I am a beginner in the field of wavefront shaping and I was attracted by your paper of LC-SLM-based transmission measurement.

Checking your paper ‘ controlling light through optical disordered media: transmission matrix approach’, the region of SLM is divided into two parts: one is reference part and the other is controlled part.

So my questions are:

1) the measurement is proceeded by sending the hadamard basis on the controlled part while the reference part remain unchanged and then get the signal from CCD. So once the measurement is completed, equation 7 can give us the observed transmission matrix directly, right? But the intensities are obtained from hadamard basis instead of canonical basis. So the equation 7 maybe is not for the hadamard basis so that a unitary transformation is needed? If so, how to reconstruce the transmission matrix by getting the signal generated by hadamard basis?

2) if I want to optimise for optical focusing, following equation 16 can calculate the corresponding optimised phase pattern, so is the phase pattern displayed on the controlled part of the SLM directly? if so, will there be a problem leaving the reference part unchanged? I am concern about that the reference part can still reflect light into the scattering medium, which will interference with the optimised phase pattern on the controlled part.

looking forward to your reply.

Hi Jian,

1) Already replied in another thread. You measure the matrix in the basis you use for the excitation. If you use a basis different from the canonical one, you need to perform a change of basis afterward.

2) Indeed, leaving the reference will give an addition background to the focusing, hence reducing the signal to noise ratio. The reference is useful for measuring the phase, once you have measured the matrix, you do not need it. If you can remove it, you would get better focusing without it.

With a phase only modulator, you cannot directly switch of the pixel. But you can for example send a high spatial frequency grating (sequence of 0 and pi phase) on the reference part that would be filtered out in the Fourier plane to get rid of it.

Best,

Sebastien

Hi Sabastien,

Thanks a lot for your advices.

And I have realized it experimentally.

Thanks