carl smith Posted January 28, 2005 Share Posted January 28, 2005 You can do a world of post processing but there is a strong advantage to the camera that has the largest out of camera dynamic range. Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 Too bad in real terms, this camera doesn't exist:) Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 Just in case my point was missed, any advantage film might have is lost because of the effort it takes to get a decent negative to the computer. Digital whips film today as a standard, half-way decent D-SLR will run rings around film in post as to convenience and the ease in which one can get full ranging dynamic range out of decent D-SLRs as opposed to the inconvenience to processing and getting the neg into the computer. Out of camera is out of camera and you can't get anymore out of camera then digital. Film ain't usable until it's been processed and that's part of the out of camera, post-processing experience. Standard ASA 400 print film is horrible in the dynamic range department. I know cause my wife uses film on her vacations and the images have noooooo dynamic range. Either the shadows are loaded up or the sky is blown, that's not dynamic range. Me thinks some here are trying to turn this thread into a film Vs digital debate as opposed to the original question: "Which DSLR is suppose to or has the best dynamic range right now?" "What is holding them back from making a sensor that can capture more dynamic range?" The question had nothing to do with film. In response to the original question: There's nothing holding anybody back from creating a sensor that has more dynamic range in it. What's holding folks back is that they don't know how to process the image in which to bring the dynamic range out of the file that the sensor captured. There's plenty of dynamic range in a digital capture of today.<div></div> Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 How easy was the above image to create? Load the image into Photoshop, mask the shadow area, level. Use the history brush at about 50% opacity in which to blend the dividing line between the two halves of the image and Voila!; done. A bit of playing for contrast/saturation/USM and airbrushing some wires. No multilayers masks or all the other falderal that I've read about. No second images to capture and blend or any sort of special photoshop skills needed. Just some good basic tenants to follow. It's done in it's most simple form by masking and leveling the area in question. Leveling the masked portion and blending with the history brush, couldn't be simplier and if done light to medium, the process doesn't destroy the image like all the other more complicated techniques do. I can't believe all the garbage I've read as to what needs to be done. It's like these folks worked hard at getting as many steps into the process as possible. Heck, half the time, I couldn't get either the computer to follow their program or I couldn't find all the steps they were describing:) "Where did they find that step?" "I can't find that command anywhere." And after I did follow their process, the image was so unuseable, it was laughable. It ain't hard people:) Give it a try. Recipe: Make capture set to RAW, in camera settings/perameters on lowest setting expose the histogram one line (stop) from right side of graph. Process in BreezeBrowser. In it's most simplest form: Load image into PSCS, mask shadow area>level right side of slider to taste>blend with history brush at 50% opacity>done:) Each step in post chews up dynamic range and the order which each step is applied matters as to how much dynamic range of the image is chewed up. If you don't believe me, process the image with the histogram in the post-processing window and watch what happens to your histogram as you apply different steps. The graph will march right out the window to either the left of the right; clipping. And once it goes out the ends, the information is lost. Now you're, through your own personal PS efforts, clipping dynamic range off the image, after the fact. Mask and level the shadows first. Then apply saturation/contrast masks. Watch the range of the image expand out. After these first two steps, apply USM, which further expands an images dynamic range. Lastly, after resizing or USM'g an image, finalize the leveling of the image. If not done in this manner, you're sure to clip shadows and highlights, reducing dynamic range. Watch the histogram in Photoshop as this is your only hope of protecting dynamic range. Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 Oh! And ya gotta convert RAW 12bit to 16bit TIFF. You need the extra bit depth the 16bit provides in post for smoothness in transition as opposed to the choppy nature of a 8bit JPEG. Link to comment Share on other sites More sharing options...
rayfraser Posted January 28, 2005 Share Posted January 28, 2005 Ah! Not just one but several intelligent responses from Thomas!!! My mind is now saturated with his dynamic thinking. While he indicates it is easy, I expect I'll need quite a bit of time to try (may be impossible using raw data from my cameras). I still feel severely limited by 24 bit (3*8) computer monitors and have posted a question in Digital Darkroom. <p>Here is a link to an article I do not fully understand but found interesting: <a href="http://www.scantips.com/basics14.html">Dynamic Range by Wayne Fulton</a>. Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 "Ah! Not just one but several..." I'm not in trouble am I? :) Thanks for the comment and the link. After the several postings, I realized that it was easy because I had spent so much time reading articles such as the one you linked to and had done a whole lot of experimentation. Now that I've pushed through the learning curve, it's easy:) But it's also easy if one were to pratice with the magic lasso, creating a mask which will allow one to work selectively with shadow areas via the level command. Simply magic lasso the shadow area you wish to work, bring up the level command and use the right slider back and forth to change the level setting. Usually, when there's a large blank/empty or flat area on the right side, one should bring the right slider in about half way or less. Any more than that and the image quality is degraded considerably. Then change over to the history brush, set opacity at about 50% and erase the leveling action around the boundry to blend the two areas in. All in all, once one get's the hang of it, it takes five or ten minutes to do and the results are quite rewarding.<div></div> Link to comment Share on other sites More sharing options...
beeman458 Posted January 28, 2005 Share Posted January 28, 2005 The second image is the unprocessed image. Image was reduced in size and saved for the web.<div></div> Link to comment Share on other sites More sharing options...
jeff medkeff anchorage, a Posted January 29, 2005 Share Posted January 29, 2005 <p><i>You can do a world of post processing but there is a strong advantage to the camera that has the largest out of camera dynamic range.</i> <p>An utterly meaningless statement. The camera that has the largest out-of-camera dynamic range is the same camera that has the largest dynamic range. Period. If 1 to 4095 in a Canon raw file gives you eight stops, then that's the dynamic range. If it gives you nine, then nine stops is the dynamic range. Either the given range of flux is recorded meaningfully in the file, or it isn't. The file comes out of the camera, so that's that. <p>Saying that the dynamic range "out of camera" is different from the processed dynamic range doesn't mean <i>anything</i>. Except possibly that your monitor is too crappy to reproduce that range, your eyes are insufficient to see the detail in either end of the dynamic range on your monitor, that you've noticed your printer can't reproduce that dynamic range, or something else along the lines of "my viewing device <i>X</i> doesn't allow me to see the dynamic range of detail recorded by my camera <i>Y</i>." If it is only when you start altering the image that you see detail that previously looked saturated or uniformly dark, then the problem is in your viewing apparatus, not your camera. <p>The "strong advantage" of a camera that gives you the largest visible dynamic range is only an advantage if you really want your camera to compress your images to the dynamic range that your monitor or printer can reproduce - and only if you want your camera to do so without any control or intervention by you. <p>If you are a P&S snapshooter who only takes photos oudoors at noon then this might indeed be a strong advantage. But do I want dynamic range compressed when I'm shooting portraits? Geez, NO! I don't want the person's face to be four large posterized blocks of color because the camera, too stupid to know what I wanted it to do, took a range of 300 values from the middle of the histogram and converted it to two or three values so that it could all fit in the dynamic space of my screen or printer. Does it follow that I want the camera to decide how to compress the image when I'm shooting a landscape? No! Would I trust a computer that doesn't know what the heck I am trying to accomplish to properly place a gradient filter in front of my lens? That is what dynamic range compression is analogous to. <p>How is the camera going to know the dynamic range of your monitor (which may or may not be calibrated, and may have a different dynamic range than some other monitor even if calibrated)? Or your printer? And how will it know the aesthetically best way of accomplishing the necessary dynamic range compression? <p>For the original poster: The native dynamic range of a sensor is determined mainly by the photosite well depth minus the dark current. If that statement doesn't make sense, <a href="http://en.wikipedia.org/wiki/Charge_coupled_device">this article</a> is a good overview; in basic terms, well depth refers to how many electrons a photosite can hold. Some scientific CCD cameras have a dynamic range of several stops more than a digital SLR, but the photosites are gigantic by comparison. So what holds them back is apparently the trade-off between photosite size and resolution in a ca. 35mm or APS-C sized sensor. Link to comment Share on other sites More sharing options...
carl smith Posted January 29, 2005 Share Posted January 29, 2005 It's not utterly meaningless, but you obviously failed to understand the point of what I was saying. It's not worth it to try to clarify at this point. Link to comment Share on other sites More sharing options...
beeman458 Posted January 29, 2005 Share Posted January 29, 2005 "It's not utterly meaningless, but you obviously failed to understand the point of what I was saying." We understand the point you're trying to make but the simplicity of the statement you're making is incorrect in-so-much-as the conditions you allude to, don't exist in reality. How? Film can't be used straight out of the camera unlike digital which can; the memory card being popped directly into a printer and printed without any post-processing applied, using only in-camera settings. One can't do that with film. In your thoughfulness, maybe you're being unintentionally overly simplistic in a field that lends itself best to conversations of a technical kind and these technicals do matter in the final so they can't be left out for the convenience of a web based conversation:) Both film and digital need post-processing after exposure of some nature, whether in-camera and then posted on a monitor or wet processed and then printed. Maybe if you took the time to clarify your point, we might be able to better understand your drift:) Link to comment Share on other sites More sharing options...
rayfraser Posted January 29, 2005 Share Posted January 29, 2005 If Carl does not clarify, then I will assume he fully supports SMaL's Autobrite as the best out of camera dynamic range! Sorry if that appears confrontational - I would gladly delete it should someone object. At least I should if I could. Just trying to increase my understanding by furthering discussion on this thread. Link to comment Share on other sites More sharing options...
jeff medkeff anchorage, a Posted January 30, 2005 Share Posted January 30, 2005 Thomas makes excellent points. I'd just add that taking a CF card out of a camera and popping it into a printer for direct printing, with no user intervention in image processing, is not an example of using an image directly from the camera. (Nor is displaying it on a monitor, for that matter.) The printer is making its own decisions about how to interpret the image in order to reproduce it given what its processors "know" about the ink and paper it is using to make the print. The printer is applying a curve to the image in a case like this, and that may be convenient, and it may even do the job well. But it is not printing an image straight from the camera; it is heavily processing it first. Same with monitor display - the image color and intensity space gets mapped to a range that a monitor can display, almost always with some clipping, some nonlinearity, and some compression. This, again, is a sort of automatic, hidden post-processing that cannot be avoided with digital any more than development can be avoided with film. Now if I "obviously failed" to understand the point you were making, Carl, I apologize. But I don't see how you can use the words "largest out of camera dynamic range" to mean anything, since you can never, ever see an image straight from the camera that hasn't been post-processed one way or another. Link to comment Share on other sites More sharing options...
barnaby_norris Posted February 2, 2005 Share Posted February 2, 2005 G'day all This is a fascinating discussion which I have been following closely, however I am confused about all this post-processing talk. Surely - and correct me if I'm wrong - the dynamic range of a camera (or its sensor) is simply dependent on the range of brightness information it can capture... the ratio between the lowest useful value (ie above noise) and the highest. This is a real-world number - the actual amount of light needed to trigger the lowest and highest possible values. So how can post processing affect this? You've only got the information you've got. No matter how much you push shadows, mask highlights, fiddle with curves and cast mysterious incantations, all you're doing is making the amount of information you started with look more acceptable. Surely no matter how brilliant your post-processing technqiues they cannot improve the actual dynamic range of your camera? So, at the end of the day, I can shoot a nice landscape on colour neg, scan it in a nice scanner, and then fiddle in photoshop to make the overall contrast suitable for my printer or monitor, but (using masks or curves or whatever) keep loads of detail in fluffy white clouds and also see the texture of grass in the shadows. I can do this only because this highlight and shadow detail/information exists in the original image. The BIG QUESTION to me is - can I accomplish the same with a DSLR? Cheers Barnaby Norris Link to comment Share on other sites More sharing options...
beeman458 Posted February 2, 2005 Share Posted February 2, 2005 "Surely - and correct me if I'm wrong - the dynamic range of a camera (or its sensor) is simply dependent on the range of brightness information it can capture... the ratio between the lowest useful value (ie above noise) and the highest. This is a real-world number - the actual amount of light needed to trigger the lowest and highest possible values."<p> So far, spot on:)<p> "So how can post processing affect this? You've only got the information you've got. No matter how much you push shadows, mask highlights, fiddle with curves and cast mysterious incantations, all you're doing is making the amount of information you started with look more acceptable."<p> Ya got that red mark again. Do you shoot darts:)<p> "Surely no matter how brilliant your post-processing technqiues they cannot improve the actual dynamic range of your camera?"<p> That's it, you're a ringer, can't get anything over on you. I'm not playing darts with you anymore:)<p> "So, at the end of the day, I can shoot a nice landscape on colour neg, scan it in a nice scanner, and then fiddle in photoshop to make the overall contrast suitable for my printer or monitor, but (using masks or curves or whatever) keep loads of detail in fluffy white clouds and also see the texture of grass in the shadows. I can do this only because this highlight and shadow detail/information exists in the original image."<p> In a nut shell. Darn, you're good:)<p> "The BIG QUESTION to me is - can I accomplish the same with a DSLR?"<p> In one word, yes.<p> Most folks don't take the time to understand how a digital file, in Photoshop works so they are unaware as to how to meter/expose a digital file and how to bring this information to the forefront when post-processing a digital file/image. Therefore they erroneously think the image has no dynamic range to it and the image file usually has about two more stops, if not more range than most folks realize.<p> <a href="http://www.luminous-landscape.com/tutorials/understanding-series/understanding-histograms.shtml">lumious-landscape.com, understanding histograms</a><p> Understanding histograms is probably the most important aspect of owning a digital camera. The second most important aspect of D-SLR ownership is learning how to bring all the potential out of the digital capture. I tried to outline some of this in my above comments as to using the magic lasso, levels and the history brush to blend the line of demarcation.<p> Digital cameras (APS-c) at this point rival all 35mm film but one, high contrast copy film as that's a speciality film, to the best of my knowledge. How do I know this is accurate, everybody in the fashion magazine industry have converted to digital and they've done so because the quality is there.<p> Digital is a learning process, that's much more tech in nature than the analogue world of the wet film process where one just drops the film off and picks the finished prints up. This is not to say that there's no skill involved in the wet process, I have a strong wet film background, it's just to say that digital is all tech in nature, from start to finish. One needs to gain the necessary skill sets if they're going to be able to fully enjoy the benefits of "today's" digital photographic world. Down the road, this will all change and digital captures, like film, will be good to go, straight to the printer, with minimal intervention on the part of the printer but those days have yet to come and yes, today's photographer has to gain the print skill set of yesterday's color lab tech who did all the darkroom magic for the photographer who lacked print skills.<p> Hope my above helps.<p><div></div> Link to comment Share on other sites More sharing options...
beeman458 Posted February 2, 2005 Share Posted February 2, 2005 Sorry. Overly sharpened image was posted. This one's a bit better.<div></div> Link to comment Share on other sites More sharing options...
barnaby_norris Posted February 3, 2005 Share Posted February 3, 2005 G'day again Ah so I was on to something, eh? Well there's a first time for everything :). I guess the upshot is that a given DSLR has a certain amount of dynamic range, but the real skill is to use this dynamic range most effectively. I just read this fantastic article at http://www.normankoren.com/digital_tonality.html which is very techie, but totally gets to the bottom of this. (Sorry if this link has been referred to elsewhere, I'm new to this forum). I guess it's time for me to quit yabberin and go out and shoot somethin' on different formats, and see what happens! Cheers Barnaby Norris Link to comment Share on other sites More sharing options...
beeman458 Posted February 3, 2005 Share Posted February 3, 2005 "Ah so I was on to something, eh? Well there's a first time for everything :)."<p> And such an eloquent writer for "just" a first timer:)<p> <a href="http://www.normankoren.com/digital_tonality.html">"Making fine prints in your digital darkroom Tonal quality and dynamic range in digital cameras" By Norman Koren</a><p> Thanks for the above link. In the case of digital photography, it's all about reading and getting techie info into the ole gray matter. The read will be most appreciated. Most digital aficionados want to just "cast mysterious incantations" as their form of a digital solution and unfortunately, digital hasn't risen to that level of sophistication of yet. But we all eagerly look forward to that day arriving. Until then, it's read, read, read, shoot RAW, convert in BreezeBrowser, the magic lasso tool, levels and the history brush for my "mysterious incantations". Link to comment Share on other sites More sharing options...
trobee Posted February 4, 2005 Share Posted February 4, 2005 I bought a Fuji S3 recently and am amazed at the dynamic range. Better than anything else I've seen at any price. Link to comment Share on other sites More sharing options...
beeman458 Posted February 4, 2005 Share Posted February 4, 2005 "I bought a Fuji S3 recently..." And when are they going to be coming out with an S3/c mount version? I'm not quite ready to throw all my Canon glass overboard yet:) But the S3's spec for dynamic range, sure is interesting. Link to comment Share on other sites More sharing options...
trobee Posted February 7, 2005 Share Posted February 7, 2005 yep Fuji could rack up with a S3/c Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now