Eugenia, to measure the luminance of an OLED you should really use an integrating sphere, this is the standard method, there are many papers showing how this should be done, but note the edges of the glass substrate should be covered so that only the forward direction light is measured not the waveguided light from the substrate, as then if you make the back calculation for internal efficiency you very badly over estimate. If you can not get access to a sphere the second best option is to have a detector as big as possible compared to the OLED and place the detector right on top of the OLED, this reduces many errors in trying to calculate the effect solid angle your detector is sampling.

Are you trying to measure an LED or an OLED?

It seems that the the surface area inside a closed curve, within which the active zone is located, should appear in the denominator. For a LED, for example, this surface area is aperture area of the light flux from the output window.

For luminance measurements it's better, if the sample area is bigger then detection area. But you still can estimate the luminance. It is necessary to multiply luminance, measured by the photodiode to ratio Sdetect/Ssample, where Sdetect and Ssample are detection and sample area. Sdetect is area in the "wisible" distance (look file). If you don't know Sdetect, then measure your sample in direct contact and take Sdetect as input area of diode, or aperture of diaphragm (if it's installed). Such estimation works only if you don't use the focusing lenses.

Thank you all for your comments. Unfortunately, the detector is 4 cm^2 bigger than the sample and it's the only one that I have, so I need to find out how to perform accurate measurements with it.

Andy, I am measuring OLEDS which have a planar surface.

Ruslan, the photodiode gives me the generated photocurrent in Amps, so I have to divide it by the area (that of the detector or that of the sample?). I understand from your answer that you suggest to "correct" the luminance values by the area ratio between the detector and the sample, isn't it? So if I install the photodetector in direct contact to the sample, then the area that I am looking for is that of the sample, right?

it's not right. In direct contact to the sample, area that you are looking is area of diaphragm (but only if it is equal or less of the sample area). Photodiode with diaphragm or lense detects light in solid angle. In the case when sample area is bigger than detected area, diode measured lignt from only solid angle. In the case, when sample area less than detected area, diode will measure less light, so you have to place diaphragm with small aperture in some distance from diode and recalibrate your detector (because the solid angle is changed).

Eugenia, to measure the luminance of an OLED you should really use an integrating sphere, this is the standard method, there are many papers showing how this should be done, but note the edges of the glass substrate should be covered so that only the forward direction light is measured not the waveguided light from the substrate, as then if you make the back calculation for internal efficiency you very badly over estimate. If you can not get access to a sphere the second best option is to have a detector as big as possible compared to the OLED and place the detector right on top of the OLED, this reduces many errors in trying to calculate the effect solid angle your detector is sampling.

Thanks Ruslan, your explanation is very clear.

Sincer my photodetector is a mounted silicon photodiode without diaphragm, can I assume that, in direct contact, the area that I am looking is that of the sample?

Yes, but this is inaccurate method. As said Andy Monkman it's better to use integrating sphere. We use in practice CCD-spectrometer with fiberoptic and collimating lense. We made radiance calibration and nonlinear correction. Also I tried sphere and Konica Minolta LS110. Results was identical and I stay in variant with optical fiber and collimating lense because it's very compact and more sensitive then sphere. Some OLEDs in threshold and high voltage has different spectra and we have to register weak signals

Thank you all for your contributions!

by multi-apertures:

reference Luminance-to-intensity Measurement Method