But DXO don’t care about this do they? If there would be support for a new camera in DNG it shouldn’t make a difference in Photolab if Photolab was lacking a profile for that camera. Photolab would still refuse to open files from that camera.
Deleted by user after finishing to read this long topic.
Response has been found.
@Marie
Default factory rendering for D850 is “automatic” . So it seems D850 choose by default what it judge to be the more appropriate picture control (Nikon terminology) for the current shoot.
(I don’t want to reset my body, I’m afraid to forget some settings, but I’m pretty sure it was like this when I bought it).
I suppose you choosed the one named “standard” by Nikon as reference. (Flat could have been a good starting point too).
Can you confirm ?
Thanx.
Hello @platypus ,
we talk about RAW images here, so we don’t need to apply a rendering then removing it to apply another. RAW doesn’t have a rendering at first so we directly apply the one selected in the menu. As long as image is not exported nothing is definitive.
Regards,
Marie
Hello,
I am wondering, if this is really the case? Even though the RAW file does not have a rendering, still each camera manufacturer uses different bayer color filters. I think this is what @platypus was referring to, that there must be some form of calibration to equalize the difference in color filters, before any color profile is applied.
For example, if I use a camera of manufacturer A that has weaker green filters, the pixel might have an intensity of 80%, and for the same scene, a different camera of manufacturer B with stronger green filters will only have an intensity of 60% for that same pixel. If I apply the same profile on both raw files, the output image will not be identical, as it is claimed.
Or are those difference negligible in practice?
Hello,
@JoPoV , you are correct.
@maderafunk the RAW does not have a rendering but here I’m speaking of the rendering we apply to mimic the jpg from the camera with the standard picture control (Nikon terminology) . Because if we don’t apply a rendering at all you would have a really dark image with some color but nothing characteristic, nothing you can work with as a basis.
Regards,
Marie
1 Like
@Marie
I think what maderafunk is telling is that when a sensor records raw datas light coming from lens is filtered by bayer filter. And each sensor has it’s own bayer filter colors.
If raw formats were the same for every sensor (i know it is not the case), there would be differences between raws datas for a same image in this unique format due to different bayern filters (forgerting to consider the differences of photosites, etc …).
Sensors don’t record readable images, but in some way, we can tell they each have their own “color space”.
Madefarunk is asking if those difference of “color space” are taken into account when applying sensor B setting to sensor A, or if when applying sensor B settings to sensor A, exactly the same “algorythm” is apllied to sensors B than the one applied to sensor A.
Hello,
applying sensor B setting to sensor A is not done brutally, we have adaptation calculation that take into account properties of each sensor. We do measurements of each sensors so we know how to switch color renderings.
Regards,
Marie
3 Likes