Hi, thanks for the question. The purpose of sending the red photon of the first source (idler A) through the bottom crystal is to introduce indistinguishability between the idler photons that could come out of the bottom crystal. If the paths of the idlers are aligned, there is no way to know from which crystal it was produced, this means that the states are interfering with each other. This is the key concept of the scheme. The object introduces an interplay between distinguishability and indistinguishability of the red photons, which combined with the fact that there these are tightly correlated to the green photons (signal photons), allows the green photons to produce an image of the object, non-locally. Best, Andres Vega

@@castaway1889 thanks for the clarification. In your picture, you have a 50/50 splitter before the camera and only one camera to its right. In the Zeilinger paper, there are two cameras after the splitter, one for each output. Is that an oversight in your picture or is there actually just one camera? In case there are two cameras, is the image generated by recording only coincident photons? Regards

@@tictacX1 One camera is enough to record the image in either output of the beam splitter. Coincidences are not necessary to retrieve the image, as occurs in the case of ghost imaging. In the Zeilinger paper there is one camera in each output because if both recorded images are subtracted, it is possible to get rid of the background that does not carry information of the object. Cheers, Andres

@@castaway1889 do you have a paper that goes with the presentation? It's better I read it to get the details. Thanks

@@tictacX1 Lemos, Gabriela Barreto, et al. "Quantum imaging with undetected photons." Nature 512.7515 (2014): 409-412. Lahiri, Mayukh, et al. "Theory of quantum imaging with undetected photons." Physical Review A 92.1 (2015): 013832.