autodev query

[Lawrence]

During the wipe process, autodev can be set to 'temp'; this lets us quickly see whether a process will be improved by a later autodev. However, I frequently find that when I actually do autodev for the second time after wipe, the result is totally different. I often stop at the first stage because the second autodev over-emphasises the background noise. Am I missing something here because the 'temp' autodev looks potentially good.

regards

Lawrence Harris

[admin]

Hi Lawrence,

The temporary AutoDev in Wipe purposefully stretches a bit less, as a signal with any bias in it (usually light pollution) makes it hard to distinguish the "real" signal on top if stretched too much. It is more a visualisation/courtesy tool in this instance.
The second AutoDev (e.g. proper module) uses an sRGB curve on top of the linear result from AutoDev. It is the "proper" result. On a well calibrated monitor, with good data, it should achieve an optimal result for your chosen Region of Interest. Noise may indeed throw a spanner in the works in severe cases though.

If you like the AutoDev from the Wipe module, simply apply the result from the AutoDev module and then apply a gamma correction of about 0.45 in the Develop module (don't redo global stretch in the Develop module - we just want to correct the output from AutoDev). The 0.45 gamma correct will undo the sRGB stretch.

Hope this helps,