In case you are using the relative methode in dilute solutions from Parker & Rees (C. A. Parker, W. T. Rees. Analyst 85, 587 (1960) ), the following holds true:
With respect to optical density, it is constant up to fairly high values, but only if corrected for inner filter effects. With respect to wavelength, it is also constant over the complete range (think about Kasha rule...), but only if you correct the fluorescence spectra for excitation intensity.
Please us perchloric acid als solvent, as irregularities have been observed with quinine sulphate dissolved in sulfuric acid.
In general, the 1971 review on fluorescence quantum yield determination from Crosby and Demas is still worth reading:
According to book C.A. Parker Photoluminescence of solutions Quinine sulphate has very large concentration range in which the fluorescence signal can be linear.This possibility is a result of weak energy transfer between excited and unexcited molecules that realized only at very small distances. In this case a decrease of signal due to difusing self-quenching and energy transfer can be realized at concentration of quinine about 0.001 M. So, I think that the signal could be linear in the range 10(-6) - 10(-3) M or 0.1 - 400 microgram/mL.
you can use 0.1M solution of quinine sulfate (QS) as standard for Q.Y calculations for the samples having emission wavelength ~400-600nm. I think it will be good for micro molar concentration of the sample but you have to fixed the absorbance of the QS solution 0.1 at the particular excitation wavelength ( what ever may be 300, 310,330.....etc) and same sample you have to use for fluorescence measurement.
for details please refer J. Phys. Chem., 1961, 65, 229