saneproduction wrote:
We always hear about the d800 sensor is amazing for shadow recovery, but can anyone post some images that show the real world use and why the 5dIII is not good enough? Here is a random high contrast shot that I adjusted the raw on (note that I am on my surface pro and have not calibrated the display yet). How much more range do I really need? Would it really improve things?
the question is wrong.
whenever the subject/motive has a wide, large dynamic range you can expose the Nikon D800 after the highlight to get this highlights reproduced without any clipping and with a full color reproduction,if there are also are highlights more than 4 stop above middle grey you can under expose to get this highlights reproduced.
At the same time you are under exposing you are also moving middle grey down towards lower levels/ shadows
and you must adjust the motive with for example curves etc etc.
With Nikon d800 you have at least 2 stops more DR from middle grey and down in to lowest levels compared to Canon, when you then are adjusting the motive this will shows in the the shadows and also in shadows-up against middle grey = noise from Canon and patter noise and also banding.
Here is sun set with one of my Canon 5dmk2 and D800
Both cameras identical exposed and the same handling in camera raw
As per the OP's question, I'd hazard a guess that we won't be seeing DR like the human eye can until way beyond any of us are shooting...anything.
However while being a Canon shooter that doesn't me I'd kick more out my digital bed. It would be fantastic to have even more exposure latitude and I completely disagree with anyone that equates the usage of more DR as somehow a crutch for the lazy photographer.
That's like saying give up your tripods because you're carrying smaller bodies, not 8x10 or 4x5 box camers...what BS. Even with super duper enhanced DR we'll still not be happy, someone will pixel peep at plus 1000 and complain about something.
Exmor wins at the moment, Nikon took advantage of it, Canon lost...BFD...if you must have that DR go get it, make your clients happy, make more profit...why wouldn't you.
Monito, the Lytro was an example of a technology that deferred processing. That's all it's ever been in this discussion: an example. I don't care about the details. All I asked was, if you could adjust DOF and focus after the fact, why wouldn't you? Why is that so hard to answer? You're trying to win an argument I'm not making.
macrobild wrote:
the question is wrong.
whenever the subject/motive has a wide, large dynamic range you can expose the Nikon D800 after the highlight to get this highlights reproduced without any clipping and with a full color reproduction,if there are also are highlights more than 4 stop above middle grey you can under expose to get this highlights reproduced.
At the same time you are under exposing you are also moving middle grey down towards lower levels/ shadows
and you must adjust the motive with for example curves etc etc.
With Nikon d800 you have at least 2 stops more DR from middle grey and down in to lowest levels compared to Canon, when you then are adjusting the motive this will shows in the the shadows and also in shadows-up against middle grey = noise from Canon and patter noise and also banding.
Here is sun set with one of my Canon 5dmk2 and D800
Both cameras identical exposed and the same handling in camera raw
StillFingerz wrote:
As per the OP's question, I'd hazard a guess that we won't be seeing DR like the human eye can until way beyond any of us are shooting...anything.
However while being a Canon shooter that doesn't me I'd kick more out my digital bed. It would be fantastic to have even more exposure latitude and I completely disagree with anyone that equates the usage of more DR as somehow a crutch for the lazy photographer.
That's like saying give up your tripods because you're carrying smaller bodies, not 8x10 or 4x5 box camers...what BS. Even with super duper enhanced DR we'll still not be happy, someone will pixel peep at plus 1000 and complain about something.
Exmor wins at the moment, Nikon took advantage of it, Canon lost...BFD...if you must have that DR go get it, make your clients happy, make more profit...why wouldn't you....Show more →
LARGE DR gives you freedom to exposure a contrasty motive = larger exposure latitude= like negativ film= your own skills=larger DR in the end product
well , it tells little bit what you can do or not, and where noise, pattern noise and banding reduces the potential for image processing in the shadows up against middle grey in a contrasty motive
macrobild wrote:
LARGE DR gives you freedom to exposure a contrasty motive = larger exposure latitude= like negativ film= your own skills=larger DR in the end product
Yeah I now about slide vs negative film, it's exposure latitude, think that was my point, why would anyone not want it, and no having it wouldn't make you lazy, it would indeed give you more freedom to create.
Why did I shoot slide film vs negative...I actually liked the contrast, shadows seem to have lost their allure it seems. Kodachrome 25 when paired with Cibachrome paper gave you killer dense blacks in color, Fiji Velvia 50 does much the same these days but with more vivid colors/colours
alexdi wrote:
Monito, the Lytro was an example of a technology that deferred processing. That's all it's ever been in this discussion: an example. I don't care about the details. All I asked was, if you could adjust DOF and focus after the fact, why wouldn't you? Why is that so hard to answer? You're trying to win an argument I'm not making.
Because the point that there will always be tradeoffs seems to escape you. That is the flaw in the argument you are making. Your question is incomplete because it is not fully thought through. It assumes that non-Lytro technology will stand still.
If the more complete question is "Would you adjust DoF and focus after the fact with 2031 Lytro DSLRs if the alternative were 2013 current DSLRs", then the answer would be Yes. If the more complete question is "In 2031 would you use 2031 Lytro DSLRs or 2031 non-Lytro DSLRs", then the answer (for me) would be the non-Lytro DSLRs.
The tradeoff is that Lytro reduces roughly a 16 MPx sensor to 1.2 MPx output (I'd have to look up the actual numbers). That 10% / 90% tradeoff makes the current discussions about Nikon's "dynamic rage" (as one write put it) and the 22% pixelage difference look like tempests in teapots.
Yes, at some point, people will be able to adjust DOF and focus after the fact, with the same resolution as DSLRs today. But by that time regular DSLR sensors will have advanced too, and skilled photographers will generally prefer those DSLR sensors to Lytro type sensors.
Look at it this way: you can get 1995 DSLR image quality with a Lytro today in 2013 (1.3 MPx). In 2031, the Lytro will give you 2013 DSLR image quality (18-24 MPx). But in 18 years DSLR sensors are not going to stand still.
StillFingerz wrote:
Yeah I now about slide vs negative film, it's exposure latitude, think that was my point, why would anyone not want it, and no having it wouldn't make you lazy, it would indeed give you more freedom to create.
Why did I shoot slide film vs negative...I actually liked the contrast, shadows seem to have lost their allure it seems. Kodachrome 25 when paired with Cibachrome paper gave you killer dense blacks in color, Fiji Velvia 50 does much the same these days but with more vivid colors/colours
that was the time, I have some 100000 slide films and also Kodacrome 25 where my Imacon scanner have great difficulty to scan through
This Lytro thing is really distracting you. Let's talk about white balance instead. Until about fifteen years ago, we couldn't adjust it in post. There's presently no quality trade-off in deferring the decision. So why not?
alexdi wrote:
This Lytro thing is really distracting you. Let's talk about white balance instead. Until about fifteen years ago, we couldn't adjust it in post. There's presently no quality trade-off in deferring the decision. So why not?
I do defer it, when possible, which is most of the time. I shoot Raw.
There are also a number of scenarios where it is not possible to ignore white balance at the time of making the shot. In those cases I would not or do not defer it.
There are always tradeoffs. To assume that there would be no tradeoffs for some advanced 18 year hence hypothetical super-duper DR sensor is to make the hypothetical worth less than the barely above zero value it has already.
The "shoot now, think later" hypothetical assumes no parallel advances in camera systems that would offer working photographers more advanced choices in the field than available now. Those systems would tend to be chosen by many or most working photographers over the "shoot now, think later" systems even if "think later" would benefit some other working photographers over "think now" for their line of work.
Exactly. You do defer it, most of the time. I defer it all of the time, because it distracts from capturing a fleeting moment. But I'm an event photographer, that's what I care about. Maybe I'd spend more time on setup if my subjects didn't change. I disagree with your assumption about tradeoffs (the only tradeoffs in your raw converter involve lens corrections), but I'm tiring of this. Let's not continue.
macrobild wrote:
that was the time, I have some 100000 slide films and also Kodacrome 25 where my Imacon scanner have great difficulty to scan through
I'm at about that same count in slides, but with those I actually print above 8x10, I don't print from a scanned image, I've a local print shop that is still in business; it opened in the 60's, one of only a few still going strong. There is a development/printing/scanning pro shop down San Diego way that is fantastic. When I've a batch of film to process I send it there, they even will do individual high-res drum scanning and it's quite good. Here's the link if you're interested...NCPS is their name, KR turned me on to this shop, they are fantastic, product and service wise...
Photography works because the human brain is easily tricked with optical illusions, such as a 2D pattern of contrast is a real 3D object when the pattern matched what is expected / normally seen in the context of the photograph.
When looking a photo our brains are conditioned to perceive anything lighter than the maximum black as having detail, because in a photo tone above max. black usually has detail. The fact the brain is fooled that way comes in handy when shooting a backlit scene which exceeds sensor DR. In post-processing the tonal values between around 20 if amplified and lighten will made it seem like the camera record detail everywhere the eye would normally see it in person. But that trick only works when the shadows are small and unimportant and the noise resulting from the amplification isn't noticed.
With respect to tonal range in a photograph what "looks normal" is subjective decision based on the context seen in the background and the lighting angle clues. Sometimes a full range is perceived by eye (mid-day) but other times in isn't (at light under a street light). For me whether or not the scene fits the camera range is part of the overall creative process. That's because I started out doing B&W photography with the Zone System, then soon afterwards was shooting weddings on color negative film with dual flash.
With the B&W zone system negative development was matched to scene contrast so the negative Zone 1 shadow detail - Zone 9 solid white highlight density range was about 3.0, the ten-stop range of #2 print paper. The other approach typically used for roll film was to develop for "normal" outdoor contrast (a 10-stop scene range) then if the frame had lower contrast (overcast) or higher contrast (cross-lit snowy scene) change the grade of the print paper. I did both at various times and wound up using a color enlarger and densitometer enlarging meter for my B&W so I could tailor the Polycontrast filtration perfectly for any range of density on the negative quickly and predictably to always get a full range of detail with rich blacks and paper white specular highlights.
Solid white objects on prints (Zone 9) are actually light gray, but that is necessary to create the illusion of 3D on white objects via placement brighter specular highlights on the "off white" objects in the print. The digital equivalent is putting white paper or other white solids at around 245-250, about 1/3 stop below clipping. That's how I expose all my digital shots nowadays. I adjust exposure until I see solid white content clipping then cut exposure by 1/3 stop. Then a quick look at the playback and LEFT side if the histogram tell me how well the scene fit the sensor. If it doesn't experience tells me how much I can pull the shadows-midtones of the RAW file up before any noise is noticed.
Right after dropping out of college where I did the ZS I went to work assisting Monte Zucker, a well-known wedding shooter and teacher. Monte had been an innovator in the use of dual flash for wedding candids. He had started in 40s with B&W and a Speed Graphic. Like most when switching to color negative film in the 60s he discovered a color print had about half the DR of a B&W print and it couldn't be manipulated. So as with digital when shooting a bride and groom with a single flash on camera one could expose for the dress, the suit, or split the difference but there wouldn't be detail everywhere as seen by eye. About the same time Graflex marketed a portable flash and Wien developed a photo slave trigger. Zucker took those new tools and used them to use what was then a common studio lighting practice to get full range images: Even fill from the direction of the camera on a camera bracket set for shadow detail and an overlapping off camera key light on a stand to match the highlight densities on the negative to the 5-6 stop range of the print. The added benefit is that moving the "key" flash off axis produced more realistic natural 3D modeling vs. the flat 2D lighting which had been the norm for indoor flash use.
What Zucker did with flash on color prints was the same as what Adams did with B&W film development: match negative range to print range. Zucker did it by changing the scene range in the foreground with flash. We used identical direct flashes. Putting the OCF light at 6ft and shooting from 8ft resulted in a 3:1 reflected ratio. Moving out to 11ft for a wide shot the OCF was moved to 8ft to maintain the 3:1 ratio and the lens opened a stop. Moving in from 8ft to 6ft for a close up the OCF was moved to 4ft to maintain the ratio and the lens closed a stop. Those 3 scenarios covered about 95% of all the flash shots at a wedding reception producing full range photos with natural looking 3D modeling.
That was back in 1972 and since then I've only rarely taken any photo indoors or out without a flash over the camera with a bracket and another for use as needed. I incorporate the use of flash and Photoshop manipulation into any shot I take. This is just a vacation shot, but it illustrates the workflow:
The ambient exceeded the camera DR. My strategy in those situations is to put a focal point in the foreground within the range of flash as a distraction so the loss of detail in the scene behind isn't really noticed that much. As in any shot with foreground showing the flash overexposed it so I had to darken the foreground in PP to make it look more normal. I also darkened distractions like the people in the background and tried to pull more detail out of the trees. Not shown are the "global" corrections I did in ACR before opening the shot in Photoshop to use the adjustment layers for local corrections.
I'd love to have a digital camera that could handle a backlit scene without flash like B&W with Zone System control could, but I'm not holding my breath waiting for one and I still keep a powerful flash on the camera at all times
Flash has its place, but generally fill flash wont look like the original scene lighting. Ask me which I'd prefer, and this is the order:
* push the subject in post
* expose for the subject and let the rest go
* add fill flash
There are cases when fill is the right thing to do and I do use it, but it doesn't produce the same result.
PhilDrinkwater wrote:
Flash has its place, but generally fill flash wont look like the original scene lighting. Ask me which I'd prefer, and this is the order:
* push the subject in post
* expose for the subject and let the rest go
* add fill flash
There are cases when fill is the right thing to do and I do use it, but it doesn't produce the same result.
With landscapes, fill flash is usually useless unless the shadow area is very close to the camera. In general with landscapes it is usually GND filters or multiple exposure blending.
I love waterfall photography.....and I'd love the extra DR to lift shadows. Water is "hot" even in low contrast situations like a cloudy day. I can lift the shadows fairly acceptably on the 5D II given the right conditions, but I'd love the flexibility of another stop or two. I don't do HDR....
PhilDrinkwater wrote:
Flash has its place, but generally fill flash wont look like the original scene lighting. Ask me which I'd prefer, and this is the order:
* push the subject in post
* expose for the subject and let the rest go
* add fill flash
There are cases when fill is the right thing to do and I do use it, but it doesn't produce the same result.
-1
I completely disagree.
Well executed fill-flash is almost or completely un-noticeable and grossly manipulating a file is the very last thing I want to do.
Get it right in-camera if you can is always the best policy IMO.
