by Rogelio Bernal » Fri Oct 21, 2011 11:32 pm
For whatever is worth...
Color processing to this image has been done uniformly across the whole image. No selective color adjustments or "color picker" operations have been done.
Being a mosaic, I would expect that color calibration may show some differences between panes, as the 10 panes were captured with cameras of different sensitivity and at nights of different atmospheric conditions, despite we tried to account for all this. In any case, any differences wouldn't be particularly dramatic so as to make "purple" something that should have probably appeared "red".
The entire 2x5 mosaic used LRGB data. The panes comprising NGC 1499 also included H-Alpha data. Ha+ RGB integration was done at the linear stage, with the proper rescaling options to ensure some balance between the existing RGB data and the added H-Alpha data. So the assertion that the nebula has been "superimposed upon a background" is misleading. A more correct statement would be that data from different filters has been integrated for this particular region.
This didn't in fact add more "red" to NGC1499 (you could process just the RGB data and it will probably have an even more "pure" red color), it mainly emphasized the structures in the nebula and its visibility. The California nebula REALLY is much brighter than its surroundings - a linear stretch would saturate the nebula immediately, while the surrounding dust would still be barely visible, if at all. In other words, if the California nebula appeared less bright in this image - which is what I think some people may feel it would be "more natural" or "realistic" - that could actually be even more "misleading" than today's APOD image. A non-linear stretch can easily throw all that away, yet there really isn't any "blame" being thrown at that - it's just our perception:
the nebula is WAY TOO BRIGHT. Well, it is. And proportionally to all other structures, it is even brighter than that.
But let's leave that aside...
The biggest attention-grabber in NGC 1499 is not the color, which is what it is - red - and I've already mentioned its brightness as well. It's the high level of color saturation. Color and saturation are different things and must not be mistaken. Still, color saturation too was applied uniformly to the entire image, with the only exception of star color for which saturation was adjusted separately. All dust and nebulae in the image received the same amount of color saturation - every single pixel. A toned-down saturation would show most of the dust of a much pale color. By pushing saturation it, it can be argued that one could actually learn more than if the image was less saturated. What some find misleading, I find it to add value, but of course, I won't expect everyone to understand "value" equally, and there's nothing wrong with that.
Still, I would like to add that one of the reasons for the strong saturation (again, applied globally to the entire image) is actually not so much for aesthetic impact but because this image will be printed on a duratran-like transparency, and usually, it is a good idea to increase color saturation for this type of media. Once the image was finished, I personally felt it also looked "good enough" for web presentation and I decided to leave it as it is.
Ad yes Chris, there's something a bit masochist about all this
Regards,
Rogelio
For whatever is worth...
Color processing to this image has been done uniformly across the whole image. No selective color adjustments or "color picker" operations have been done.
Being a mosaic, I would expect that color calibration may show some differences between panes, as the 10 panes were captured with cameras of different sensitivity and at nights of different atmospheric conditions, despite we tried to account for all this. In any case, any differences wouldn't be particularly dramatic so as to make "purple" something that should have probably appeared "red".
The entire 2x5 mosaic used LRGB data. The panes comprising NGC 1499 also included H-Alpha data. Ha+ RGB integration was done at the linear stage, with the proper rescaling options to ensure some balance between the existing RGB data and the added H-Alpha data. So the assertion that the nebula has been "superimposed upon a background" is misleading. A more correct statement would be that data from different filters has been integrated for this particular region.
This didn't in fact add more "red" to NGC1499 (you could process just the RGB data and it will probably have an even more "pure" red color), it mainly emphasized the structures in the nebula and its visibility. The California nebula REALLY is much brighter than its surroundings - a linear stretch would saturate the nebula immediately, while the surrounding dust would still be barely visible, if at all. In other words, if the California nebula appeared less bright in this image - which is what I think some people may feel it would be "more natural" or "realistic" - that could actually be even more "misleading" than today's APOD image. A non-linear stretch can easily throw all that away, yet there really isn't any "blame" being thrown at that - it's just our perception: [i]the nebula is WAY TOO BRIGHT[/i]. Well, it is. And proportionally to all other structures, it is even brighter than that.
But let's leave that aside...
The biggest attention-grabber in NGC 1499 is not the color, which is what it is - red - and I've already mentioned its brightness as well. It's the high level of color saturation. Color and saturation are different things and must not be mistaken. Still, color saturation too was applied uniformly to the entire image, with the only exception of star color for which saturation was adjusted separately. All dust and nebulae in the image received the same amount of color saturation - every single pixel. A toned-down saturation would show most of the dust of a much pale color. By pushing saturation it, it can be argued that one could actually learn more than if the image was less saturated. What some find misleading, I find it to add value, but of course, I won't expect everyone to understand "value" equally, and there's nothing wrong with that.
Still, I would like to add that one of the reasons for the strong saturation (again, applied globally to the entire image) is actually not so much for aesthetic impact but because this image will be printed on a duratran-like transparency, and usually, it is a good idea to increase color saturation for this type of media. Once the image was finished, I personally felt it also looked "good enough" for web presentation and I decided to leave it as it is.
Ad yes Chris, there's something a bit masochist about all this :-)
Regards,
Rogelio