I haven't made up my mind about using the calibration target. http://www.neatimage.com/testtarget.html
I had one printed 11 x 14 and mounted on foam-core. Calibrated a well lit shot, and the auto-calibration didn't get any of the upper or lower values.
I emailed tech support on this, wondering if I should have shot some frames a few stops above and below, so that I'd have some samples for the upper and lower values?
They were against this, said to let NeatVideo extrapolate the curve from the existing data points - and this would be better.
Another thought - as the calibration target squares are solid, they probably compress quite well, and as the goal is to calibrate for noise ... I'm not sure it's going to correspond to actual noise generated by the sensor/compression combination when shooting a real-life scene, especially in low light or in a complex scene.
Can anyone shed light on the best NeatVideo workflow for calibration, etc.?
BTW - they also mentioned that you should shoot the target each time you shoot, in that lighting, shutter, etc. It doesn't seem you can create and save a calibration, and then re-use it - unless it's for an identical scene.
I also noticed that NeatVideo behaves differently in different programs.
In AE (or maybe it was in Premiere - or both?), it would say there was 'clipping in the X channel' quite often when dragging a rectangle. NeatVideo said this wasn't' really clipping, but not enough noise.
In Eyeon Fusion, it didn't object in the same way (I guess it found enough noise).
Great software, but ... I haven't learned the best way to use it yet.



