plasticman Posted November 18, 2006 Share #21 Posted November 18, 2006 Advertisement (gone after registration) Marc - could you possibly post a 100% crop showing a part of the adult's (your own?) face in the picture? I'm fascinated by the fact that it appears that the profile somehow rescues a lot of detail in the image, and i'd like to pixel-peep. ;-) Link to post Share on other sites More sharing options...
Advertisement Posted November 18, 2006 Posted November 18, 2006 Hi plasticman, Take a look here Thomas Knoll on the M8 and profiles . I'm sure you'll find what you were looking for!
mwilliamsphotography Posted November 18, 2006 Share #22 Posted November 18, 2006 Okay, (it's not me, it's my client holding his newborn baby). This was shot ISO 320, and exhibits a little camera shake blown up this far as I shot it handheld at 1/60th with a 90/2. But I also noticed more detail also. Same with other shots from that day. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/9447-thomas-knoll-on-the-m8-and-profiles/?do=findComment&comment=97278'>More sharing options...
plasticman Posted November 18, 2006 Share #23 Posted November 18, 2006 Okay, (it's not me, it's my client holding his newborn baby). oh - sorry :-) being a bit dumb today, due to an exceptionally good bottle of wine last night... The image seems indisputably better with this profile - strange thing is that i really thought Knoll must be right about this subject: seemed that IR shouldn't really be just a matter of magenta-cast so much as color-contamination affecting contrast and ALSO perceived color. The fact that images would therefore be generally 'muddied' by the IR shouldn't (i thought) be fixable with a profile... but this begins to suggest that's wrong. Interesting. Marc - thanks for taking the trouble to post the crop! Link to post Share on other sites More sharing options...
Jamie Roberts Posted November 18, 2006 Share #24 Posted November 18, 2006 oh - sorry :-) being a bit dumb today, due to an exceptionally good bottle of wine last night... The image seems indisputably better with this profile - strange thing is that i really thought Knoll must be right about this subject: seemed that IR shouldn't really be just a matter of magenta-cast so much as color-contamination affecting contrast and ALSO perceived color. The fact that images would therefore be generally 'muddied' by the IR shouldn't (i thought) be fixable with a profile... but this begins to suggest that's wrong. Interesting. Marc - thanks for taking the trouble to post the crop! Marc, Mani-- First, Marc--great shot, and I'm glad the profile helped. Secondly, thanks for all your kind words here and elsewhere on the forum! You know, Thomas Knoll is still right One of the reasons *I think* the new profiles rescue detail, especially in skin tones, was precisely because the original C1 profile was producing such truly out of balance results to begin with. In other words, there may still be IR pollution in other colours, and all the profile may be doing is mapping a clipped color / certain colour response to one set of magenta-associated neutrals *and* balancing, more or less, the color response. In other words, I think actually manually profiling the 'normal' LAB profile values with the colour checker in Profile Maker did the trick. It's amazing what you can do with 18 colour patches; once I have the DCC it'll be even easier to make sure there's no "gotchas"... Every time I'd just tweak one thing in the C1 profile editor, there would be a deficit somewhere else. So I tried to get as close to the CC / LAB values as possible, and as a control, I used the output from a variety of profiles (even for other cameras) that are known to be excellent. So that's it. I've just heard back from a landscape photographer who does think the greens are a bit weak. Fair enough; but that's truly easy to fix now with a neutral base. Link to post Share on other sites More sharing options...
mwilliamsphotography Posted November 18, 2006 Share #25 Posted November 18, 2006 You go boy ! Link to post Share on other sites More sharing options...
blakley Posted November 18, 2006 Share #26 Posted November 18, 2006 Jamie, you don't need to apologize for the use of the term "impossible" in reference to color - it's Dan Margulis' term (from "Photoshop LAB Color: The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace"), and Dan knows a lot about these things too... My thread on profiling and impossible colors (which you cited above) is here. In a sense I think everyone's right here. It's possible to have a device with a spectral response which allows you to definitively distinguish IR from visible colors, but the M8 sensor is not such a device. Therefore there is no "perfect profile" which will remove IR contamination from all images without affecting any visible colors. But, as your profile clearly demonstrates, the M8 sensor's spectral response does allow a profile to removal of most of the IR contamination from most images with only minor shifts in visible colors. Is this good enough? Well, that's a subjective judgment - which we should probably make in the knowledge that no digital camera on the market reproduces all visible colors accurately in all photos. Link to post Share on other sites More sharing options...
gogopix Posted November 18, 2006 Share #27 Posted November 18, 2006 Advertisement (gone after registration) This discussion may uncover a lot of reasons why the profiles DO work, and may be more satisfying than we think. The profile, if it just supresses the red channel will shift all colors, and would act somewhat like the 'color washer' plugings for PS. That works well in ALL cases (except black and grey as we will discuuss below). In a bayer sensor only the 1 in 4 R sensors will record the excess photons of the IR since we hope the green and blue filters work as advertised. These channels are ALL OVER the sensor, not just the blacks. It is just that the bayer interpolation spreads that red to the neighbors however slight hence the cast on black AND the red shift in all other colors. SO removing some amount of red, corresponding to the excess should be ok, except where ther may be a dark grey that would subtract the red from the x,x,x and leave x-IR,x,x which would introduce a color teal cast (blue/green) that seems to happen in the straight IR filters at the edges of WA lenses, since all things are not the same. If you think about it, a software shift in the red and filtering more IR in a bayer sensor world are the same thing. So why isnt it that simple. Well, it is because in NEITHER case is simple subtraction good. Now, I dont know the profile mapping, but color washing software is SMART as far as whites, grey and black are concerned. If the feeling is that real neutrals are correct, it leaves them alone Here we want the opposite we want to take FROM the ALMOST neutrals (and it seems white suffers just as much but the contribution of the red in the whites is less seen due to another M8 feature! the 16 to 8 bit compression. This is a long winded way of agreeing with Guy. I would like ly NOT have my M8 fixed for the IR but only for the banding. LOgically, I see NO reason why a SW algorithm can't be BETTER than a straight filter, except the filter costs money but is easier on workflow. IR excess in, red channel increases (in proportion of course to the reflective propertirs of the object)and that happens EVERYWHERE. IR filter takes out EVERYWHERE Profile can certainly be told to take out EVERYWHERE QED. but maybe it isn't working inspite of the evidence. Jamie, I've been convinced all along a SW solution exists. Your profile says it is easy. Are you in any way reluctant to give the final image to a client? Looks good to me! BTW all, what's wrong with these images? except the extra workflow (and maybesome tweaking) I see MORE differences in mixed tungsten/fluorescent lighting - a WB nightmare! regards Link to post Share on other sites More sharing options...
gogopix Posted November 18, 2006 Share #28 Posted November 18, 2006 Just as an example of what I would NOT want to lose these are tack sharp, 3 dimensional and the colors are spot on. no filter, no SW, jpg out Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/9447-thomas-knoll-on-the-m8-and-profiles/?do=findComment&comment=97489'>More sharing options...
malcolm Posted November 18, 2006 Share #29 Posted November 18, 2006 How these profiles affect "normal" pictures?Do the profiles affect the color balance when there are greens of blues in the picture? Are they acceptably "neutral" in normal cases? Capture One LE does not allow to apply profiles per picture. Even black and white conversions are difficult. It is useless to me, and Capture One Pro is too expensive for a non-professional. Ruben, I've used Jamie Roberts M8 profile (the higher contrast one) in Capture One LE, on an individual image - and I'm VERY clumsy with digital post-processing. See attached "before" (-orig2) and "after" (-JR2) images for comparison. Even though the original example seems a bit overexposed, and the after a bit underexposed - I haven't made any adjustments to exposure, contrast, brightness, or sharpening; but, I'm sure the relative brightness could be easily equalized by someone more accomplished in post processing. The comparison only shows the affect of using the JR profile - the "after" is virtually identical to how my human eyes saw the subjects. Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/9447-thomas-knoll-on-the-m8-and-profiles/?do=findComment&comment=97590'>More sharing options...
gogopix Posted November 18, 2006 Share #30 Posted November 18, 2006 Malcolm Nice job. Picture worth ten thousand words and all that... If one can fix the few shot that need it as you did, why give up the inherent sharp, detailed look of the M8? I will still use filters in some circumstances, but may not want any degradation. Thanks! Victor Link to post Share on other sites More sharing options...
Jamie Roberts Posted November 18, 2006 Share #31 Posted November 18, 2006 How these profiles affect "normal" pictures?Do the profiles affect the color balance when there are greens of blues in the picture? Are they acceptably "neutral" in normal cases? Capture One LE does not allow to apply profiles per picture. Even black and white conversions are difficult. It is useless to me, and Capture One Pro is too expensive for a non-professional. Ruben-- a couple of points here... First, the profiles are very neutral with greens and blues, in my experience. I have some landscape shots to work with (when I get a chance--gotta get some work done!).. Personally, I prefer a neutral RAW develop and if I want more saturation well, there are many ways to do that! Anyway, C1 *does* let you quickly apply more than one profile per picture, and LE and Pro work the same. In the process tab, when you add a shot to be developed, it freezes *all* of the settings that go into the develop. So here's what you do. Pick the profile you want, drop into the develop queue; change the profile, change the exposure, or sharpness, or whatever, drop into the queue; continue as much as you want. Very soon you'll have a queue with a color shot with new profiles, a BW profile shot with a different WB, and maybe a high saturation shot as well... When you hit develop (or have them developed when you drop them in the queue), then they will all be different. In other words, C1 develops each shot with the settings you make *when you develop it*, which is very cool. Same for LE, IIRC. Link to post Share on other sites More sharing options...
Guest guy_mancuso Posted November 19, 2006 Share #32 Posted November 19, 2006 Jamie what we need is a profile for when we have the cut filter on also in C1 Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.