Why you shouldn't be worried about DR

[dlleno]

* the scene contains detail that is important to the meaning and effectiveness of the photo, that is beyond the DR capability of your sensor

To me, blown highlights (unless intended by the photog) always look bad, not matter if this part would contain "important detail" - but of course that's just perception.

yes, thats what I mean by "important detail" for example, small areas of blown highlights, reflections and such, can be acceptable depending on their size and shape and the way the blend into the photo. If they are not acceptable, then it means there is important detail that you can't capture and the photo fails to meet the tog's objective.

I'd like to add that of course no sensor in the near or mid-term future will allow you to capture sunset with a dr of 18-20 stops, but unfortunately I'm often just missing a little bit, so I'm positive that even a little gain in dr would help a lot.

that a tripod and/or other suitable support is not out of the question!

That's mutually exclusive with your points containing "you must capture the shot with one exposure" - my main problem with hdr outside are moving grass & leaves, and if these are in a high dr area of the shot even manual blending in postprocessing doesn't help.

not necesarily... what I mean here is the just because you must capture with one shot doesn't mean you are not already using a tripod. that is, the conditions that favor the need to capture a high DR scene are often compatible with using a tripod. I dont' mean all the time, of course . yea totally agree with moving leaves especially that this is a real challenge, no matter how good the ghost reduction is. and today's HDR offerings are pretty impressive in that regard.

I also agree that many might never encounter that problem, the real question is what tradeoff a higher dr would produce, i.e. as far as I understand it filling a higher dr range with low dr data produces worse gradients, this tradeoff could only be fixed by a 15/16bit resolution?

 

[Sporgon]

I'd like to add that of course no sensor in the near or mid-term future will allow you to capture sunset with a dr of 18-20 stops, but unfortunately I'm often just missing a little bit, so I'm positive that even a little gain in dr would help a lot.

And if you did have this amount of latitude in a single exposure you would produce a very flat, boring image with little visual punch.

The irony of the DR debate is that the compression of DR is what often makes a photograph worth looking at.

 

[Sporgon]

I'd like to add that of course no sensor in the near or mid-term future will allow you to capture sunset with a dr of 18-20 stops, but unfortunately I'm often just missing a little bit, so I'm positive that even a little gain in dr would help a lot.

And if you did have this amount of latitude in a single exposure you would produce a very flat, boring image with little visual punch.

The irony of the DR debate is that the compression of DR is what often makes a photograph worth looking at.

you do not know what a selective curve is - do you

Yep, and I know what a lunatic is too.

 

[insanitybeard]

you do not know what a selective curve is - do you

You do not know when to stop - do you?

We get it. Nikon is better. How many more times need you tell us?

 

[dlleno]

I'm having second thoughts maybe I should sell popcorn afterall!

 

[Pi]

ankorwatt is ruining this place, you can't have a thoughtful conversation, you can't offer advice, you can't just talk to each other without his incessant DR is the only important imaging metric, D800 this, QPcard that. I am out of here until he gets banned again, it just isn't fun, informative, or helpful.

In a thread called Why you shouldn't be worried about DR, his posts are not out of line.

 

[Apop]

I'm done, have fun.

ankorwatt is ruining this place, you can't have a thoughtful conversation, you can't offer advice, you can't just talk to each other without his incessant DR is the only important imaging metric, D800 this, QPcard that. I am out of here until he gets banned again, it just isn't fun, informative, or helpful.

Banning people because you do not agree with them?
If you dont like his opinion , dont read it ?

As long as there are no serious personal insults I see no reason why people should be banned .
His participation in topics is very selective anyway , I dont see many other comments from him that arent DR related.....

You running away from discussions or one sided arguments from a DR fixated person is rather weak!

 

[Marsu42]

And if you did have this amount of latitude in a single exposure you would produce a very flat, boring image with little visual punch.

Of course any high-dr image before applying a global tone curve, global tonemapping or local adjustments (no matter if done by you or by "intelligent" software) looks extremely boring, that's why you're supposed to use postprocessing ... if you want punchy out-of the box shots use a mobile phone as you immediately get satisfying blacks and blazing whites :->

I'm done, have fun. [...] it just isn't fun, informative, or helpful.

I'm sorry, you're one of the most competent and helpful members around here - but of course it's your free, spare time and it certainly *should* be fun, informative and helpful. I hope you reconsider as ignoring parts of a thread is possible, usually longer-running threads degrades into silliness anyway, ankorwatt or not.

 

[tcmatthews]

Lets face it this post was troll bait to begin with.

 

[Alrik89]

Still fun to see, how much time you waste on this stupid theme.

 

[dlleno]

ankorwatt is ruining this place, you can't have a thoughtful conversation, you can't offer advice, you can't just talk to each other without his incessant DR is the only important imaging metric, D800 this, QPcard that. I am out of here until he gets banned again, it just isn't fun, informative, or helpful.

its true that the common denominator in these derailed threads that degenerate into nothing is, well, recognized by most. The signal to noise is painful in some posts, to be sure. But as for offering adivce, engaging in thoughtful conversation and just plain talking -- these are wonderful objectives, to be sure. and there is an easy solution to the S/N problem: ignore (i.e. don't respond do) the ad hominum. If we did more of that around here we could have some engaging conversations I suspect.

 

[LetTheRightLensIn]

I like noise. Sometimes when I am feeling crazy I even add more!

I doubt you add pattern banding noise though . And I doubt you add so much as to completely erase details that you want to be present .

I sometimes add some film-like grain to B&W white shots and stuff too, but that's something different than not wanting more DR for some things.

 

[Famateur]

you do not know what a selective curve is - do you

You do not know when to stop - do you?

We get it. Nikon is better. How many more times need you tell us?

He won't quit until he gets 14 "stops". ;D

 

[Joe M]

I would love it if my 5D3s had 12 fps, 15 stops of DR, had two blazing CF slots and had all double cross focus points. But if Canon gave me all of that today, what would they sell me in a couple of years? They hand out features in dribs and drabs to keep the customer coming back for more. Do I want more DR? Yes. Can I shoot with what I have now? Yes. Should Canon give us more? Yes. Will Canon give us what they want to give us and not what we want? Yes. Am I worried? No.

 

[Pi]

But if Canon gave me all of that today, what would they sell me in a couple of years?

Well, the flip side is: if they do not give you what you want, what would they sell you today? In my case, nothing. I did not upgrade because I was not excited enough.

 

[qwerty]

you do not know what a selective curve is - do you

You do not know when to stop - do you?

We get it. Nikon is better. How many more times need you tell us?

He won't quit until he gets 14 "stops". ;D

What halfway serious photographer would be content with only 14 stops???

Last time I was shooting for fun, I needed something like 19-20 stops of DR (graffiti in an abandoned basement at EV 1-2; looking out into daylight at EV 13-14). To get the shadows indoors and the highlights outdoors on my 5d III, I had to take 5 exposures at 2 stop increments. (For those of you keeping score at home, DxO reports the DR @ 8MP; the DR at full resolution will be lower by almost a stop. In theory I could have gotten away with taking exposures in an 8 stop range instead of 10 stops.) Obviously some scenes can not be captured in multiple exposures like this.

To get this extra 8 stops, all we need are bigger sensors; doubling the edge length of a sensor (and keeping the per-pixel tech the same) will give you one stop of DR (at the original resolution), with a free resolution boost in the midtones. So, I just need Canon to come out with a sensor of (36*2^8) mm format (about 30 ft), and I will have all the DR I needed for that shot (assuming I was content with a resolution of 20-some MP in the shadows). On the other hand, an Exmor sensor with its built-in 2 stop DR advantage would only need to be (36*2^6)mm (about 7 1/2 ft) -- small enough to mount on your truck.

Looking at it another way, all Canon would need to do to match Nikon's 2 stop DR advantage would be to make sensors with their current tech that are about 100x150mm. (And you would get a huge boost in resolution when you are not limited by read noise.) Or maybe, in the interest of fairness, we could convince Nikon to replace their FX line with a 4x crop sensor so us Canon fanboys won't feel inferior.


You know, sometimes I can't tell where I am on the serious-sarcastic spectrum... time for the weekend.

 

[Loswr]

it requires knowledge

Back when I taught in medical schools, and with some coaching from my wife (who has taught science to preschoolers through adults, and is a professor), I learned that beating people over the head with a fact repeated over and over was not an effective way to instill that knowledge in people who require it.

That seems to be a piece of knowledge that has eluded you.

 

[Jim O]

But if Canon gave me all of that today, what would they sell me in a couple of years?

Well, the flip side is: if they do not give you what you want, what would they sell you today? In my case, nothing. I did not upgrade because I was not excited enough.

This may be a quite logical choice. It's always about costs and benefits and there are always those who will skip a generation or even two. That's a built-in part of the equation and Canon, like any corporation, understands that that maximizing profits is not about selling the most units. Sometimes they might get it right, sometimes they might not. Sometimes managers get promoted, sometimes they get reassigned. They do not care whether you upgrade. They care that they sell enough units to meet their predictions in terms of cost and price. If everyone upgrades they haven't priced it high enough. If no one upgrades, they've priced it so high that there's no value in upgrading, even for the new feature set.

Do you upgrade your car every year? Probably not. I'm driving a 2006 E class Mercedes with 33,000 miles on it. I practically stole it from the dealer three plus years ago when it was four years old and had only 5500 miles on it! One owner who really never drove it. It's got a lot of life in it. When I bring it in for service they give me a loaner and it's usually a current or one year old E class. It has blue tooth, satellite radio, a more sophisticated sound system in general, a more sophisticated navigation system (as does my iPhone 5 with Google Maps for what that's worth), and a nicer sunroof. It also has a rear view camera, something I've lived without all of my driving life, but wouldn't mind having. The ride is pretty similar, especially since I put on four new tires that had dry rotted from age but still passed inspection. Am I going to "upgrade"? No, of course not. Why? It's not enough to convince me. Maybe if they add in jet packs so I can fly over traffic, that might convince me. In the meanwhile, I'll keep driving my "nearly new" almost eight year old car. And you'll keep shooting with your older camera(s) until there is something to which upgrading makes sense.

 

[Jim O]

it requires knowledge

Back when I taught in medical schools, and with some coaching from my wife (who has taught science to preschoolers through adults, and is a professor), I learned that beating people over the head with a fact repeated over and over was not an effective way to instill that knowledge in people who require it.

That seems to be a piece of knowledge that has eluded you.

So well said.

There is also a thing called courtesy and repeatedly insulting people with "you do not know what a selective curve is - do you" is not a good way to achieve a goal, unless the goal is to get them to stop hearing what you have to say.

 

[Famateur]

you do not know what a selective curve is - do you

You do not know when to stop - do you?

We get it. Nikon is better. How many more times need you tell us?

He won't quit until he gets 14 "stops". ;D

What halfway serious photographer would be content with only 14 stops???

Last time I was shooting for fun, I needed something like 19-20 stops of DR (graffiti in an abandoned basement at EV 1-2; looking out into daylight at EV 13-14). To get the shadows indoors and the highlights outdoors on my 5d III, I had to take 5 exposures at 2 stop increments. (For those of you keeping score at home, DxO reports the DR @ 8MP; the DR at full resolution will be lower by almost a stop. In theory I could have gotten away with taking exposures in an 8 stop range instead of 10 stops.) Obviously some scenes can not be captured in multiple exposures like this.

To get this extra 8 stops, all we need are bigger sensors; doubling the edge length of a sensor (and keeping the per-pixel tech the same) will give you one stop of DR (at the original resolution), with a free resolution boost in the midtones. So, I just need Canon to come out with a sensor of (36*2^8) mm format (about 30 ft), and I will have all the DR I needed for that shot (assuming I was content with a resolution of 20-some MP in the shadows). On the other hand, an Exmor sensor with its built-in 2 stop DR advantage would only need to be (36*2^6)mm (about 7 1/2 ft) -- small enough to mount on your truck.

Looking at it another way, all Canon would need to do to match Nikon's 2 stop DR advantage would be to make sensors with their current tech that are about 100x150mm. (And you would get a huge boost in resolution when you are not limited by read noise.) Or maybe, in the interest of fairness, we could convince Nikon to replace their FX line with a 4x crop sensor so us Canon fanboys won't feel inferior.


You know, sometimes I can't tell where I am on the serious-sarcastic spectrum... time for the weekend.

LOL...touché. You've convinced me -- I'm not going to be satisfied until I have at least a truck-mounted 20-stop camera that fits in the abandoned basement to get that shot and avoid bracketing.

These dynamic range flame wars have been entertaining at times. Thanks for keeping it that way. (I mean that genuinely.)