Thought experiment; help needed

Jock Elliott

Hall of Famer
Location
Troy, NY
I like to take pictures of the daytime sky (clouds, etc) and the sky at the margins of the day (sunrises, sunsets).

Suppose we have two different cameras. Camera A has dynamic range of 10.8 evs and color depth of 19.1 bits. Camera B has dynamic range of 12.8 evs and color depth of 24.3 bits.

If i were to shoot the same scenes (a sky full of clouds and also a sunset) at the same time with the two different cameras, how would the differences in dynamic range and color depth be reflected in the images? Further, would those differences be noticeable without pixel peeping?

If anyone has photos that would show those differences, please post them.

Thanks in advance.

Cheers, Jock
 
Dynamic range is about light/dark. The larger the dynamic range is, the larger the difference between the brightest value that's just short of clipping, and the darkest value that's just short of blocking. In your sunset shot, if you want the "foreground" (any object that's not part of the sky, really) to be a black silhouet, and there're no really dark clouds in the sunsetty sky, you won't need a particularly wide dynamic range. If you do want to show detail in the foreground, or if there're both dark clouds and brightly lit up clouds and/or the sun, you might need good dynamic range to be able to show detail in both bright and dark parts of the scene.

The colour depth is how many gradations of saturation in a certain colour can be recorded. Each bit is a doubling of the number of saturation levels that can be recorded, just like each EV is a doubling of exposure. If the colour depth is limited, you'll get banding, because even if there is variation in saturation levels in the actual scene, the camera isn't seeing "finely" enough to see that difference, and will record two (or more) different saturation levels as one.

Neither dynamic range nor color depth are particularly pixel-peep-y; if significant parts of the scene exceed the dynamic range or colour depth of the camera, you'll see it at any print size. [edit] smooth colour transitions do become more important at large viewing sizes simply because there're more pixels available to display the transition, so a low number of recorded saturation levels shows up a bit earlier.

Note that most output media, including computer screens and print, can't show 12ev of dynamic range, probably not even 10. Same applies to colour depth. So even if the camera is capable of recording a certain dynamic range or colour depth range, the raw conversion or jpeg engine will still have to be smart enough to compress the recorded data into a range that the output medium can handle, preferably without the whole thing becoming bland and lifeless.
 
Bartjeej,

Let me "play this back to you" and see if I have got it right.

With more dynamic range, I might be able to see more detail in the dark area in the shot below:

G12_clouds_Oakwood_cemetery_020-001_copy_Medium_.jpg
Join to see EXIF info for this image (if available)


And with this shot,with greater color depth I might see more variation in color, except that most computer screens won't display it:

G12_Dawn_s_early_light_029_DxO-1_Medium_.jpg
Join to see EXIF info for this image (if available)


Is that approximately correct?

Cheers, Jock
 
Yeah, that's basically right. In the case of clouds, you can even have thunderheads that are very bright where reflecting the sun and dark as night on the opposite side where all hell is breaking loose - so a lot of DR can help even without "going to ground" with the lower part of the shot. Much of the data available from a higher dyanamic range sensor raw file is not seen in an initially processed shot, but the point is it's THERE and can be pulled into play and manipulated in ways to make the shot look the way you want it to. If you pull up all of the shadows and pull down all of the highlights, you can show almost all of it, but it will get to a point of looking un-naturally flat. So you're always gonna compromise between showing as much dynamic range as you want and still maintaining some drama in the shot. But the bottom line is the more data you start with, the more options you have for how to do that and the more likely you'll find a pleasing combination of settings that gets you where you want to be.

I never fully GOT how much of a difference a file with loads of dynamic range could make until I made the fatal mistake of shooting with an RX1. Those files blew my mind and continue to every time I use them. That experience was my downfall - now I have three cameras, two of which (the RX1 and Df) are full frame, and the third is an APS based camera (Coolpix A) with still quite a bit of dynamic range. I got so spoiled by the RX1 files, that some of the m43 and APS and smaller compact files I'd been perfectly happy with before no longer satisfied me. Part of this is because I really like to dig in when I'm processing - I'm not a "realist" or an accuracy freak - I'm usually going for an effect in my processing. So not everyone would find the high DR files possible from some of today's best sensors as advantageous as I do. But I simply don't want to go back now that I've tasted (and had full gluttonous meals from) today's amazing full frame sensors.

It's all in the pixels, but it's not about creating final images that are only different to pixel peepers. Its about having gobs of data to use the way you want to in creating shots as dramatic or as subtle as you want to in processing. And those differences are usually pretty obvious at 1024x768 or 1600xwhatever...

-Ray
 
Ray,

Thanks!

My goal is to capture the drama that I see in the skies and somehow transmit/translate that to the person who sees my image.

Sometimes I get images that I really like, and sometimes II get frustrated because the image doesn't come close to having the impact that the original scene did. Then I wonder: is it the camera? the post-processing? It couldn't possibly be the "nut behind the shutter release" could it? :D

Cheers, Jock
 
Ray,

Thanks!

My goal is to capture the drama that I see in the skies and somehow transmit/translate that to the person who sees my image.

Sometimes I get images that I really like, and sometimes II get frustrated because the image doesn't come close to having the impact that the original scene did. Then I wonder: is it the camera? the post-processing? It couldn't possibly be the "nut behind the shutter release" could it? :D

Cheers, Jock

As always, it's a combination. Without "seeing" the light and the shadows and the colors (if a color shot) and ultimately the composition and getting the camera settings right to record it, none of the rest of it matters. But if you get that part right, the other stuff can make a difference. Here are a couple of RX1 shots I took on the same evening last month. I got lucky with the content of the shot, but the camera and files it gave me to work with made a huge difference. They're not even the toughest shots from a DR standpoint, but having all of that information in the files STILL made a big difference when it came to holding up under my heavy handed processing. Some of it is the resolution (which the RX1, A7, and D610 have and the D810 and A7R have in spades) but a LOT of it is the overwhelming amount of information in the raw file that I was able to bring out to do what I wanted it to do. My Df doesn't have the resolution or the DR of these sensors, but it has overwhelmingly good high ISO capabilities that made it a good call for a DSLR for me, rather than the D610, which would have doubled down on the strengths of the RX1. In a perfect world, I might have both, but I'd never carry two DSLRs (which I sometiemes did with smaller mirrorless cameras), so it would just be a gross indulgence.

I couldn't have gotten these shots to look the way they do with the APS or m43 or smaller sensor cameras I've used. I might have gotten close with some APS sensors, but there's data in these files that just isn't there in most sensors. It doesn't always matter, but sometimes it does. And, of course, many people wouldn't WANT to do the same things to these files that I did - that's just a matter of taste (or lack thereof). But I DID want these results, so it matters to me... That's my story and I'm sticking to it!

14551712857_d9e59c95dd_h.jpg
Last Night Sunset-15-Edit by ramboorider1, on Flickr

14551710627_5c1d15627f_h.jpg
Last Night Sunset-131-Edit by ramboorider1, on Flickr

-Ray
 
As always, it's a combination. Without "seeing" the light and the shadows and the colors (if a color shot) and ultimately the composition and getting the camera settings right to record it, none of the rest of it matters. But if you get that part right, the other stuff can make a difference. Here are a couple of RX1 shots I took on the same evening last month. I got lucky with the content of the shot, but the camera and files it gave me to work with made a huge difference. They're not even the toughest shots from a DR standpoint, but having all of that information in the files STILL made a big difference when it came to holding up under my heavy handed processing. Some of it is the resolution (which the RX1, A7, and D610 have and the D810 and A7R have in spades) but a LOT of it is the overwhelming amount of information in the raw file that I was able to bring out to do what I wanted it to do. My Df doesn't have the resolution or the DR of these sensors, but it has overwhelmingly good high ISO capabilities that made it a good call for a DSLR for me, rather than the D610, which would have doubled down on the strengths of the RX1. In a perfect world, I might have both, but I'd never carry two DSLRs (which I sometiemes did with smaller mirrorless cameras), so it would just be a gross indulgence.

I couldn't have gotten these shots to look the way they do with the APS or m43 or smaller sensor cameras I've used. I might have gotten close with some APS sensors, but there's data in these files that just isn't there in most sensors. It doesn't always matter, but sometimes it does. And, of course, many people wouldn't WANT to do the same things to these files that I did - that's just a matter of taste (or lack thereof). But I DID want these results, so it matters to me... That's my story and I'm sticking to it!

14551712857_d9e59c95dd_h.jpg
Last Night Sunset-15-Edit by ramboorider1, on Flickr

14551710627_5c1d15627f_h.jpg
Last Night Sunset-131-Edit by ramboorider1, on Flickr

-Ray

Wow!

Cheers, Jock
 

Similar threads

Back
Top