I have heard that the dynamic range of this camera is great, that you can't make a bad exposure with this camera and read in many different places that (forgive me if I'm not says this right) that DXO says it's testing showed a dynamic range of 14 EV.
I recently purchased the new Sekonic L478DR and used the DTS software to find the dynamic range of the camera and calibrate my meter to my camera. Now the resulting dynamic range came up at 6.7 EV (give or take the little bit of leeway they give you for a safety margin on both ends of the scale). That's not anywhere close to what I was expecting.
Please forgive if I'm getting this all wrong. I have too many hobbies and trying to keep up with all of them is a constant challenge, I'm always up for a challenge!
Thu 19-Dec-13 09:31 PM | edited Fri 20-Dec-13 01:46 AM by KHR
It sounds like I've offended you in some way by your statement. I didn't BUY IT TO TEST IT. I've had it for over a year now and just purchased a new light meter with the exposure calibration capability to get the most out of my equipment and was surprised with the results.. Now if trying to get the most out of my equipment and asking question about it is a problem, then I'm always going to be a thorn in your side when you read my posts! Because I will ALWAYS strive to achieve the best I can! I joined Nikonians for just that reason, to get advice and technical support from others when needed and to become a better photographer.. So I could produce better images and hopefully I can pay it forward to others in the future. Not to be criticized for asking a question.
I have had it for over a year. But I also have had to send it back for repairs, once for focusing issues then once because it would occasionally lock up on me which was repaired about 8 months ago and worked fine until this last weekend where it locked up on me again. So I just sent it away again today for repairs. I must have just ended up with a lemon. While it's in, they are going to look a the sensor as well.
The reason I asked the question is because of the results of the Sekonic DTS software. It's not what I expected. I'm just trying to figure out what should I expect from the camera and is the DTS software usually pretty accurate? Or where the results of the DXO testing unrealistic to the actual results in the field?
Now something just came to mind. Are they speaking of in camera dynamic range or the capabilities that can be achieve in post processing with this sensor?
>The reason I asked the question is because of the results of >the Sekonic DTS software. It's not what I expected. I'm just >trying to figure out what should I expect from the camera and >is the DTS software usually pretty accurate? Or where the >results of the DXO testing unrealistic to the actual results >in the field?
I know nothing about that software, but the reason I think it's DR is so good is that I can take a low ISO exposure (say 100-200) in lighting where I would normally need to do multiple exposures for an HDR, and pull up the shadows and down the highlights and get decent shots. FAR better than any prior camera, and at least a stop or so better than the D4.
I haven't tried to scientifically measure it, but the reality is there's all sorts of detail in those shadows that normally is just blank/grey mush.
You'll get different results with the DTS software depending on how you process the images and what tone curves you apply. Quite honestly, that feature of the meter isn't that useful unless you're only shooting jpegs. The D800 has an extremely good dynamic range, no matter how it's measured.
So it is more of what you can do with it in post. I definitely had noticed that there was more detail in the shadows when post processing which has changed how I shoot. I used to try and get as much tonality as possible in camera But now I push my high lights as far to the right as possible without blowing out anything knowing I can pull up the shadows so much.
Just trying to figure out if I was getting the most out of it.
Fri 20-Dec-13 02:20 AM | edited Fri 20-Dec-13 02:23 AM by GiantTristan
I think yours is a very valid concern. Characterizing the DR of a camera in qualitative terms is not very helpful. The DR of the D800 is definitely much greater than the 8 stops displayed in the camera histogram. The "blinkies" just refer to the in-camera JPEG associated with the raw file and are pretty useless for determining DR.
You can use the program "RawDigger" to display the true raw file of your D800. I find that the camera histogram truncates about three stops on the right side. This means that you can over-expose by about three stops before the Highlights are blown. The limit on the left side is more difficult to asses. Here the limit of DR is determined by the dark current. Usually, one assumes that a signal can be detected, if it is about three times the average noise level. Apparently, the dark current is somehow "nulled" in the camera. From trial and error I conclude that one can extract around two additional stops from the shadows. This would amount to a DR of around 13 EV at base iso. This is fairly close to the DxO value.
I wanted to know what others have see on there cameras and if it was unreasonable for me to assume that I would see that kind of DR when I used the DTS software. I only have one of these cameras and don't know anyone else personally that has one to compare notes with. As I stated, I've been having trouble with mine and wasn't sure if this should be a concern or not? It's off for repairs for the third time now and I have asked if they could check that the sensor was performing as it should while they have it. It's a bummer that it's just before the holidays but at least I kept my D300 as a back up.
I've never thought about testing my D800E and have had some good examples of some good dynamic range with my camera. Here is a couple of test shots that I played with just to see what I could get with LR4.4:
D800E 85mm F/1.8G @ f/1.8 iso 220 1/100th:
Straight out of LR with no edits:
Corrected (about 3.5 stops of exposure added) to get a more evenly lit image:
Another example of the dynamic range:
D800E 50mm f1.4G @ f/2.8 iso 400 1/800th:
Again straight from LR4.4:
Now pushed 5 stops and pulled highlights down a little:
Hopefully these are helpful on what to expect as these I feel are good examples what the D800 is capable of.
Very nice examples. I can honestly can say that I've never attempted to pull the shadows up that much. For one thing I would not be able to achieve focus with any of my lenses in that low of light as the back lighting one in front of the window. I hope that my camera comes back working as it should.
> I hope that my camera comes back working as it should. > > >Kelly
Not sure if this is applicable to you, but my sense is there is the very occasional software-bug driven "lock-up" buried in the D800 firmware (even the most up to date one). The one I am thinking of happens upon image review.
I have always found that pressing the shutter release and taking another picture eliminates the lockup. It happens to me once every 6 months or so. I could be wrong about this, maybe it's operator error on my part . But I do suggest if you get locked-up, just take another picture and see if that does the trick.
The lock up I was having would completely disable my camera. Right after taking an image an hour glass symbol would show up on the LCD and the LED that lights up when data is being read to the card. Now this would last anywhere from a few seconds to several minutes. The first time I sent it in was because I waited for almost 10 minutes until I finally had to turn the camera off. When I turned it back on, the camera would not recognize the card as well as my computer. I had to send the card to the Lexar to see if they could recover my images ( which I must say, they did and also replaced the card with a new one. Great service).
>Hi Steve > >The lock up I was having would completely disable my camera. >Right after taking an image an hour glass symbol would show up >on the LCD and the LED that lights up when data is being read >to the card. Now this would last anywhere from a few seconds >to several minutes. The first time I sent it in was because I >waited for almost 10 minutes until I finally had to turn the >camera off. When I turned it back on, the camera would not >recognize the card as well as my computer. I had to send the >card to the Lexar to see if they could recover my images ( >which I must say, they did and also replaced the card with a >new one. Great service). > >The same locking up started again.
That sounds almost like a card issue.
I've had mine lock up, but more as described by SteveK - it happens only when displaying, and PRIOR to displaying anything (I don't recall if an hour glass appears but the green light is on).
My impression (no data here, just impression) is that it may be at a time of action on the card, as though the camera is starting a new folder, maybe scanning all folders, etc. There are other times when this happens, such as inserting a new card with a lot of data already on it. Consider an android phone - insert a SD card, and it won't let you use it until it scans the file system. I wonder if something isn't triggering the D800 to do the same sort of thing.
I find I can just take another image and move on, and when I look back in a couple minutes it is fine. It has never prevented me from exposing an image (with the buffer not full of course).
It happens rarely to me -- maybe once every 10-15 shoots, never at the beginning, always after quite a bit of shooting. I'm always using a backup card also.
I've come to just ignore it as normal. But it never goes on for minutes (at least not more than 1-2). And it has never affected card readability (but I've never pulled a card out -- did you? Or just power off? I certainly have powered off and back on).
When it first started doing this I thought myself that it may be a card issue. So I removed the CF card and just shot with the SD card for around two weeks before it did again, which eliminated the thought it was the card.
There would be times that it wouldn't happen for a month or two, then it would do it 2 to 3 times in one week. And each time seem to last a little longer.
I just received my camera back from Nikon service for the second time for the same issue and auto focus problems. Now their description of what was done was pretty limited. They said "Checked and cleaned camera (I'm assuming the internal electronics) and checked and adjusted auto focus". So the only thing I know was that I was right about the focusing because now when focusing at f2.8, it's tack sharp now.
>So it is more of what you can do with it in post. I >definitely had noticed that there was more detail in the >shadows when post processing which has changed how I shoot. I >used to try and get as much tonality as possible in camera >But now I push my high lights as far to the right as possible >without blowing out anything knowing I can pull up the shadows >so much.
The whole "expose to the right" adage has become accepted fact, but I personally believe the higher dynamic range of the modern sensors call for pulling highlights back toward the center a bit more. I'm a bit proponent of "expose correctly" or "Expose to the center" instead.
Try pulling down the highlights in something like a performer in a spotlight. Even if you can pull them down to where they don't clip, they end up flat with inadequate numbers of tones (a one-color-forehead type effect). The high end of the tone mapping is is somewhat robbed of detail normally.
Or a more concise way of saying it, with the high DR of the D800, I find I can pull shadows up more accurately than I can recover detail in (nearly) blown highlights.
PS. I've talked to several people who didn't realize that the D800 (and other later Nikons) have a feature to show the histogram of just part of an image. Shift to the RGB histogram display, and zoom in. What you see on the display is just for the portion of the zoom, unlike the regular histogram which is the whole frame. Very helpful in high contrast shots if you want to see where each channel is on the key subject. Again, a performer on stage is a good example. Or land in a landscape that includes bright sky.
Sorry, Kelly, I guess it came across that way. Yes, I get very amused by the focus on equipment and the urge to measure everything. A camera is for making photographs. It either does that or it doesn't. It all comes back to Cartier-Bresson's statement:
"I am constantly amused by the notion that some people have about photographic technique – a notion which reveals itself in an insatiable craving for sharpness of images. Is this the passion of an obsession? Or do these people hope, by this trompe l’oeil technique, to get to closer grips with reality? In either case, they are just as far away from the real problem as those of that other generation which used to endow all its photographic anecdotes with an intentional unsharpness such as was deemed to be “artistic.”
What matters a whole lot more than the dynamic range the camera can handle is how well you can compose a photograph and how familiar you are with whatever equipment you're using. As Rick pointed out, you're going to get different results from different ways of measuring -- not just dynamic range, but most other measurable things about a camera. Again I say: What difference does it make? Does the camera do what you want it to do? If so, relax. If not, either find a different camera or learn to MAKE the camera do what you want it to do.
I think you are referring to, Nikonian Bill Claff His research into Photographic Dynamic Range for determining a real world effective DR that is somewhat less than the typical Engineering Dynamic Range. Bill's measurements suggest that Nikon is significantly better than other brands in low ISO DR with the D600/610 and D800 leading the industry.
It is fun to post process D800 files, they are so flexible in retaining so much clean detail and color from essentially black frames. The first time I tried to rescue a very dark venue shot where the flash did not fire, seemed like magic. From totally black image to a usable printable image with a single slider swipe in LR. Before that the D7000 was only beaten by the D3x but not by much. One feature that never fails to impress it the DR, more than resolution.
>I have heard that the dynamic range of this camera is great, >that you can't make a bad exposure with this camera and read >in many different places that (forgive me if I'm not says this >right) that DXO says it's testing showed a dynamic range of 14 >EV.
Isn't it great that we have so many camera choices? To respond to your post, here are my thoughts based on my own personal experience and approach to photography. 1. The dynamic range capability of the D800 Series is the best (widest) I have ever experienced from a Digital Sensor. 2. Actually, you CAN make bad exposures with the D800 Series cameras … I guess I could say that I certainly have been able to achieve that result. 3. While the dynamic range has, indeed, been measured (by others) as high as 14 EV's, that result only occurs at the native base ISO of 100. My understanding is that, in general for most shooting situations, you can expect a 1 EV reduction for every increase in full ISO value…i.e for an ISO value of 200, the dynamic range could record as much as 13 EV's and so on. 4. After some experimentation and evaluation, I tend to expose to the right (ETTR) and then draw information out of the shadows if necessary. Because of my D800E, I am not going the HDR multiple exposure route nearly as often as before for my "Extended Dynamic Range" (EDR) images. 5. I believe that the D800 series camera is best shot on a super sturdy tripod with the best "glass" you can find, with your camera set at 100 ISO while employing a remote release in an environment with no air. I appreciate that could be limiting and that is why I am fortunate to have more than one camera.
I performed the same test using the Sekonic L758 and also used the DTS software to measure the dynamic range of my D800E using studio lighting, 18% grey card, and the X-rite’s Color Checker Passport. The results of the dynamic range I measured was at 12.6 EV.
So I think it may have to do with the lighting you use!
>Hi Kelly, > >I performed the same test using the Sekonic L758 and also used >the DTS software to measure the dynamic range of my D800E >using studio lighting, 18% grey card, and the X-rite’s Color >Checker Passport. The results of the dynamic range I measured >was at 12.6 EV. > >So I think it may have to do with the lighting you use! > >Rick
I also used the Color Checker Passport, and did a white balance using the card in the CCP. Did the first JPEG at f8 @ 1/15 sec, ISO 400. Then the next 2 at f2.8 and f22. Then ran it thought the DTS software In the basic mode and the results were 6.7. Now not sure if I was doing anything wrong. I followed the steps as I watched a webinar on using this software. I have sent an email to Sekonic with question over a week ago and have not heard back from them yet. I asked if using the advanced section would give me more accurate results and if there was any resources on this process.
When my friend and I performed this test, we followed the exact procedure described in the Sekonic Data Transfer Guide. My friend who owned the studio and the Sekonic L758 light meter helped me set-up the shooting with both studio lights calibrated at the same intensity and both the same distance from the camera at 45º from the target at one meter. The target was placed on a non-reflective service to eliminate any reflection back into the lens and was perfectly perpendicular to the camera and the same height.
We performed this test several times using Nikon's 16-35mm, 24-70mm and a 70-200mm lenses at various focal lengths. We used TIFF files instead of JPEG's, thought this would provide us with the best tonal range. We used ISO 100, and f/2.8 on all lenses. All three lenses performed very well with a range of 11.3 to 12.6 EV.
I'm very satisfied with the results of the testing we performed on the D800E and the lenses we used. So under ideal conditions, I can see getting a dynamic range of 14 EV that DxO Labs stated.
I followed same exact procedure described in the Sekonic Data Transfer Guide. We tested this camera like a science project, even measuring both studio lights to have them matched in intensity. The target and both studio lights were measured down to the millimeter and the same height. We wanted to make sure there was no glare or light refraction back at the camera.
Not all cameras will have the exact same test measurements even if they are the same model and lot code, I know this from experience. All the electronic and mechanical components used in the design of the D800E have tolerances that are not all the same and that also applies to all test equipment like the Sekonic L758 light meter, they will all slightly differ.
I used continuos lights that aren't very bright. Ill do the test again when I get it back from service with my flashes or outside on a bright day. As stated that may have been all the DR that was in the image! And I'll used the lowest ISO possible. Does anyone have an opinion on weather the three stop difference matters if it's achieved with the shutter speed or the f stop? And is it better to shoot tiff?
As an example, this photo before conversion to JPG has a DR of about 13EV according to RawDigger. There's not much data at the bottom end, but, it's there.
Taken under terrible conditions: about 2 or 3PM local time, in the Australian summer. A couple things to notice: there's quite a bit of detail in the planters up front. You can also see the moon. Use the trellis as a reference, and look between 1 and 2 oclock.
TL;DR - pictures like this make me think the dynamic range of the 800 is not a concern.
>The human eye can see stars and bright sunlight. This >corresponds to a DR of about 1,000,000 or 20 stops. This means >there is only a 7 stop difference between the human eye and >the D800.
I don't think that's exactly fair, to compare what is essentially a movie camera to a still.
The D800 can see really dim stars, and bright sunlight, just not in the same image. And neither can the human eye. It can only see about 5-8 stops at any given time, from what I understand, depending on your age and eye sight and type of image.
For a given instant in time captured by both, I think the D800 has the eye beat in pure dynamic range.
There is a lot of confusion regarding the dynamic range of the human eye. The 20 stop DR is derived from the simultaneous viewing of the full moon and "weak" stars. In order not to violate the copyright, I cannot quote directly. The link is http://www.clarkvision.com/articles/eye-resolution.html . This is quite an interesting article and apparently scientifically sound.
>There is a lot of confusion regarding the dynamic range of >the human eye. The 20 stop DR is derived from the simultaneous >viewing of the full moon and "weak" stars.
It's an interesting article (and note if you have trouble with the link remove the period).
What I wonder is whether it is reasonable to test in such a fashion, as the very bright moon causes that area of the retina to light-adapt. If curious try doing just this, then shift your vision so the sport where the move was, is over a star, and try to see a dim star. It's also a bit different of a test as if the stars are off center, they are in the portion of the eye more sensitive to light in general (your peripheral vision can see in less light than the center of your eye).
But as it really says -- it's not testing two sensors, the eye is just different.
I still stand by the idea that the eye in a normal scene cannot see as much as a modern sensor. Look at (say) a bright waterfall scene in sunlight with trees or better still caves surrounding it. Take a photo of it, and you can raise the shadows and see what's in very deep shadow on the image, but your eye cannot see into those same shadows.
But... I haven't got scientific references to quote.
Happy New Year! Off to kill a few more retina cells.
You misread the article, the DR of an eye is at best with dark conditioning about 15 stops. The 20 stop suggestion was total range when conditioning the eye to very dark and very bright, but not at the same time. But few people can even manage 15 stops. I studied human perception a lot, mostly sound since my career was recording music and I was very interesting in developing techniques to help paint aural images in a listener's mind. It tended to work, with many gold and platinum record awards(my studio where I trained most of the engineers had over 200) and an Oscar nomination. That study carried over to vision as well and discovered that the brain is very much involved with image stacking and use of memory to create the impression of continuous focus and wide DR. If the same method of capturing an image and seeing one were used, modern DSLRs would have the advantage. The eye supplies the data that is built into a scene memory by sampling at different iris diameters to build a DR map, much the way that focus is much shallower DOF than we believe because of the constant sampling at different focus points to build up a focus stacked impression in the brain. A good test is to artificially freeze the muscle of the iris to measure the light level range detected in one iris diameter and it turns out to be rather limited. If you ever had your eyes dilated for a eye exam you know how intolerant you become of bright light when still being able to see mid tones. The DR in that case is down around 8-9 stops. A D800 can handle 13 stops DPR with one aperture setting and 14.4 stops Engineering DR Stan St Petersburg Russia
>The eye supplies the data that is built into a scene memory by >sampling at different iris diameters to build a DR map, much >the way that focus is much shallower DOF than we believe >because of the constant sampling at different focus points to >build up a focus stacked impression in the brain.
Can you cite a couple of articles on that? I'd be interested to learn more, especially the frequency of the iris changes.
I also recently bought a Sekonic L-758DR, my first light meter since my 1960-70 film days. I did do the DTS calibration with the Color Checker Passport and followed the Sekonic webinar with the same DR of 6.7 results. However, no studio lights just every movable light in the house aimed at the target using a 24-120mm F/4 lens on a D800 at an ISO of 200. Not real scientific. Interesting - but I set the clipping points of the L-758DR manually for -5 and +4 stops with the Out of Range icons set about 1/2 stop beyond that. With the icons set for “Outside Range” (custom setting 7) I have a reasonable DR guide.
The real reason for wanting this meter is for the spot function. I am starting to use ND filters and the spot function helps me to understand the the maximum, mid, and shadow zone values of my composition. The ability to select the mid-range reading and having the +/- range displayed helps me to learn to see which filter might be right or even if a filter is appropriate. I expect that in time I will see it au naturel.