Difference between revisions of "Hugin Exposure tab"
(tweak) |
(correction) |
||
Line 24: | Line 24: | ||
The exposures for all images (except for the anchor image determined by selecting | The exposures for all images (except for the anchor image determined by selecting | ||
'''Anchor this image for exposure''' in the [[hugin Images tab]]) are optimised | '''Anchor this image for exposure''' in the [[hugin Images tab]]) are optimised | ||
− | (TODO actually the anchor is too?) | + | (TODO actually the anchor is too?). |
− | |||
− | |||
− | |||
=== Low dynamic range, variable white balance === | === Low dynamic range, variable white balance === | ||
Line 35: | Line 32: | ||
Similarly to above, the exposures and white balance is optimised for all images except the | Similarly to above, the exposures and white balance is optimised for all images except the | ||
− | ''anchor'' | + | ''anchor''. |
− | |||
− | |||
=== High dynamic range, fixed exposure === | === High dynamic range, fixed exposure === |
Revision as of 22:41, 11 June 2007
hugin has a brightness and colour correction system which is completely independent of the old PTStitcher tool. This improved system is only available with the default nona stitching tool in the hugin Stitcher tab.
The hugin system works by sampling a spread of points for each pair of overlapping images. The optimiser then tries to model a system of camera response curve, exposure, white balance and vignetting that fits the values of these points.
So for this to work, the photos in the project have to be already aligned. Align photos by managing control points in the hugin Control Points tab and optimising geometric image parameters in the hugin Optimizer tab.
Optimize Preset
Use the Optimize Preset combo box to pick one of several pre-set photometric optimisation schemes, then click the Optimize now! button to calculate the best available brightness and colour adjustments.
Low dynamic range
This will optimise vignetting, the camera response curve and exposure (EV) for all the photos in your project.
The exposures for all images (except for the anchor image determined by selecting Anchor this image for exposure in the hugin Images tab) are optimised (TODO actually the anchor is too?).
Low dynamic range, variable white balance
This will optimise vignetting, the camera response curve, exposure (EV) and white balance for all the photos in your project.
Similarly to above, the exposures and white balance is optimised for all images except the anchor.
High dynamic range, fixed exposure
High dynamic range, variable white balance, fixed exposure
Custom parameters below
Image Variables
TODO
Exposure
TODO
White balance
TODO
Camera and Lens variables
The Camera and Lens variables are the photometric analog of the geometric lens correction model, hugin assumes that all input photos with the same lens number have identical values unless they are unlinked in the hugin Camera and Lens tab.
Vignettting
Vignetting is dependent mainly on your lens and the aperture. Usually the centre of the image is brighter with a falloff towards the edges, hugin can calculate this falloff curve as part of the photometric optimisation process or you can enter it manually in the hugin Camera and Lens tab as the three numbers shown here.
Vignetting Centre
The centre of vignetting is rarely the exact centre of the photo, hugin can optimise this position or you can enter it manually in the hugin Camera and Lens tab. The scale is in pixels, with 0,0 indicating the centre of the photo (TODO is this relative to the d & e parameters?)
Camera Response
hugin can optimise the camera response curve by comparing differences between overlapping images. To do this your photos need to either have significant vignetting or have variable exposure. If your photos have perfectly even exposure and zero vignetting, then you would have to calibrate the camera response separately and then enter it manually in the hugin Camera and Lens tab.
The camera response curve is used both for mapping the images to a linear colourspace when creating HDR output, and for normalising the colourspace for internal vignetting, brightness and colour corrections when creating 'normal' LDR output. If your pictures don't require such corrections then you don't really need a calibrated response curve.
Hugin uses the EMoR response model from the Computer Vision Lab at Columbia University which simplifies the full response curve to five empirical coefficient numbers.