Gossips

Great details and tones in the large version Pat and I like your comp also. Processing wise I only know so much,others here will know more, but by artifacts do you mean halos. Silver fx produces them a lot I find. If your working through a mask for the sky it sounds like the mask needs to be more refined than a quick selection if that's how you're doing it. If you want email me might be easier. Nice shot though.

Very nice comp that presents a real feeling of huddled whispering among the figures in the formation. The large version needs to be viewed. well done, Pat. Chet

Love the comp and backlit effect. The only color I'm really seeing is in the sky so I'd like to see some color hint in the rock (I'm guessing highly unsaturated sandstone reds.) Then again this works just fine the way it is too.

Beautiful !

Bill,

Thank you for the offer. It's halos and I'm seeing them created in LR, PS, Silver Efex and even to a lesser extent with TK's basic and expanded midtone actions. What I'm actually curious about is the algorithms that are being used for the contrast adjustments in these applications. May be proprietary but I was hoping someone knew of a reference that goes into the nuts and bolts of the image processing code.

pat

Pat: The hint of color makes this for me. Love the detail in the larger version and the overall comp. I really enjoyed our evening last night. When I grow up I want to learn to process like you so this looks pretty good to me. 2 thumbs up! >=))>

Thanks Bill. Same here. Hope you don't freeze in RMNP this weekend. It was seven below zero this morning when I took Molly to the dog park and I guess there's supposed to be another wave of this coming through on Saturday. Take care.

pat

I like the image, Pat, and the selective color. I think anyone who does conversions in any program is going to fight that battle. I know nothing of the algorithmic formulas, but I do find that if I remember to do chromatic aberration corrections in lightroom (raw conversion) and really blow things up on exposed ridge lines, it seems to help. Some lighting conditions are really hard to deal with and I think you did an admiral job on this.

The subject is interesting and I really like the effects of the lighting in the sky. The detail look good in the larger version. Nice composition.

Gary,

Thanks to your prior suggestion, I do always now use the chromatic aberration adjustment in LR and it does help. There are basically two kinds of halos that I routinely see: a sharp one to three pixel bright halo next to dark edges and a wide gradient sort of halo (maybe as much as 50 pixels) that appears around vertical shapes in an image. The first is at least related to chromatic aberration but it's hard for me to imagine that the second type is. The second is more like noisy turbulence - it's smoky looking and very faint. Both are obviously artifacts created by what I consider to be buggy algorithms in the applications. My argument is that if I can see the artifacts, the algorithms should be sophisticated enough to account for them and eliminate them; or not create them in the first place. Hope that makes sense.

pat

I really like the dark and foreboding mood of the image. It works really well.

Pat, This does work really nice! I like the darkness of it and that glow in the sky looks great. Excellent comp and the large version is wonderful to look at. Love the textures in the rock and the amazing details. Pretty impressive image to me!

Nick Bristol
Lone Rock, WI

Pat,

I think this image is of rocking good quality. The tones are so sweet, and the color left in the sky produces such a nice subtle contrast in the canyon walls.

All the technical conversation is intriguing to me, but I just don't want it to happen in the end, lol. Maybe you can email the adobe team, and let us all know what you come up with?


I think this one a lot

Love the monolithic character of the rock tower and the backlighting really accentuates that feel. The touch of color gives an eerie feel to the image which works well. Wish I too understood better why halos and artifacts crop up even when using the feathered masks of Tony Kuyper's masking techniques at times. Really nice image.

I really like the mood of this Pat. Feels like it could be somewhere in Middle Earth where Gollum might exist. The larger version shows a lot of the details and textures in the rocks. Nicely processed.

Hey, Pat, this is a neat take on this set of hoodoos. The almost b&w approach gives a lot of character to the gossips.

As to the issue of halos, I suspect that your 1-3 pixel halo may very well be the human vision effect called "Mach bands" (check out the wikipedia description) and that we see the brighter side more easily than the darker side. I see that regularly even in RAW files and have wondered about on-chip effects. My limited knowledge of CCD performance says that there shouldn't be any charge leakage from one pixel to adjacent pixels until the central pixel starts to saturate (>90% of the maximum charge) and that the result would be a spread of brightening around the saturating pixel. That's not what I see in my images were there seems to be extra brightness on the bright side of any strong contrast area.

I see the larger halos (20 - 30 pixels) when there's a strong brightness change and I'm using TKs triple play in PS. I wouldn't be surprised to find that other software packages are using similar algorithms, just keeping it hidden. BTW, I see several light dust spots in the larger version. I find a fair number of those that are very hard to spot in the raw file, but show up way to nicely when I'm doing high contrast processing.

Mark,

Thank you very much for that input.

I agree with you about the saturation leakage. And for further confirmation, I don't see it until I'm in post processing and then usually after a tweak to contrast - either global or local.

In this image, the larger halo appeared with almost anything I did but unless it was invisible in one of the channels in the raw image, it wasn't there in RAW. So it was a processing artifact. Interestingly, at least to me, the GIMP algorithms are documented and If the descriptions are accurate, I'm surprised at how simplistic many of them are. I would have expected a high quality contrast algorithm to involve edge detection and I didn't see that in the documentation. I also didn't see anything in the algorithm description that would prevent a calculation that drives the lumination to an unrecoverable limit. In short, they don't appear to be reversible functions. I may not be seeing an adequate description though.

I guess in addition to just wanting to know how the algorithms work, I'd also like to consider a design that would improve performance. I wonder how many of the developers are actually practicing photographers.

Take care.

pat

Just right. The angle works well, touch of light strong, and suggestion of color a fine choice.