[moved] Optical interference

Interference, on which sense ?
Both devices are emitters, not sensors.

Interference occurs with coherent light, e.g. laser light. It's effectively irrelevant for LEDs, either when using one or multiple LEDs.

Interference effects can be still observed when non-coherent light interacts with dielectric layers of wavelength dimension, e.g. the "dichroic" colors of a soap bubble.

Ok. I have mixed two lights in an integrant sphere couple with an optical fiber and i have placed a camera (which is a linear sensor) at the end of the optical fiber for readout, usually the resultant power must be sum of the individual power, but it is not the case here, i have a power leak up to 10%. At low power i had good results, however once i increase the LEDs power this phenomena occurs in a destructive sense.

Consider non linear detector characteristic.

No, it is not possible. To see any interference you have to use (more or less) coherent light. Light from two sources like LEDs will not be coherent at all. Although it is possible to see interference pattern just by using light from one LED. You can build Michelson-Morley interferometer and use LED as a light source. But then path length of two arms must be equal. It is possible to see such a pattern but you must pay an attention to balance two arms of the interferometer. Further details in google.com under "white light interferometer" and "white light interferometry".