead of observing 160 million photons each second, you'll only see about 0.25 million = 250,000 photons per second. In order to get the best results for this step, your desired target should be the same spectral type as the model spectrum used earlier.
Okay, but in real life, most of us observe from the ground. We won't see 250,000 photons per second from HD 20291, because the Earth's atmosphere will scatter some of them. Suppose we observe at an airmass of 1.2 in the V-passband. Then we can expect
(1.2 airmass) * (0.2 mag/airmass in V-band) = 0.24 mag of extinction
In this case, the star will be fainter by 0.24 mag = 10^(-0.4*0.24) = 0.80 in intensity. So, instead of recording 250,000 photons per second with our telescope, we'll only record about 200,000 photons per second.
