RDKirk wrote:
Interestingly, we had an analogous situation with film. The higher resolving films had shorter dynamic range (and less latitude) than the lower resolving films.
I never knew this. I always shot very low ISO print film and transparencies (usually ASA 64 or 25 for the latter!). This is useful information when trying to gauge digital.
RDKirk wrote:
With color negative film, their concept of dynamic range is more about latitude (how much the lab can correct poor exposure) than about dynamic range.
If you similarly pull all the stops with digital--proper exposure, post-processing, et cetera--you can certainly do at least as well as color negative film.
Ok about the latitude, thanks for the correcton, not so sure though about getting more out of digital than film
dhphoto wrote:
The results have been just great, but I'm not expecting miracles out of the gear. I think some here are. I think David's shot is a great example of what dslr's can do.
I agree
Current DSLR's give great results. Although some do beter than others.
Take a look at this D90 sample. Perfectly exposed for the highlights. Meaning, there is still detail in the stained glass. Since I was in a tomb (US cemetary Normandy), I couldn't use flash.
In PP (LR 2.5) I adjusted the fill light by a whopping +60!!! This is the result:
I can show you 100% pixel-peeping crops of how good the D90 files held up such a heavy treatment. I wouldn't dream of doing this with my 5D2. There would be banding all over the place
Current DSLR's give great results. Although some do beter than others.
Take a look at this D90 sample. Perfectly exposed for the highlights. Meaning, there is still detail in the stained glass. Since I was in a tomb (US cemetary Normandy), I couldn't use flash.
In PP (LR 2.5) I adjusted the fill light by a whopping +60!!! This is the result:
I can show you 100% pixel-peeping crops of how good the D90 files held up such a heavy treatment. I wouldn't dream of doing this with my 5D2. There would be banding all over the place
5D2 destroys the D90 in other categories. Different cameras, different strengths and weaknesses. You guys going to keep posting about this or just accept reality?
As I recall color neg film was able to record a vast DR on the actual negative and view it on a light box, but it was impossible to actually print the full range. Print paper just can't do it, so you need to "compress" the range and this usually meant blocking up the shadows. Also, I was never able to scan the full DR from a neg either. Negs were certainly more tolerant of overexposure.
Future Man wrote:
5D2 destroys the D90 in other categories. Different cameras, different strengths and weaknesses. You guys going to keep posting about this or just accept reality?
Sure, different cams with different strengths and weaknesses... But we were discussing DR, right? Just to put everything into perspective, I posted a D90 sample... which blows the 5D2 for these kind of purposes. If you don't like to read it, just move along.
FYI I use a 5D2 with great pleasure for low light event work. I find here it really shines, because of the excellent and very usable ISO1600-6400.
Future Man wrote:
I don't get people saying "Why could I push shadows more with the 5D when the 5DII is supposed to be better?"
5DII is a different camera. It has roughly double the MPs, smaller pixel density, better high ISO performance, video capability, etc.
I agree some people are expecting a perfect camera. Maybe we'll have that in 10-15 years, but for now the 5DII is a dream for me.
Where do the excuses stop though? "D3X doesn't count because it's more expensive" "Point and Shoots don't count because they aren't 21MP" "D90 is worse in every other aspect"...excuses can be made indefinitely...
Dawei Ye wrote:
Where do the excuses stop though? "D3X doesn't count because it's more expensive" "Point and Shoots don't count because they aren't 21MP" "D90 is worse in every other aspect"...excuses can be made indefinitely...
I wasn't making any excuses. I clearly understand there's a banding issue when pushing the 5DIIs files. No perfect camera. And I don't mean to try and derail the thread, it's an interesting topic, but it's covered ad nauseam on here.
I can show you 100% pixel-peeping crops of how good the D90 files held up such a heavy treatment. I wouldn't dream of doing this with my 5D2. There would be banding all over the place
I see banding if I tamper with exposure at very high ISO. But Fill Light does not cause banding on my raws. Just tested it with ISO 1000, moved the slider to 100%. No banding at all and the noise is very good. That is at 21MP:
Mirek Elsner wrote:
I see banding if I tamper with exposure at very high ISO. But Fill Light does not cause banding on my raws. Just tested it with ISO 1000, moved the slider to 100%. No banding at all and the noise is very good. That is at 21MP:
I don't get this type of banding in every situation. I get good results like in your sample pic. But it seems the banding is primarily situated in the red channel. Try to do similair fill light adjustments against a dark/underexposed (containing red) background and then see what happens.
A simple way to determine your camera DR is to shoot a gray card, bracketing exposure to make it a range of tone from near white to the point where detail disappears into the noise. I did it with my 20D a few years ago and got this result:
The card filled the viewfinder and as I stopped down the lens from 2.8 the histogram spike from the card indicated by the red lines moved left. The tone the card was reproduced at and the RGB readings are shown. Conclusion? The camera/lens combo can resolve detail in a range of about 6 stops.
I did B&W zone system work and metered scenes with a 1-degree spot meter. An average flat-lit scene is 10 stops. Cross-lighting will increase the range to around 12, snow or sand even more. Overcast lighting conditions will reduce the range. The way the ZS works is that the zones are tonal values on the print. Negative development is varied to fit various scene ranges onto a single paper grade by making the highlight-shadow density range the same.
Color Print paper can record a scene range of about 5 stops with detail. The latitude of the negative is about 2 stops greater, which is why Granny's Instamatic worked: color negatives overexposed by 2 stops can still make good prints. DR for digital is a bit longer than color negative. It is much longer than transparency film.
The bottom line is any scene over 6-7 stops either needs to be exposed "perceptutally" to make what is most important in the midtones look good at the expense of detail in the highlights and shadows, or have the range adjusted with flash. The way to reduce the range of a scene with flash is to shoot into the shadow side of the ambient and then lift the shadow side independently. Flash can't reduce contrast if it hits sunny highlights and shaded areas at the same time.
YMMV, but I keep a 580ex flash on my camera on a bracket at all times outdoors and use it whenever the range exceeds the sensor. That is very easy to determine:
1) FIrst set exposure using the overexposure warning to keep all non-specular highlights below clipping.
2) Then look at the left side of the histogram. If its piled up on the left and dipping on your foot the range exceeds the sensor and its time to shoot into the shadows and add flash to reduce the scene range, or find less contrast lighting conditions (e.g. open shade).
Indoors situations like a bride and groom next to each other will exceed the range of the camera when one flash is used. Moving a single flash off axis creates shadows which increases the scene range. In that situation the solution is using two flashes in an overlapping key over fill arrangement.
1) Add fill until the darkest detail is recorded
2) Overlap the key light on top of fill to create the highlights until the brightest ones are just below clipping.
Since most people add key light then fill they don't grasp that the contrast is reduced by the fact the key light is actually overlapping the fill. It was easy for me to grasp because I learned flash using two of them in a key over neutral fill arrangement: fill on a flash bracket, key off axis.
I don't have the patience to read the whole thread but I get the impression that it's mostly people who have never seriously used film and only know the crappy lab prints they used to get for vacation photos.
I've been shooting medium format film for a long time and have just recently got into digital with a 5DII. As much as I love the Canon, I'm not getting rid of the film stuff as there's some situations where the Canon is not match for portra 160nc color print film
With film I can shoot a strongly backlit subject on a sunny day right into the sun and (by simply exposing for the shadows) get a beautiful soft blue color and gradation in the sky. Do the same thing with digital and you get a clear white, blown out sky. Mind you, I scan the film on a high end scanner and then process in PS so I'm talking about negative film as a simple raw source for a digital image, not bw film with zone system and darkroom techniques.
By the way, I don't know if this has been noted already but IMO the photo by the OP is not really taken in an "extreme" situation. Shooting the moon and clounds is not that tricky even with a fairly cheap digital compact.
As mentioned color negative film has a longer range (straight line section in the DlogE plot) that the print paper, so you are likely getting much more range out of the scan on a high-end photomultiplier scanner than you would on a color print.
With film I can shoot a strongly backlit subject on a sunny day right into the sun and (by simply exposing for the shadows) get a beautiful soft blue color and gradation in the sky. Do the same thing with digital and you get a clear white, blown out sky.
If you exposed that scene for the shadows with Kodachrome, you'd have gotten a clear white, blown out sky, too.
RDKirk wrote:
Then how about when we compare digital with film, we keep it straight.
What's your point?? I was comparing the 5DII's DR to slow/medium speed medium format color print film. I have no idea why you're bringing up slide film?? It was a comparison between a specific digital camera with a very specific type of film specifically regarding the aspect of DR.
