|
Forums >
Digital Art and Retouching >
How Close did I get?
I love this grade:  This is the best I could do to emulate it:  I have a couple of questions for the pros. How close did I come? What did I do wrong? What can I do to improve it? It looks way too "unreal" to me but with this particular photo it seemed like anything I did gave it a hyper real look. I bounce back and forth using Ps for edits but organizing and grading in Lr. I am by no means a professional retoucher. There will be terms and techniques I'm not familiar with. Apr 27 16 12:11 am Link Welcome back  Carlitos Huanes wrote: 9/10 What did I do wrong? The light is different which makes it difficult for you to match the look accurately. What can I do to improve it? It looks way too "unreal" to me but with this particular photo it seemed like anything I did gave it a hyper real look. It seems to me you have a bit more magenta in your shadows and highlights compared to the target. Note how the sea foam in your case is orange-ish and in the first photo it is not. Of course that is related to the different light too but still something you might want to check. Also as a whole your version is somewhat more saturated in the blues too. I also don't see so much green in the water of the target. Your blacks are darker too. Apr 27 16 02:54 am Link Thanks Anchev! The top photo seems very specular to the point that it looks "crispy" or "sparkly". What characteristic am I responding to? Adding any more Clarity to my photo didn't achieve a similar effect and actually made it too "unreal" for my taste, like a Dragan photo. Apr 27 16 09:48 pm Link Carlitos Huanes wrote: Hey Carlitos, Apr 27 16 10:43 pm Link Never ever use Clarity. This is a destructive operation and as you see it creates halos. I wonder why they made that function at all, so many photos were destroyed by it. It is a form of filtering, not color grading. Even highlights/shadows is "dangerous" in that sense because it is a form of tone mapping and that is raster operation too. If you really care about quality - the only operation you should do in raw conversion is exposure and white balance and they should be done very carefully too. Specularity is the physical characteristic of surfaces to reflect directional light. Metals and dielectrics are different in this but that's a separate discussion. I will just mention that metals can have color reflections and dielectric reflection is normally with the same spectrum as the falling light (i.e. "white"). In any case you should remember the energy conservation law: Falling light = Diffuse reflection + Specular reflection + heat dissipation This means that if you want something to look specular and physically correct you should try to reduce the diffuse component (color) and increase the contrast of luminosity. For that to work properly you need to separate luminosity from color and tune them separately. But that is not possible in ACR. When working with curves or any other tool in LR, increasing contrast will always add saturation and that is just the opposite of what you want. Even if you reduce saturation globally after that it is still not the same, you simply have no control with the ACR tools. For that reason as I mentioned in the earlier thread it is unlikely that this can be done properly in LR. Apr 27 16 11:51 pm Link Thanks Tom, I tried a black and white layer set to multiply. I don't know what is actually happening when I do it but I got some very interesting results which did give it a "dirty" look. I'm certain I'll be using that technique. What do you mean by "noise ofc"? Thanks Anchev, I must say, I'm familiar with conservation of matter and energy, momentum, electrical charge etc. Falling light is a new one for me. Separating luminosity from color to manipulate them is interesting but outside of my knowledge base. Lr certainly has it's limitations. I'm not a huge fan of Clarity but when you say "destructive" do you mean it cannot be undone or it's distasteful? I didn't think any Lr operation affected the original raw file. Apr 28 16 11:56 pm Link Carlitos Huanes wrote: I mean incident light, the one that falls on the object directly or indirectly. Separating luminosity from color to manipulate them is interesting but outside of my knowledge base. Lr certainly has it's limitations. For creating looks, especially more complex ones, it is crucial to understand the structural elements of the image. LR is definitely not an app for color grading. It is an image database organizer with interface to ACR. I'm not a huge fan of Clarity but when you say "destructive" do you mean it cannot be undone or it's distasteful? I didn't think any Lr operation affected the original raw file. ACR operations are recorded in the XMP files by default. So you don't really touch the raw file. However once you create a halo and export it to another image format - obviously that cannot be undone without starting from the raw file again. You can apply ACR operations to other file formats too (JPG, TIFF, etc.) and they are recorded in the internal XMP data of the file. Apr 29 16 02:13 am Link Carlitos Huanes wrote: Noise "of course". Apr 30 16 01:40 am Link You guys are a wealth of knowledge. Thanks so much. And, thanks for giving that a try Tom. I decided to continue to use the inspiration but go more naturalistic. Understanding that it's purely subjective tell me if in your opinions it's a step in the wrong direction:  If you're curious to take a wack at the raw file yourself I'm happy to make the raw file available. I'm always interested to see how different artists use the same source material. Thanks again. May 01 16 02:14 am Link I would recommend that you have a look at the learning materials of Dan Margulis. This will give you a good starting point in understanding image channels, color and open new doors for you. May 09 16 02:43 am Link There's on-axis flash in the reference photo. So even if you got your processing to match exactly, you won't get the exact same results. May 09 16 10:56 am Link Sunlight hitting the back right of model. Hard light source hitting model on axis as said above. probably just a shoe mount flash. The cross lighting is the biggest thing missing in my opinion. The original image has dimension yours does not. From there, things are underexposed a little, saturated a little in the mids and highs, just on the model, everything else is desaturated, curved for warmth and some contrast. May 09 16 11:33 am Link Most of what's left was in camera stuff, that created the original image with the original light. You came very close with the post as you could. You might want to d&b the hand though, she's straining it and it really shows, you need to soften all the tendon action. May 09 16 09:34 pm Link J O H N A L L A N wrote: Good call John. I'll keep that in mind for the rest of the images. May 10 16 03:30 pm Link |
