Exposure & Fog of Death

<p>At one time I briefly touched on this subject; but never really jumped in with examples.</p>

<p>The two photos here have all EXIF data intact to avoid questions like what ISO, WB, lens, f/stop shutter etc..etc.</p>

<p>One photo is as is (UN-corrected) as shot.<br>

The other photo is corrected w/ levels & curves..That's all.</p>

<p>I have seem this time and time again and now wonder, WHY..WHY the "fog" as I call it? Slime might be a good word too.<br>

I fully understand the need for PP in nearly all digital images, but C'mon..Is this the life of digital?<br>

Is it impossible to have achieved the (corrected) image w/o ANY PP? </p>

<p>Looking at the histogram, ya; it's a tad overexposed, but I just don't believe this is the reason for the "fog."<br>

Whoever looks at this, compare the two images; switching from A to B while watching the histogram.<br>

I see a small shift to the right, BUT, the levels remain relatively unchanged.</p><div></div>

<p>I can't make heads or tails of what you are trying to show here. Fog? Slime? I just don't get what you mean by those terms.</p>

<p>I can't see any obvious differences between 1 and 2 except for some darkening and correction of overexposed highlights from the first one in the second one. The slight overexposure would normally follow from metering the water rather than the gravel and the child's shirt which are off center. I'd try metering off the lighter areas and then holding and recomposing the shot. Or use exposure compensation.</p>

<p>Maybe the "fog" refers to the out of focus background? Otherwise the images have no real technical issues that I can see.</p>

<blockquote>

<p>Fog? Slime? I just don't get what you mean by those terms.</p>

</blockquote>

<p>Poor descriptors maybe.</p>

<p>The UN-corrected image though slightly over-exposed has the appearance of (lower contrast) AND, looks as though a hazy plastic covering needs to be scraped off to reveal the "cleaner" image.<br>

Hope that helps describe what I'm talking about.<br>

My 2nd question dealt with how to shoot this if possible w/o PP which clears this image up immensely IMO.</p>

<p>I DID try <strong>SIX </strong>seperate f/stops to prove or disprove my theory. In other words, + exp comp and - exposure comp in 1/3 increments<br>

The haze, fog or whatever I choose to call it, remained; even with a technically good histogram.<br>

The mystery remains; at least for me.</p>

<p> </p>

<p>Look at the LH side of the histograms of the two images. In the uncorrected version, there are no truly black pixels. The darkest pixel is a dark grey. In the corrected version, the LH end of the histogram starts almost exactly at level = 0, as it should (...er... in most cases...). </p>

<p>Such corrections are most easily made by moving the LH slider in the "levels" tool of PS so that it is under the LH end of the histogram.<br>

Cheers,</p>

<p>Tom M<br>

Washington, DC</p>

<p><em>"Is it impossible to have achieved the (corrected) image w/o ANY PP?"</em> <br /> <br /> It's a bit (no pun) ..like having the tools but not accepting they are valid.<br /> <br /> Most labs may adjust a film image prior to or at print, most digital printers give you the opportunity to adjust an image at print. Many photographers, use under and over exposure to stretch the limits of a camera then apply a modification at or before print..or www. ...etc....etc.<br>

<br /> My suggestion is to print both files as they are now..and see which you like. You may well come to like what you have called "fog". Depending on the printer you use..adjustment in tones can make or break an image effect.<br /> <br /> http://www.cambridgeincolour.com/tutorials/photoshop-curves.htm<br>

<br /> Eventually the technology re:computation ability, of a digital camera, will make many of these adjustments automatic. eg;Imagine the face recognition software being able to recognise race or ethnicity..then the ability adapting that same system to identifying the difference between midday and evening or morning..and adjust colour/exposure/contrast...or the different types of trees..in a landscape ..therefore adjusting leaf colour to match species identified by the software and hardware in combination.<br>

<br /> It won't be long now..some clever person will supply profiles for each brand/model of camera, matched to different brand/model/printers. Eventually ..this will all be "in camera".<br /> And at that point...the answer to your question...may well be yes.</p>

<p>Tom; Anthony,</p>

<p>Thanks for looking.</p>

<p>Your opinions help to reinforce my original thought;<br>

On rare occasions, we may get a "perfectly" exposed shot that requires no PP...I've had a few myself; but generally some PP will be needed to clean up the shot. RAW is of course a different story and I expect to have to work the digital negative.</p>

<p>Ahh, if only we could shoot gray cards all day. ;), boring subject, but the exposure would be bang on; pretty much. ;)</p>

<p> </p>

Sure, crank up your camera's contrast and use your in-camera jpg. Seems kinda silly to rope yourself into zero lattitude by

having your post done up front, but whatever floats your boat.

<br><br>

Just because some of the noise hits 0 doesn't mean you've got a pleasing black point set. And sometimes crushing a little

shadow detail ends up making the picture as a whole pop better.

<br><br>

A camera that I couldn't get "fog" from is a camera I'd take back. The more decisions I can shift from exposure time to post

time the more brain cells I get to use on things that can't be fixed.

<p>Thanks Roger,</p>

<p>..but I'm not talking about the camera "nailing" it every time; but I see your point.<br>

If you look at my EXIF data, even though I shot the sample images in JPEG, I rarely allow the camera's PP algorithms to molest the image, save sharpening due to the lens I was using that evening. I tend to keep my jpegs neutral, but that is personal preference only.<br>

Maybe I need to crank up contrast etc when shooting jpeg.<br>

What I consider a (average) scene may not be what someone else considers average in tonality, contrast etc.<br>

In the sample I provided, there are no wild swings in exposure, so I know I was well within the camera's dynamic range.<br>

I notice this phenomena even with controlled studio lighting, although not quite as obtrusive.</p>

<p>The answer is very simple - the "fog" or white veil you see on digital images is because the sensor has a 1:1 gamma slope, and you're used to seeing a film characteristic curve as "normal." When you correct the image using "Curves" in a software program you are changing the gamma slope to match what you see on film - toe, straight line portion, shoulder. That changes the contrast relationships over the entire image. Many people (my self included) do not like the look of the linear response (1:1 slope) of digital images - therefore, I apply both levels and curves as part of post processing. </p>

<p>This is not much different that working with an image in the darkroom and finding the correct exposure and color balance through test prints for color prints - or the correct exposure and contrast for B&W prints. I really don't understand the "life of digital" comment? Making simple corrections is too much work? If you have a problem having to work with images - then send them out or quit taking photographs. </p>

<blockquote>

<p>Making simple corrections is too much work? If you have a problem having to work with images</p>

</blockquote>

<p>Thanks Steve; good response that helps me reach a conclusion.<br>

No; I have no problem working with or editing images, been doing it a long time.<br>

Perhaps my question stemmed from a little laziness since I do a lot of product work and am trying to cut the time I spend in PP.<br>

I love batch processing, but obviously one can not do this when the camera "looks" at scenes differently with no predictable outcome given so many variables.<br>

Given what many of us have come to know as the "perfect" looking histogram still does not guarantee a result we were hoping for.</p>

<p>So thanks to all,<br>

I feel satisfied with the answers and will continue to "tweak" the curve when needed; which seems to be often. ;)</p>

<p>Pete, if I'm reading the EXIF data correctly, you used a variable aperture zoom (18-200 VR) wide open at f/5.6, at the 170mm focal length. Try stopping down, if only 1/3 to 2/3 of a stop. Few lenses are at their best wide open, even when "wide open" means only f/5.6 Flare, including internal flare, will rob the photos of saturation and contrast. Virtually all lenses have some internal flare, despite sophisticated multi-coatings. That may be what you're seeing as "fog" - an overall lack of snap caused by flare.</p>

<p>Run some test photos, keeping the exposure value constant while stopping down. Run the lens through the entire aperture range. Compare results. You should see some improvements in what we perceive as "sharpness" between f/8-f/16, which includes increased saturation and contrast due to reduced internal flare. Even my 105/2.5 AI Nikkor - the best lens I own - produces a rather hazy look wide open even tho' the resolution is very good. Stopped down as little as to f/2.8 improves that snap considerably. (You'll also see a decrease in apparent sharpness due to diffraction at minimum apertures, but that's a whole nuther matter.)</p>

<p>Also, check your computer system calibration and relevant settings. The photo shows an embedded AdobeRGB profile. I'm looking at a copy of your photo converted to sRGB and it seems okay to me, considering the overall conditions. But in AdobeRGB it may look different to you.</p>

<p>Except when I'm shooting sports, I almost always have my cameras set to produce both a JPG and a raw file for each image. When I first switched to digital, I used to adjust the camera so that the JPG had moderately low contrast, low sharpening, moderately low color saturation etc. My reasoning, probably similar to yours, was that such settings would more accurately let me see "what was in the file", minimized the probability of clipped highlights, etc. Much as you describe, the JPGs all looked like they were taken through a slightly dirty window. Of course, I would then always be stuck with the task of tweaking the settings in the RAW converter to bring each image to life.</p>

<p>However, in the last couple of years, I have gone to almost the opposite extreme. I still usually produce both a JPG and a raw file for each image, however, I now have the camera set to produce JPGs which are moderately high contrast, sharpening and saturation, much like the images one used to get from a 1 hr film minilab. Often, the shadows (and hopefully to a lesser extent, the highlights) are a bit clipped, but as someone pointed out earlier in this thread, this can give an image more "pop".</p>

<p>To me, the advantage of doing it this way is that for many uses (eg, family snaps), images adjusted this way are quite adequate, and I don't have to do any post processing on the majority of my images. This contrasts to my previous method where it was mandatory to tweak each and every image. </p>

<p>A second advantage to producing higher contrast, sharper JPGs in-camera is that it gives me better idea of what a post processed version might look like for those images which are going to receive such treatment. Of course, in-camera contrast and color enhancement is rarely as good as targeted local contrast enhancements, targeted color adjustments, and the more sophisticated sharpening and noise reduction algorithms used in post processing of a raw image, but I find that being able to preview an image without it looking like it was taken through a dirty window is more appealing to me. It might be to you as well.</p>

<p>Sincerely,</p>

<p>Tom M<br>

Washington, DC</p>

<p><em>"...The answer is very simple - the "fog" or white veil you see on digital images is because the sensor has a 1:1 gamma slope, and you're used to seeing a film characteristic curve as "normal."..."</em><br>

Steve - The in-camera microprocessor gamma corrects the raw image before it is output as a JPG file. If cameras didn't do this, all in-camera JPGs would look utterly horrible, not just the small amount of "fog" that the OP is talking about.</p>

<p>Tom M<br>

Washington, DC</p>

<p>Lex & Tom,</p>

<p>Your comments are appreciated and valued.<br>

I've shot professionally for years, so it was difficult to post my question. I don't feel so bad now as it is becoming clear that "out of camera" jpegs will rarely produce a shoot it and print it scenario. This seems to be not a stupid question with a obvious answer I may have overlooked.</p>

<p>I shoot jpegs for fun..and lots of 'em..When a client is paying, RAW all the way.</p>

<p>Tom, I'm glad someone else was able to see this "Dirty Veil." My eyes are a little older, but not THAT old yet. LOL..and you might be right Tom, I may have to adjust some of the "in Camera" processing to get a image that eliminates or reduces this effect I dislike. Personally, with my bias to RAW shooting, I find it hard to molest even a jpeg with forced algorithms..such is life I suppose with jpeg.</p>

<p>Lex, great observations...and yes, my camera profile is set to Adobe RGB, I just convert it to sRGB at PP for web uploads etc. Interesting comment on lens flare. The lens here (18-200) while hardly of super high quality seems to control flare fairly well and the sun was at my back as I thought about this test shot, elimination ghosting too. Your analysis has some credence though; I don't see this "veil" effect as much with my 50mm prime, but it is still there.</p>

<p>Thanks again,</p>

<p>Pete</p>

<blockquote>

<p>Steve - The in-camera microprocessor gamma corrects the raw image before it is output as a JPG file. If cameras didn't do this, all in-camera JPGs would look utterly horrible, not just the small amount of "fog" that the OP is talking about.</p>

</blockquote>

<p>I wasn't addressing JPEGs. RAW files are unprocessed and are 1:1 gamma - look at a raw file in your favorite processing program and you will see a 1:1 slope that is totally uncorrected. I have no use for JPEGs for processing as I own an M8 and the JPEGs from that camera are totally unusable.</p>

<p>Steve - I'm sorry, I must be missing something, but I just can't see how your comment about gamma=1 conversions from NEFs to JPGs pertains to the ongoing discussion. There was no indication anywhere in this thread that the original poster's "original" image was anything but a JPG that had been converted in-camera (ie, with a normal gamma). He states this in two ways: (a) by saying that he strongly prefers to use the JPG straight from the camera with as few adjustments as possible; and (b), in the post that immediately precedes your 1st post in this thread, he explicitly stated that the sample images are JPGs, not raw files that he tinkered with during a post-processing RAW-JPG conversion step. In his original post, he states that the only thing he did to his second image was to apply levels / curves to the 1st image.</p>

<p>In fact, they couldn't possibly be the result of a NEF to JPG conversion with a gamma of 1 because if they were, they would look ridiculous, whereas in fact, they look quite reasonable ... not perfect, but very close. In fact, in any raw converter that I have ever used, one would have to explicitly override a normal gamma (ie, 1.8 - 2.2) default value and set it to unity to see what the raw data "really" looks like. Pete, please chime in to clarify, but to me it seems clear that Pete did not do such a thing to his sample photos.</p>

<p>Tom M</p>

<p>Pete - A long shot ... By any chance, might Nikon's "Active-D-lighting" feature been enabled on your camera when you took the original shot. If so, it could have made the shadow areas not as inky black as you expected. </p>

<p>Don't forget, there is no such thing as a virgin, un-corrupted JPG from a digital camera. The only difference between the JPG that comes from your camera and one that comes from a raw converter software program is that in the former, you are using someone else's conversion presets, whereas in the latter, you are using yours. :-)</p>

<p>Cheers,</p>

<p>Tom M<br>

Washington, DC</p>

<blockquote>

<p><a href="../photodb/user?user_id=4873900"><em>Pete Harlan</em></a><em> </em><a href="../member-status-icons"><em><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/1roll.gif" alt="" /></em></a><em>, Sep 01, 2009; 09:29 p.m.</em><br /><em>C'mon..Is this the life of digital?</em><br /><em></em></p>

</blockquote>

<p>Pete, in a nutshell, yes, analog or "wet" photography too because of numerous factors, far too many to get involved with here. Suffice to say that there's no way you (or anybody) can say, "Yup that's EXACTLY what I remember seeing". So choices have to be made using test strips in "wet" photography, or twiddling in the digital domain. First, the recording medium is roughly linear, the human optical system is closer to logarithmic. Throw in different color gamuts, and things get REALLY nutty !<br />Interestingly enough, nobody mentioned the Digital Haze phenomenon inherent in all Bayer Matrix systems, which is a major contributor to the "fog" you mention.<br />So we live in an extremely imperfect world, then we try to photograph it "accurately".</p>

<p>Quick example......<br />Is photo 1 in your example what you remember, or is photo 2 what you remember ?</p>

<p>Hmmmmm....................</p>

<p>Bill P.<em> </em></p>

<p>Tom:</p>

<p>The test images were shot as JPEGS. The ONLY correction I made was <strong>CURVES</strong> to the photo labled<strong> Corrected.</strong><br>

No; <strong>ADL</strong> was not turned on. I rarely use it. Good guess though, I have left it on by accident in the past which can cause probs.</p>

<p>William: The <em>corrected</em> image is far closer to what I recall.<br>

I don't shoot thru "dirty windows." :)</p>

<blockquote>

<p ><a href="http://www.photo.net/photodb/user?user_id=4873900"><em>Pete Harlan</em></a><em> </em><a href="http://www.photo.net/member-status-icons"><em><img title="Frequent poster" src="http://static.photo.net/v3graphics/member-status-icons/1roll.gif" alt="" /></em></a><em>, Sep 03, 2009; 07:48 a.m.</em><br>

<em>William: The corrected image is far closer to what I recall.<br />I don't shoot thru "dirty windows." :)</em></p>

</blockquote>

<p>Pete, I shoot much the same way, with similar results, and I have to do the same approximating as you do. It's just part of the "way it is".<br>

Nice work !<br>

(*_*)/<br>

(Japanese emoticon)</p>

<p>Bill P.</p>

<p>is the 'fog' not just the outcome of having your camera's contrast set to low? or having a natural low contrast lens?</p>

<p>something that should be fixed later in PP. I set my digital to about a quarter sharpness and contrast then boost later, this gives the 'fog' effect you are talking about.</p>

<blockquote>

<p>Digital Haze phenomenon inherent in all Bayer Matrix systems, which is a major contributor to the "fog" you mention.</p>

</blockquote>

<p><br />Now <strong>THAT</strong> seems to make sense.</p>

<blockquote>

<p>is the 'fog' not just the outcome of having your camera's contrast set to low?</p>

</blockquote>

<p>believe me; I considered that sometime ago, but the answer is (no).</p>

<p>If I may back up a bit.</p>

<p>First, my posting has nothing to do with shooting RAW.<br>

I shoot RAW (High Priority Images) when it counts and <strong>expect </strong>to be PPing.</p>

<p>When I shoot what I consider (Low-Priority) images, birthday parties, family gatherings, etc; I shoot jpeg. I have neither the time nor the inclination to PP RAW files for photos like that.<br>

Someone will no doubt ask "why not shoot both?" Personal preference; I don't want the additional "Fat"<br>

of RAW + JPEG taking up HD space for no real reason. 'Nuff said.<br>

Besides, a camera like the D-300 produces wonderful high quality JPEGS anyway when extensive editing is not in my plan. (i.e) birthday parties.</p>

<p>Setting "in camera" parameters to coax a (cleaner) image seems too variable to me it it's final outcome.<br>

Setting contrast to a higher value is fine in <strong>SOME </strong>situations<strong>.</strong> The result though is one I prefer not to have; that being "clipped" shadow areas.<br>

Hence; I find myself coercing the image in PP anyway.</p>

<p>Based on all the great input so far, I've come to the conclusion a custom curve may be my answer and leave the "in camera" processing pretty much neutral when shooting JPEG.</p>

<p> </p>