|
Anti-bloom chips have added circuitry to each pixel that siphons off captured photons as the pixel approaches saturation which if left unaddressed will eventually cause the pixel to over flow into neighbouring pixels in the same column (and very well demonstrated with your posted image). Anti-bloom technology tends to be a desirable feature for astrophotography taken for aesthetic purposes - or night-time security cameras.
However, for those who wish to perform photometry - for example - it has the undesirable effect that the pixel value may not be linearly proportional to the number of photons actually landing on that pixel.
If you have two stars in an image and one is brighter than the other you could take a sequence of images and the ADU count will double for each star for each time you doubled the exposure. But with anti-blloom - there will come a point where the brighter star will not double its ADU count and instead of displaying a straight line graph (ADUs vs. exposure) will start to droop.
The point where the non-linearity starts will depend on the chip (and actually kicks in gradually) thus there is a significant part of the ADU scale (at the upper end) where scientific measurement is very unsafe.
There may also be the disadvantage (and I'd appreciate other's comments on this point) that the anti-bloom circuitry will take up real-estate space that might otherwise be made available to the pixel itself - so I would expect that the full-well capacity (max number of electrons that can be collected) to be significantly smaller (mainly due to electrons being siphoned off - but also possibly due to smaller active pixel area) and reduced sensitivity (smaller active pixel area).
I don't strictly know that with anti-blooming that the active pixel area is measurably smaller - but I do suspect that it has to be.
It's a good question - look forward to any other comments.
Mark C.
|