|
Forums >
Photography Talk >
Why is the D800 camera profile so bad in CO7?
Why is the D800 camera profile so bad in CO7? Skin tones are VERY unnatural. While my older Canon 1Ds and Canon 5DII files open quite well at standard settings D800 files look wrong. If I open my Canon images in Canon's Raw converter, ACR and CO7 all at default setting they look very similar. If I open d800 files in NX2,ACR and CO7 the NX@ and adobe ACR are quite similar with NX2 looking significantly nicer the CO7 conversion at default setting looks way off with "brownish" skin tones and patchy skin.  The actors true skin tone is what you see in the Nikon shot. He's blond and fair skinned. If I want a fake tan look like the CO7 conversion I can do easily in post. Straight RAW conversions should be accurate and natural. Feb 03 14 04:15 pm Link Although the one on the left looks alittle more appealing than the slightly more yellow/red image on the right... I'm thinking your client would probably be happy with either one. What amazes me is how you captured the EXACT same image at the EXACT same time with two different cameras... OR... are you just opening up the same file with two differnet applications...  Feb 03 14 05:13 pm Link I like the image on the left, I think we could agree that if he comes out like a smurf, regardless of what we shoot, we will agree that it is wrong. No real human (without MUA) looks blue, unless lit by blue light. For a comparison like this, my wrong is your right. Feb 03 14 06:15 pm Link There are proprietary algorithms in Nikon' software to process its NEF files. Feb 03 14 06:43 pm Link Why do people think skin tones should be perfect 'out of the camera'? If you get the limits of imaging sensors and how the temperature of light affects color, you just shoot RAW and adjust. It's easy to come up with a simple workflow to nail skin tones with minimal post processing. Works on my D70, D700, D800, V1or LX-5. I'm not sure of the OPs workflow, but in 99% of other cases the problem is a result of technique, not hardware. It's just easier to blame hardware. Feb 03 14 06:50 pm Link Feb 03 14 06:53 pm Link Everyone except Nikon is guessing at the algorithms used by Nikon. Granted Adobe's guessing is pretty good, there's nothing like knowing the actual ones. I always use NX2 to open .NEF files and then import into CS. I've had any problems with accuracy. Feb 03 14 06:59 pm Link Robb Mann wrote: Robb this is not a technique issue. Feb 03 14 07:36 pm Link Robb Mann wrote: I'm going to start to blame mixed lighting conditions and temps on equipment from now on. Feb 03 14 07:46 pm Link oops Feb 03 14 08:05 pm Link Have you checked the default color profile the programs try to open with? When I tried to open my d800 raw files for the first time, everything defaulted to an adobe 1998 profile. I guess that's alright if I'm putting my pictures on a sega or nintendo from the 80s. Feb 03 14 09:59 pm Link I feel your pain and frustration Fred. I also shoot Raw with a D800. I have the same prob but in Adobe LR4 and CS5 / ACR The shadow areas only in skin tones look fine in NX2, actually great; but get more contrasty and almost blotchy in any Adobe software. It's subtle, but noticeable, and drives me nuts. I'm glad you mentioned it, because I have not heard any complaints anywhere else, including the Nikon forums. Other Nikon photographers I know don't seem to have that problem, or they just haven't noticed. I just assumed it was something I was doing wrong or missing. I was hoping maybe the latest versions would have better conversion algorithms but I don't currently have the money to upgrade or try any other Raw converter software; So I have been dealing with retouching when needed. What's worse: Because of this, I revisited some of my older NEF images, (shot with a D700) and they have the same prob now. I swear they looked fine when in LR3 and CS4 but changed after I upgraded and imported into LR4 and CS5. I am going to do some comparisons soon. I think there is something wrong with all the other software NEF conversions. Unless there is something we are totally missing. Feb 03 14 10:44 pm Link Jacob in Japan wrote: Adobe RGB 1998 or just Adobe 1998? Feb 04 14 09:06 am Link The D800's skin tones are amongst the worst I have seen, either in ACR or NX2? Even my ancient D200 looks better! Feb 04 14 02:59 pm Link Here is a reply I got from a Phase One "Crew" staff member. Re: Why is the D800 camera profile so bad in CO7? So Phase One made a deliberate choice to brown tint the Phase One ICC profile of the D800 according to Christian Gr Feb 04 14 03:10 pm Link DaeNaturals wrote: I see the same thing you are seeing between ACR and NX2, but as you say it's a subtle difference. The difference I'm seeing in Capture One is not subtle. Feb 04 14 03:21 pm Link London Fog wrote: I recall my friends d200 often had red artifacting in the gradint near deep shadows on skin. Feb 04 14 03:54 pm Link Since I can remember, Nikon D70, D300 and now D800e come into LR & C1P a bit on the orange-ish side. Seems to be a Nikon thing. Feb 04 14 04:25 pm Link In general, I've been more than happy with the my images using LR 4 for D800 conversions:  That said, I'm old and digital is still sort of a mystery to me. I will say, though, that I just received my first copy of a book featuring many of my images of Antique/Classic Raceboats, shot with my first digital camera, a D2X. Virtually all of these were shot on the water, boat-to-boat--in 2005--and the quality, at least to me, is more than impressive. The auto-focus--I was in a moving boat, shooting a moving boat, sometimes very fast (they're raceboats, remember, and the fastest competitive boats of their respective eras, going as fast as they can) using 2005 Nikon technology, is right on. I had never seen these images in print, and I'm blown away. RBD Feb 04 14 06:51 pm Link Antique Raceboats.... that must be quite a book. When the photographer is blown away it means he's done a great job. Feb 04 14 10:50 pm Link Does C1 allow you to create your own profiles, like you can with LR? Feb 05 14 07:49 pm Link It could be the possibility that capture one is reading the ISO wrong as it looks slightly underexposed? A few measurements have shown the D800 iso100 is actually iso65 or so. Feb 06 14 04:07 pm Link Yingwah Productions wrote: Could you elaborate on those measurements please. I have noticed C1P behavior to suggest that what you say is true. Feb 06 14 04:29 pm Link Geez, why are so many commenters being so obtuse or commenting if they are completely ignorant & unhelpful. You might want to hold off on advice if you do not understand the question. This is not a technique or lighting issue, the OP is asking about discrepancies in RAW decoders for the same file. He did not shoot the same photo with two different cameras (face - palm). That said, OP, every RAW decoder works differently. For some shots with certain cameras one decoder will be superior to another. You may find this comparison to be helpful: http://www.lifeafterphotoshop.com/dxo-v … -pro-best/ Even different process versions of RAW decoders can change an image's appearance significantly. To my eye, it seems that Capture One is defaulting to converting with more contrast, clarity, vibrance and sharpness. I am a Lightroom user, but you should be able to at least compensate by changing the default import settings. Feb 06 14 04:29 pm Link Yingwah Productions wrote: No it's not an ISO reading error. Christian Gr wrote: They are dding contrast and a brown tint to skin tones .... and anything else in the photo that is in the same color range. Feb 06 14 04:30 pm Link Leggy Mountbatten wrote: Good question. Feb 06 14 08:16 pm Link No not really. You can save your correction done to a photo as an ICC profile, but as far as I know there is no proper tool for building accurate profiles from targets. Feb 06 14 09:43 pm Link Have a look at MM thread where we were discussing MFD/35mm equivalent. Mescalamba had few good points on conversion. Feb 07 14 12:55 am Link Vector One Photography wrote: those algorithms (hardware + software) are reverse-engineerable by interested parties. there are plenty of hackers with a lot of time on their hands and huge ego. Feb 07 14 04:55 am Link Just spotted this on another forum: Quote from: T.Dascalos on Today at 08:16:03 AM So Phase One comes out with a back with a Sony sensor. They make a profile for it and guess what..... using the IQ250 profile on the Nikon D800 makes the D800 look amazing. Feb 09 14 01:48 pm Link Fred Greissing wrote: Can you try it and see if you get the same results? Feb 09 14 08:42 pm Link Giacomo Cirrincioni wrote: I tried it. One of the IQ250 profiles is significantly better than the D800 profiles. Feb 10 14 08:26 am Link Fred, I was playing around with Profile Inspector (Link: http://www.color.org/profileinspector.xalter ) and there does seem to be a difference in the Capture One DataBack and their one for the D800. They make a distinction on their file folders as "DB" and "DSLR" so it was easy to find the ICMs. Profile Inspector might allow you to save the PhaseOneIQ260-Outdoor Daylight.icm (Didn't see the one you referenced?) to a Nikon one as it does allow you to Save and Modify the ICM to an extent. Haven't tried it though. I did find it interesting that some time back when Qimage Ultimate began to print RAW files directly that the converter used for the D800 and D800E were different. Seems the generic RAW software conversion programmer made the framing different between the two and the Qimage software guy/owner had to change the RAW converter code himself to make the D800E work correctly in Qimage. I don't think you can beat Capture NX2 for conversion though. Nikon probably didn't provide much in the way of an SDK for other programmers over what they hold onto. Does seem the Phase One's Capture wants to make their converter a little better than what the D800 does though. Very suspicious if they think "Their customers asked for darker orange skin in some feedback poll for the D800." If true, their backs should be the same, or so I would think. Feb 10 14 10:09 am Link GRMACK wrote: This whole brown-tinted profile is either suspicious or simply rather amaturish. Feb 10 14 11:37 am Link Fred Greissing wrote: I've always struggled with Capture One. It feels like I have to fight it to get to the right color. Overall, it feels like far more work then using LR5, and for the most part, I want to cut that RAW processing time down as much as possible. When it's on, it's dead on... but I shouldn't have to browbeat it to get there. Feb 10 14 01:04 pm Link I'm not the only one that has noticed the fishy business going on with Capture One's profile for the D800. Take a look at this: http://www.luminous-landscape.com/forum … 7799;image Notice how the IQ250 color profile looks much better and how there are more subtle tones in both the sky and in the ground below the cliff. From this it is very clear that the Capture One D800 profile is deliberately altered (as stated by Phase One 'brown tinted') and that this is to the determent of IQ from the D800. Feb 13 14 04:27 pm Link Capture One isn't so hot on converting Canon Raw's either. Especially if Highlight Protect is turned on. Feb 14 14 07:54 pm Link Interesting and dynamic post, Fred. I tried Capture One-And-the-market but I didn't like its interpretation. On some photos I will open in ViewNX but sometimes Lightroom opens good. I really want the ViewNX view as it matches my LCD view. I am running off that when I work. The very early morning I quickly processed from a commerical shoot some images shot over three minutes. Complex but I had everything the way I wanted when I shot it. A shirt was stained so I cleaned it up but I didn't change exposure or contrast. I really don't want some user profile impose that messes up fair skin. On one image I opened in Capture one it automatically made it more contrasty and darkened her and it didn't work. Feb 14 14 08:17 pm Link Smedley Whiplash wrote: I have the opposite experience - Levels in C1Pro is a far better tool than LR's Tone Curve for my work. Love the way C1 handles contrast and saturation, too. Feb 14 14 08:28 pm Link http://www.luminous-landscape.com/forum … c=87187.40 Interesting that another forum is discussing how bad CO7 profiles for the d800 are. Feb 14 14 09:56 pm Link |
