The perfect exposure

Discussion in 'Canon EOS' started by marcelo_schiavon, Feb 21, 2013.

  1. Hi,
    This is my first post so please be kind. :·
    I'm looking for a way to analogically expose my pictures at site. I'm guessing that the only way to achieve something like that on a digital Canon is to have some type of lens that allows you to manually and analogically expose at least one of the factors, say time value or diaphragm aperture. Also it should be a somewhat short mid range zoom lens? Like a 17-80mm.
    Is there a lens like that? I have a Canon 7D.
     
  2. The 7D lets you set the aperture, shutter speed, ISO, and many other exposure factors with any EOS or compatible lens,
    so I, for one, am not entirely sure of the question. You can set any of the parameters to whatever you like in "M", Manual
    Mode. There are no aperture controls on the Canon lens, though, and the only place you see the shutter speed displayed
    is in the viewfinder or the top LCD, or, in certain modes, the rear LCD. But if you want to expose at f/4 for 1/250 all the
    time, you certainly could.
     
  3. I have no idea what you mean when you say you want to "analogically expose" images.
    As for standard zooms for your 7D, there are lenses with 15-85mm, 17-50mm, 17-55mm, 24-70mm, and 24-105mm ranges currently available.
     
  4. I'm guessing you are referring to setting your camera for multiple exposures with only one variable changing (such as shutter speed or f-stop). As Larry says, your 7D can do that (and even meters for your setting, whatever it is, in M), though all those settings are addressed on the camera body. I'd suggest reading your manual in reference to using 'M' mode.
    Since the camera controls all those aspects, every EOS Canon lens allows you to manually set the exposure settings (on the camera).
    The lenses Mark mentions are only some of which encompass that range of FLs, Sigma makes a 17-80/2.8-4 (I think), Canon has a 15-85, a 17-85, an 18-55... the list goes on...
     
  5. Maybe you mean "automatically?" Your 7D will do that too. In fact I feel safe in saying ANY digital SLR will expose manually or automatically. Also ANY digital SLR will give you three possible automatic modes:
    Aperture priority (Av), whereby you set aperture, and speed is adjusted automatically
    Shutter priority (Tv), whereby you set speed, and aperture is automatic
    Program, whereby both aperture and shutter speed are automatic at the same time.
    There are other variants as well -- auto ISO and various program modes.
    Any native Canon EF or EF-S lens will give you automatic aperture capabilities. Most modern third-party lenses will do this too. Only manual lenses (mostly vintage lenses with manual focus and aperture) will lack automatic aperture capabilities.
     
  6. Marcelo, I also don't understand the phrase "analogically expose". I have recently been watching a lot of extended presentations by very well regarded professional photographers on B&H Photo's YouTube channel and just about every one of them primarily shoots in Aperture Priority (Av). Av lets you control depth of field and the camera determines shutter speed to properly expose the image. Then these pros go to the histogram to see if the shutter speed setting the camera picked is giving them the histogram (which shows the range of colors between pure black and pure white that are being captured) they want; if not they typically use exposure compensation to adjust, moving the exposure compensation to a + setting if they want to push the histogram to the right to pick up more bright colors or to a - setting if they want to push the histogram to the left and pick up more dark colors.
    Once they have the histogram where they want it, if the light isn't changing and they're working a scene and taking multiple shots, they will lock in the camera settings via either Exposure lock or changing to Manual and locking in the Aperture and Shutter Speed previously determined through the process I described. That prevents the camera bouncing the shutter speed around based on the light where it is momentarily being pointed.
    Here is something you might find of interest from long time professional John Shaw on exposure. His advice to push the histogram to the far right contradicts the advice of at least one prominent landscape and nature photographer who says he usually shoots with exposure compensation at -2/3. I emailed Mr. Shaw and asked about this conflicting advice; he responded that the other photographer is simply wrong. With film you could get more saturated colors shooting slightly underexposed, but with digital you do not. With digital, whether you want the final image (when shooting RAW and then processing) to be bright, dark, or somewhere in between, you will get the most data to work with by exposing to the far right and then processing the image. As he says in the piece linked below, you introduce noise by pulling up shadows to show more details; with ETTR you're avoiding that as much as possible, and you don't introduce nearly as much noise pulling down highlights as you do pulling up shadows. And Shaw says camera makers tend to be very conservative in showing the "blinkies" on overexposed areas (because people hate overexposed images), that with most cameras if you're getting just a few blinkies they won't show as blown highlights on your computer monitor and you should still move exposure compensation up slightly.
    http://www.johnshawphoto.com/ettr-to-the-far-right/
    I've just started trying Shaw's approach, shooting in RAW, and it works beautifully.
    As far as lenses go, with your camera you can get properly exposed images from a wide variety of Canon lenses and 3rd party lenses made for Canon cameras.
     
  7. david_henderson

    david_henderson www.photography001.com

    I'm sorry but in common with others I can't understand exactly what you require and I'm not sure guessing will help either of us. Could you have another go at explaining please?
     
  8. Don, I think the difference between the two approaches might have to do with whether they're looking at the right-most peak in an RGB histogram or the single peak in a luminance histogram. If it's the latter, you really do have to expose about -2/3 stops from right to maintain saturation in a sky, for instance. However, if these guys are on auto, all bets are off the table. There's no telling what the landscape photog is really doing, based on your description (and possibly even his).
    Even if we're talking about an RGB histogram, color and saturation may not be correct at that last fraction of a stop, where the response curve is rather abruptly rolled off. If the brightest thing in your frame is a blue sky, then you'd better back off at least a half stop from the right, unless you love milky cyan skies. If the brightest object is a bright cloud, OTOH, you can expose much closer to the right. It depends on what you're shooting. In the end, I rarely find noise so monumental an issue (especially during daylight / at low ISO / with the luxury of a tripod / etc.) that I feel compelled to grope for every last fraction of a stop on the righthand side of the histogram. If the shadows are that far down, it's far better to take a second frame on a tripod, with longer exposure to comfortably capture the shadow detail, and then to combine the two in post.
    FAIW, I have to question the usefulness of first shooting in auto, then working out an EC, which wouldn't necessarily apply well from one frame to the next, and then going to manual to lock a setting in. The full-manual alternative is so much quicker and simpler: Set the camera to M, set the aperture and ISO you want, frame the image you want to shoot, while adjusting speed so that the match-needle indicator in the viewfinder indicates a "correct" exposure. Take a test shot, and view the histogram. Make corrections, etc. Once you've determined your best settings, they're already locked in for you.
     
  9. Well, shooting raw or not, once exposure is over, the details in the highlight area are gone. If the exposure is under, one may still be able to get some details from the shadow area. So I always try to avoid over-exposure by setting exposure compensation to-1/3 EV (and always use Av, spot metering and raw).
     
  10. David, here's the thing, though: Let's compare a photo of a sun-lit cumulus clouds in a bright blue sky, cedar trees in the foreground, with a photo of a squirrel sitting on a stump, surrounded by cedar trees, no sky in the picture. If you set -1/3 stop EC, Av, spot metering, RAW and spot off of the squirrel, the exposure probably won't hit anywhere near the right of the histogram. Although it might be exposed correctly in a textbook sense, it could be exposed more for better results, owing to the low contrast. And with the first picture, if you spot off of a stand of cedar trees, your sky will be hopelessly blown out with the same settings. Automation will fail you, and there will be no recovering from it by virtue of the fact you're shooting raw.
    The whole idea of EC is to correct for how we estimate our automatic exposure will be screwed up. If we're putting that much thought into it already, why not just go manual and get it right?
     
  11. Wow! thanks for the multiple replies! I was travelling offline so I couldn't reply before but I'm already loving this forum. I'm so sorry for the confusion maybe it's my bad English I'm gonna read this all right now but let me first try to clarify what I mean when I say "analogically" expose. What I mean is to have a chance of getting a REAL analog exposure variable. The problem with digital photography is that you're limited to the thirds or the halfs of a given diaphragm value. You move from 2.8 to 3.5 to 4.0 but you can't go in between. You can set your Time Value to something more precise as well to get more to a specific EV but again you move from 1/40 of a second to 1/50 to 1/60 but you can't go to 1/55 or 1/56.

    I would love to have a way of moving through all those values without limits. Specially with the diaphragm value.

    The thing is that with film you control this variable in the dark room but that's way too expensive to be experimenting with. I'm not sure I can even get the right chemicals.
    You may say "you have that ability with RAW images and Photoshop" but I've noticed that when over or underexposing my images on the field by even a third of a diaphragm at least SOME info gets lost. Am I talking thrash here?

    Just finding out that I have some images uploaded from who know when!
     
  12. Thank you all but special thanks to Sarah and Don for your inputs.
    Don, really interesting stuff! didn't know about that! I'm currently reading about the theory behind ETTR.
     
  13. William Michael

    William Michael Moderator Staff Member

    I would love to have a way of moving through all those values without limits. Specially with the diaphragm value.​
    One way to do that would be to get a manual lens with a ring aperture on it: you might need to modify the lens’s aperture ring such that there are no click stops.
    Another option, (though not controlling the aperture) would be to use a Variable ND filter, like the Singh Ray.
    But both of these will not have a tally (a number) directly attached to the image you make – what I mean is, you might have a lens or a filter with and (in theory)an infinite number of possible increments: but you won’t have any 'number' to associate with each of those increments.
    ***
    But the question I’d ask is why would you want smaller increments than ⅓ Stops? – Sure I have a light meter which reads 1/10th stops - (I think) – but for what practical use?
    You wrote:
    'You may say "you have that ability with RAW images and Photoshop" but I've noticed that when over or underexposing my images on the field by even a third of a diaphragm at least SOME info gets lost. Am I talking thrash here?'​
    I don't think that you are talking trash but I question how significant and how useful having the choices you want, will be.
    Assuming that you have any scene, which has ABSOLUTELY STABLE LIGHT: and you make a series of nine shots each at ⅓Stop increments.
    Then you find the one shot, where you have ‘lost’ the information you want.
    So therefore, you use the shot either ⅓ over that or ⅓ under that and use it.
    The question to address is - what is the REAL WORLD - PRACTICAL BENEFIT of having a choice of two other shots which are not ⅓ Stop difference, but rather a ¼ or a ⅙ or even an ⅛Stop difference?
    WW
     
  14. Sarah, John Shaw is referring to the histogram on the camera, which I think for virtually all cameras is an RGB histogram.
    David, Shaw's point is that most camera makers are conservative about blown highlights, so that when you see only a smidgeon of "blinkies" on the histogram of your camera, the chances are good that you will see no blown highlights, i.e. no lost detail, when you open the same image on your computer. So Shaw is suggesting that for most cameras, in most situations, from where you get your first blinkie you can generally go +1/3 or +2/3 or even +1 on exposure compensation without actually having blown highlights when you open the image on a computer, despite what the histogram on the camera shows. I have found this to be the case on multiple occasions in the last few days of trying what he suggests. And he's not suggesting that you have to or even should leave the finished image at that exposure, he's saying that's the point at which you get the best RAW data to work with. But he's not saying blindly do it, he's suggesting testing what the histogram on your camera shows versus what your computer shows and if there's a consistent gap, adjust your camera exposures accordingly. If there's not, don't.
    Here's John Shaw's bio from his website:
    "John has been a professional nature photographer since the early 1970s. His work has been published in many publications and books, including National Geographic, Nature’s Best, National Wildlife, Audubon, Outdoor Photographer, and many others. In 1997 he received the first-ever Outstanding Photographer Award given by NANPA (North American Nature Photography Association). Nikon chose him as a featured Legend Behind the Lens in 2002, while Microsoft designated him an Icon of Imaging in 2006. He has been part of Epson’s Stylus Pro fine art print makers group since 2001.
    John has published six books on nature photography, plus six eBooks on Photoshop and Lightroom. He has photographed on every continent, from the Arctic to the Antarctic, from Provence to Patagonia."
    When somebody with his credentials offers free online advice on photography via a blog, and will even respond to email questions, I'm inclined to pay attention and at least give his methods a try.​
     
  15. Sarah, John Shaw is referring to the histogram on the camera, which I think for virtually all cameras is an RGB histogram.​
    Not necessarily. You have the option of displaying either on Canon cameras, and the older Canons only displayed luminance.
    David, Shaw's point is that most camera makers are conservative about blown highlights, so that when you see only a smidgeon of "blinkies" on the histogram of your camera, the chances are good that you will see no blown highlights, i.e. no lost detail, when you open the same image on your computer. So Shaw is suggesting that for most cameras, in most situations, from where you get your first blinkie you can generally go +1/3 or +2/3 or even +1 on exposure compensation without actually having blown highlights when you open the image on a computer, despite what the histogram on the camera shows. I have found this to be the case on multiple occasions in the last few days of trying what he suggests​
    The unstated basis of this argument is actually the highlight recovery capabilities of many applications such as PhotoShop. Unfortunately when highlights have to be recovered (when one or two, but not all three channels are blown out), their reconstruction is essentially "colorized," based on hue and saturation of the nearest pixels that are not blown out. I find this a very dubious practice. I'm sometimes grateful for highlight recovery when a frame would otherwise be toast, but I think to exploit it with every exposure is just asking for inaccurate highlight colors and an abundance of spoiled shots, all in the name of rather small improvements in shadow noise. (There are much, much better ways to handle shadow noise!)
    Marcelo, I agree with William that a third of a stop is already a pretty small increment. If you want a lens with a continuously variable aperture, it's going to have to be a pretty old one. I collect cameras and have many such lenses. They all seem to be pre-1970's. You would also need to meter with an analog meter, e.g. the ever popular Gossen Luna Pro CdS. Still, I would have to join him in asking, "Why?" There is honestly little need to break into a cold sweat over a third of a stop of underexposure.
    BTW, film is not a particularly expensive endeavor if you don't mind B&W (which I see from the photo at the bottom of this page that you do not). Buy it in 100 ft rolls, and bulk-load the canisters yourself. All the chemicals are still available to do your own darkroom work. The cost is not really all that significant. With digital the cost is in the equipment. With film, it's in the materials (the camera bodies being far cheaper).
     
  16. david_henderson

    david_henderson www.photography001.com

    If you can set a cameras exposure in 1/3 stop increments, then assuming that you make optimal decisions you need never be more than 1/6 stop away from the "right" exposure, and most often you will be less. Given that its tough to see the difference that 1/6 stop creates I can quite understand why camera manufacturers have not been concerned to facilitate continuous exposure setting, and why photographers tend not to fret about the issue.
    The difficult thing about exposure isn't what your equipment allows or doesn't allow you to do. Its making those optimal judgments. Added to which there is of course the fact that sometimes the most accurate exposure doesn't necessarily provide the image the photographer wants to see- and photography is supposed to be somewhat interpretative rather than just recreating nature. I think you're chasing after something that doesn't much matter.
    I should add that as far as I'm concerned you can't change aperture settings in a darkroom. You can certainly increase or decrease exposure by fine increments if you wish but changing apertures has an effect on sharpness via depth of field that you can't mimic later. Neither can you recreate the sharpness effect of changing shutter speeds.
    The discussion on expose to the right and understanding the point at which you really do lose information with your sensor is far more useful to you, I'd suggest.
     
  17. "The unstated basis of this argument is actually the highlight recovery capabilities of many applications such as PhotoShop."
    Sorry Sarah, you're just wrong. There's no highlights recovery involved. I open photos in Lightroom 4 that showed a few blinkies on the display of my Canon 5D Mark III and they show no blown highlights. I don't have to touch a slider to recover them.
    As far as dubious practices, maybe Shaw's standards aren't as high as yours, LOL. Somehow it hasn't kept National Geographic and all kinds of other publications from running his pictures for decades.
     
  18. Don, I've actually tested my cameras to see when the blinkies blink, and they blink correctly when a channel is blown. They do not blink when none of the channels have been blown. So there's no built in conservatism to the blinkies or the histograms either one. (I've tested the histograms too, just as I test most aspects of my equipment.)
    As far as dubious practices, maybe Shaw's standards aren't as high as yours, LOL. Somehow it hasn't kept National Geographic and all kinds of other publications from running his pictures for decades.​
    So you're implying that you've looked over my work and are not impressed. Fine, and I don't really care. Don't buy my work then. Who the hell are you anyway?
    I've not been in the game, professionally, as long as Shaw, but I bet I've been doing digital photography professionally as long as Shaw. I would not be so arrogant as to say another established professional photographer is "wrong" (as Shaw has arrogantly done). In fact I will say there is really no "wrong" way to expose an image, as long as the photographer is achieving whatever it is he or she is trying to achieve. (There are legitimate differences of opinion amongst expert photographers and legitimately different approaches between them.) I do believe there are ways that increase one's success rates, though, and I suspect Shaw loses or compromises a lot of shots by doing what he does. I have a different, more conservative, and probably more labor intensive workflow that works well for me. Enough said.
    I think I'm safe in saying one thing, though: Unless all your photos have a very similar dynamic range (e.g. product shots for ebay in a light tent), there is no SINGLE method that is going to give you the best results for all of your photographic scenarios. It is a very naïve and unsophisticated photographer who thinks there is.
     
  19. One thing to remember with ETTR is that it works only if you have enough light to play with. I.e., it's no use pushing the histogram to the right by increasing ISO for example (which auto settings can/will do): the extra RAW info benefit will be negated by the higher ISO noise. I'm sure ETTR is valuable in special situations, like say you're shooing in bright light or are doing stills/landscapes on a tripod with slow shutter speeds. But in practice, I've found it to be of little practical value in day-to-day shooting and just adds a step in post processing.
    WRT to the OP's original question, cine lenses (vs. photo lenses) have continuous adjustment of the t-stop (they call it t-stop to represent transmission). However, that kind of fine tuning is really useful for video, not for stills; and the cine lenses cost a ton (often tens of thousands of dollars). In any case, there is almost no value these days in going to less than 1/3 stop if you shoot in RAW since post processing allows so much latitude to recover from exposure errors (like with Adobe Camera RAW, Lightroom, etc.).
     
  20. Sarah, I implied nothing about your photography, know nothing of your photography and based on the arrogance and condescending nature you display here I have no desire to. I just find your implication that you have higher standards than a very accomplished photographer like Shaw ("suspect he loses or compromises a lot of shots by doing what he does" is your latest blather) to be laughable.
    It's clear you didn't bother to actually read Shaw's comments on ETTR at the link I posted before you mischaracterized what he actually said on multiple occasions. He said that you don't want blown highlights, that it doesn't apply in every single lighting situation, etc. But obviously there's no reason for you to read what anybody else has to say since you already know it all.
     
  21. Don, if you want to belittle me about not being as accomplished professionally as Shaw, I have no defense. I am not. (Are YOU?) However, that in no way means I don't know what I'm doing or that Shaw knows more about what he's doing than I do. Honestly, you speak of Shaw as some sort of omniscient deity and laugh (out loud, apparently) at the notion that someone so pitiful as I would dare question the Word of the Almighty Shaw. And that, my friend, is pretty darned funny. You must admit, yes? I try to offer you helpful thoughts/advice, and you insult me for it. Good on you, and welcome to PhotoNet!
    FAIW, I've already seen and read all these expose beyond the right arguments before. They've been kicking around for a few years now. And yes, I've already done my own experimentation and drawn my own conclusions. It would seem silly to test the same thing over and over, expecting different results. I apparently do things differently from Shaw and offer no apologies for it. And to be perfectly clear, he apparently does things differently from me and owes no apologies for it. If we were to swap methods, we would probably both be worse off for it. So all this talk of yours of "higher standards" is a pretty unfair mischaracterization.
    Finally, I believe you totally misread my comment about those who seek to apply a single exposure method to fit to every situation: To be clear, I was NOT referring to Shaw, nor was my comment meant as more than an observation.
     
  22. And a correction/retraction: Indeed, Shaw's argument is NOT related to highlight recovery, as indeed you said, Don. I don't take issue with anything Shaw wrote, although I do regard the closely related highlight recovery method with considerable apprehension.
    I have not yet owned/used any camera that did not give me accurate blinkies and histograms. HOWEVER, both are actually a depiction of the information in the thumbnail jpg. Camera settings will affect both greatly. In cameras I've owned/used, the top end is kept accurate and the bottom end adjusted accordingly, at least on the "picture style" modes I've tested. However, I did not test any higher contrast modes, and that might be where the disconnect between blinkies and clipping occurs. I would submit that an easier, more intuitive alternative to exposing a certain number of stops beyond the right is to simply get the blinkie and histogram representation on the LCD correct. I've done this in my various Canon cameras by dialing down contrast as far as possible, which also shows more of the shadow detail on the left side of the histogram. I honestly have no idea how any of this translates to Nikons.
    ... but you may want to run this past your deity before believing it.
     
  23. As said above you just need an older manual lens with an aperture ring on it. Also you can use the ev dial to move up and down a third of a stop. It sounds like you have a very particular purpose for having this ability? Because in actuality you can get nigh on spot on exposure using just the camera on board equipment or if you want to really get precise, a handheld incident meter which is very good for establishing ratios let's say on a face and/or 1% spot meter. Which really lets you choose what tone you want to be your "middle" grey. Curious about why you want the stepless aperture for metering.
     
  24. Sarah, there's no need to be a condescending, intellectually dishonest troll even as you backpedal. I don't recall referring to Shaw as my deity. A couple of weeks ago I had never heard of him. I came across his blog after watching an extended YouTube presentation by Art Morris on bird photography. Art is considered one of the world's leading bird photographers and I found his presentation educational (though I'm sure you could have corrected him on some points and given him some advice to improve his photography), so I took a look at his blog and saw a post in which Art, who has been a professional photographer for 30 years or more, referred to Shaw as his idol when he started shooting and said Shaw had been a mentor and a friend. Shaw posted a piece or two on Art's blog, I believe including the one on ETTR, which you've now finally read apparently, even though there was of course nothing in that or anything written by anybody else on photography that you could possibly learn from.
    Shaw's piece on ETTR seemed somewhat at conflict from something I'd heard from a very prominent landscape photographer on another YouTube presentation and the commentary on Shaw's blog was closed on the topic, so I emailed him and he was nice enough to respond back. He said nothing derogatory about the other photographer, just said that he was wrong on that issue. Presumably the other photographer disagrees. I found Shaw's rationale more persuasive, having personally shot a lot more underexposed than overexposed images (I wish I could shoot them all perfectly exposed as I'm sure you do), have tried ETTR a few times since then and found it helpful.
    I thought the piece on ETTR might be of interest to the person who asked the question that originated this thread, so I posted a comment that included a link to it so if interested he could go directly to the source and draw his own conclusions. I did not solicit your feedback and in my first brief response to you I tried to let you know that. This is the first photography forum I've spent any time on, but I've seen your type on a variety of other forums, the type who positions himself or herself as the resident expert that everybody should bow down to. If that's what this forum is about I won't be here long. If giving greater weight to professional photographers who have written books and had their pictures appear in top publications over decades than to self proclaimed experts on the internet whose work I know nothing about means I'm worshipping a "deity", so be it.
    You seem like a last word type of person, so go ahead and have it. And in case I haven't made it clear, based on your posts on this thread and others, I am not interested in any advice from you on any topic. If you will do me the courtesy of ignoring anything I post, I'll do the same for you.
     
  25. No, I'll let all the last words be yours, Don. I thank you very much for each and every one of them, each one like a beautiful pearl!
    • wrong
    • As far as dubious practices, maybe Shaw's standards aren't as high as yours, LOL. (Nice!)
    • there was of course nothing in that or anything written by anybody else on photography that you could possibly learn from
    • arrogance
    • condescending
    • laughable
    • you already know it all
    • self proclaimed experts
    • condescending
    • intellectually dishonest
    • troll
     
  26. The vast majority of the times the auto exposure gets it right.
    I've learned over time in situations with a wide range of light levels to do 3 exposure bracketing, not for HDR, but to allow me to choose the right show for what I want to work with. It has saved me numerous times with situations around saturation and exposure areas. On rare occasions I have used the 3 exposures for HDR.
     
  27. Don, may I suggest that once one is in a hole, it is time to stop digging? Sarah is one of the most helpful and generous posters in the EOS group, and has thoughtfully participated in years of discussions. Childish and insulting commentary such as "As far as dubious practices, maybe Shaw's standards aren't as high as yours, LOL. Somehow it hasn't kept National Geographic and all kinds of other publications from running his pictures for decades." has no place here, and completely disqualifies you from opining on other's styles. "Intellectually dishonest troll..." Now that is just hilarious, doubly-so, if irony is your goal. ;-)
    Chillax,
    rt
     

Share This Page