Photographing a White Bird

Just extracted the jpeg from the first white duck image

Sorry my friend, but not all cameras are suited to shooting white birds.

Don't know what that means. Are you saying my camera or Sandy's camera is not suited to shooting white birds?

 

Anyway I extracted the jpeg from the first default Raw white duck half in shade and edited it to see if by exposing for the highlights which would render a rather dark image if there's enough good image data to work with to brighten it and I'm surprised to say there is. Edited in ACR 6.7.

 

so the max peak white highlight didn't extend into the fourth zone section on the right of the incamera histogram

I generally prefer to move the histogram as far right as possible - without blowing highlights of course (which takes some guesstimates as no camera provides a RAW-based histogram). Lifting shadows brings up the noise; this is not an issue when pulling an image down. When the image at the back LCD looks a bit washed out with a reasonable amount of blinkies only, then I know I got the optimum image for later post-processing.

Just extracted the jpeg from the first white duck image

 

Don't know what that means. Are you saying my camera or Sandy's camera is not suited to shooting white birds?

 

 

Well, yes, for your camera, given the circumstances that you cited, with a bird going in out of direct sunlight. To properly expose in shade and in the open, you need a quick an easy way to adjust EV without taking your eye away from the VF. I don't know about Sandy's camera.

I generally prefer to move the histogram as far right as possible - without blowing highlights of course (which takes some guesstimates as no camera provides a RAW-based histogram). Lifting shadows brings up the noise; this is not an issue when pulling an image down. When the image at the back LCD looks a bit washed out with a reasonable amount of blinkies only, then I know I got the optimum image for later post-processing.

 

I look for a "few" blinkies, but no whole areas of the bird blinking. I know that I can recover way more than the in-camera, preview JPEG would indicate.

...you need a quick an easy way to adjust EV without taking your eye away from the VF. I don't know about Sandy's camera.

What do you think the camera's internals are going to do when you adjust EV up or down with one rotodial? It's going to pick for you shutter, aperture or ISO depending on the auto exposure setting of either Program, Av, Tv or ISO range set by the user to let more or less light in or amplify the sensor's electronics with high ISO.

 

Shooting in bright noon-ish daylight at f10, ISO 200, as I did with the white ducks, my K100D chooses outrageous and unnecessary fast shutter speeds over 1/1000's meaning too much light even under partial shade. I just don't trust leaving it to my camera's metering with everything set to auto.

 

Now in low light like Sandy's morning scenes ISO 800 is just way too noisy for exposures that preserve highlights with my camera. Sandy's first image is at ISO 2500, 1/4000's, f/14. Why would Nikon's internals choose that odd combination of exposure settings for that amount of light?

OK, I'm totally perplexed by camera and sensor technology.

 

Dieter's mentioning lifting shadows from highlight preserved exposures shows too much noise. It got me wondering about the extracted jpeg from Raw edit of the white duck above and whether that dark murky shadow area just underneath the duck produced a bunch of noise after I brightened it. Surprisingly it didn't and I even have noise reduction turned off for jpeg rendering in my camera's menu system. It's as if they seem to be moving the goal post on what really affects image quality with regard to noise. See the 100% view crop...

 

And the EV comp rotodial is a two finger affair requiring I press down on one button and role the dial with the other.

That's how it works on Nikons too - press the button and rotate a dial. Either press and hold the button or press and release the button, then rotate the dial (behavior can be customized in the menu). Also the option to not have to use the button at all and do it with the dial alone - takes a bit getting used to and after some mishaps I went away from that option. The Sony A7 series has a direct dial for exposure compensation and does not require a button press.

 

Sandy's first image is at ISO 2500, 1/4000's, f/14. Why would Nikon's internals choose that odd combination of exposure settings for that amount of light?

Sandy apparently had the camera in P(rogram) mode - something I never use on any of my cameras. I have trouble understanding how the camera could choose 1/4000s, f/14 and ISO 2500 at the same time. Certainly not my choice of parameters given the circumstances. I'd been at around 1/1600s, f/8, and ISO 400.

 

Dieter's mentioning lifting shadows from highlight preserved exposures shows too much noise.

Not what I said; never said too much. But if you compare a shot that has the highlights all the way to the edge of the histogram with one that has the highlights at the boundary to the fourth section then and you had the scenario where either of them would require lifting the shadows, then the first mentioned would have less noise than the second on account that the second needs one stop more of lifting. Similarly, if the first allowed you to actually pull the shadows whereas the second would still require lifting them, the first would again win in the noise department (albeit, we would be talking about less than a stop in that particular scenario). Depending on where you are on the ISO scale, the differences may be small and almost invisible or rather significant when towards the higher end of the ISO range.

 

Given how well your images turned out given the variation in lighting condition, I'd say you did very well and there's no reason to think that the camera isn't up to the task.

Lifting shadows brings up the noise;

That's the main point I was addressing in the 100% view. How do you explain the lack of noise even I didn't expect since I turned off noise reduction for jpeg rendering?

 

The goal post is the ISO and how much noise it produces between different cameras which seems to keep moving according to one's understanding of what's going on. My K100D is notorious for noise even at base ISO 200 and yet I'm scratching my head on the above crop. Maybe the noon day sun even in shade I shot those white ducks under produces plenty of light that can withstand underexposing to preserve white highlight detail.

Edited November 29, 2017 by Tim_Lookingbill

What do you think the camera's internals are going to do when you adjust EV up or down with one rotodial? It's going to pick for you shutter, aperture or ISO depending on the auto exposure setting of either Program, Av, Tv or ISO range set by the user to let more or less light in or amplify the sensor's electronics with high ISO.

 

Shooting in bright noon-ish daylight at f10, ISO 200, as I did with the white ducks, my K100D chooses outrageous and unnecessary fast shutter speeds over 1/1000's meaning too much light even under partial shade. I just don't trust leaving it to my camera's metering with everything set to auto.

 

Now in low light like Sandy's morning scenes ISO 800 is just way too noisy for exposures that preserve highlights with my camera. Sandy's first image is at ISO 2500, 1/4000's, f/14. Why would Nikon's internals choose that odd combination of exposure settings for that amount of light?

 

In Av mode, I know that the camera will change the SS to adjust for my changes in EV. I've chosen my aperture and ISO and the camera changes SS, in this case. I see the SS in my VF, so, if it starts to drop too low, I either open my aperture more or raise ISO. I can do those fast on my Canons, but not as fast as turning the EV wheel.

 

I'm not sure about your camera, but I'd be shooting at higher than ISO 200 for ducks, even dappling ducks. I'd suggest ISO 400 in bright sun for dappling ducks.

 

I have no idea why Sandy's camera would make such poor choices.

OK, I'm totally perplexed by camera and sensor technology.

 

Dieter's mentioning lifting shadows from highlight preserved exposures shows too much noise. It got me wondering about the extracted jpeg from Raw edit of the white duck above and whether that dark murky shadow area just underneath the duck produced a bunch of noise after I brightened it. Surprisingly it didn't and I even have noise reduction turned off for jpeg rendering in my camera's menu system. It's as if they seem to be moving the goal post on what really affects image quality with regard to noise. See the 100% view crop...

 

The noise that I'm seeing at 100% is not that unpleasant. Also, I don't find luminescence noise bothersome and will leave it in my super-high ISO shots, taken at ISO 20000 and 25600.

 

It was so dark when I took the following shot that I could hardly see the buck in the viewfinder. The BG bokeh is loaded with luminescence that only reduced slightly, in order to preserve as much fur detail as I could. It wouldn't be great as a large print, but it's fine full-screen on the internet:

 

Big Buck After Sundown by David Stephens, on Flickr

Edited November 29, 2017 by dcstep

How do you explain the lack of noise even I didn't expect since I turned off noise reduction for jpeg rendering?

What David said. In addition, you didn't do that much lifting and were at a relatively low ISO, so I wouldn't expect a significant increase.

Clearly, David, with that Canon 5D Mark IV you brought the right equipment that can render that dark of a scene that well at ISO 25600. Checked out the enlarged version on Flickr and can only wish that some day I might be able to afford a camera like that. Edited November 29, 2017 by Tim_Lookingbill

Clearly, Dieter, with that Canon 5D Mark IV

That would be David, not me :)

 

I use Nikons and am fairly certain that I have never used ISO 25600...

What David said. In addition, you didn't do that much lifting and were at a relatively low ISO, so I wouldn't expect a significant increase.

You don't think lifting the jpeg's darkest shadow detail from 000RGB to 20,26,14RGB is a bit extreme especially for a jpeg?

 

If so, then Sandy should consider exposing for the white highlights of BIF's shooting jpeg with the sun low in the sky but with a reasonable high ISO setting that at least allows 1/500's and up shutter speed. Fast shutter speed would be top priority over aperture and high ISO as long as the white bird plumage isn't blown.

That would be David, not me :)

 

I use Nikons and am fairly certain that I have never used ISO 25600...

Whoops! I caught myself the first time I posted to both of you to make sure I didn't get the names mixed up but I guess it was just a matter of time when my memory would bite me in the arse posting this much in one thread. At least I had enough time to correct it.

Clearly, David, with that Canon 5D Mark IV you brought the right equipment that can render that dark of a scene that well at ISO 25600. Checked out the enlarged version on Flickr and can only wish that some day I might be able to afford a camera like that.

 

I feel your pain. ;-( Really, I want wildlife photography to become way more affordable. I think that the more of us in the pool, the merrier.

 

Sony, IME, is moving quicker than Canon, toward an affordable, high-ISO machine. I'm shooting my, relatively expensive, Sony a9 at formerly insane ISOs. The a7R III will be out later this week, with even better dynamic range at a retail price $1,000+ less. We can only hope that this will continue to trickle down their line and break the $1000-body price barrier.

 

Here's a buck, running at 20-mph to 30-mph, shot through my windshield, in pretty heavy snow, with my EXPENSIVE a9 at ISO 25600:

 

Big Buck Runs by David Stephens, on Flickr

Let me just say that, IME, even with modest cameras, like my OLD Canon 7D, shooting in RAW will usually give you around 1-stop more dynamic range than the same in-camera JPEG. Shooting RAW and Exposing-To-The-Right ("ETTR") maximizes your data collection if gives you much greater latitude to adjust during RAW conversion. Birds and wildlife shots often have very wide DR, when you shoot in-camera JPEG, it's as if you're shooting Kodachrome all over again, and eschewing all (well, most) of the advantages of digital files.

If I had metered spot and used the DF, I believe I'd have got the shot. Have been working with Pattern for a few weeks, brought the D 750 for the extra MP, since I knew I couldn't get close. I have always found the D 750 quirkier than the DF, but it was just in on the recall, so I hoped for the best. Actually, except under very difficult conditions, my percentage of successful shots is usually quite decent, which is why I started this thread. Certainly I got a lot of ideas from those of you who contributed. Thanks again!

Sony, IME, is moving quicker than Canon, toward an affordable, high-ISO machine. I'm shooting my, relatively expensive, Sony a9 at formerly insane ISOs. The a7R III will be out later this week, with even better dynamic range at a retail price $1,000+ less. We can only hope that this will continue to trickle down their line and break the $1000-body price barrier.

 

Here's a buck, running at 20-mph to 30-mph, shot through my windshield, in pretty heavy snow, with my EXPENSIVE a9 at ISO 25600:

Not as good as the night shot of the deer with Canon 5D at the same ISO, but then I have to wonder why a manual exposure shutter speed of 1/320's with the Sony. Was this jpeg or Raw? Was the image dark and you lifted it in post?

 

It's very informative to see a number of good cameras perform differently in challenging real world situations over looking at dpreview static scene analysis shots and gallery snaps with no background on the conditions and amount of time the photographer reacted in getting a quick shot.

 

I have to thank everyone as well for posting very useful information and also especially thank Sandy for starting this thread.

Not as good as the night shot of the deer with Canon 5D at the same ISO, but then I have to wonder why a manual exposure shutter speed of 1/320's with the Sony. Was this jpeg or Raw? Was the image dark and you lifted it in post?

 

It's very informative to see a number of good cameras perform differently in challenging real world situations over looking at dpreview static scene analysis shots and gallery snaps with no background on the conditions and amount of time the photographer reacted in getting a quick shot.

 

I have to thank everyone as well for posting very useful information and also especially thank Sandy for starting this thread.

 

Good questions Tim.

 

The night shot of the deer (Canon) was in much better conditions, in that I was standing close to the buck and he wasn't running. The running buck shot was through the windshield and it pretty heavy snow and the buck was moving at least 20-mph. I happened to have my 100-400mm f/4.5-5.6 zoom mounted, with the 1.4x teleconverter, so that results in f/8 as my largest available aperture. If I'd had my 500/f/4 mounted, I would have opened it all the way up and moved SS up to maybe 1/500-sec. It was RAW and I brought the overall EV up in RAW conversion. ISO 25600 seems to be about as high as I can go and still see in the EVF. On the Canon it's a little easier to see in such darkness, but it easier. I prefer to top the Sony out at ISO 20000 and the Canon at 25600, expecting a lot of noise in both cases.

 

One other thing about noise, I use DxO PRIME noise reduction, which does a pixel-by-pixel analysis to minimize noise and not smash details. It does a darn good job. You should see the chrominescance before NR, with either Canon or Sony. The 5D MkIV cleans up way better than the 7D MkII and 5DS-R. The Sony a9 is close to the 5D4. Can't wait to get the a7R III, which is reported to have another stop of DR.

 

Those static IQ tests are useful, but really don't tell the whole story. Some files start out bad and just can't be fixed, no matter what software you use. Some other files, look like crude when you start, but clean up nicely. Real world shots, in tough positions are really the most informative, but finding some that match your subjects and shooting style can be tough. That's why I like to share on PN.

You don't think lifting the jpeg's darkest shadow detail from 000RGB to 20,26,14RGB is a bit extreme especially for a jpeg?

The results seem to speak for themselves - looks OK to me so the change was not extreme.

 

Have been working with Pattern for a few weeks

It takes getting used to; it's a tad hard to predict what it will come up with but eventually one gets the hang of it. I am sure one can come up with scenarios where it completely screws up but in many cases it does a fairly good job and in many others, the corrections needed become obvious at some point.

If I had metered spot and used the DF, I believe I'd have got the shot. Have been working with Pattern for a few weeks, brought the D 750 for the extra MP, since I knew I couldn't get close. I have always found the D 750 quirkier than the DF, but it was just in on the recall, so I hoped for the best. Actually, except under very difficult conditions, my percentage of successful shots is usually quite decent, which is why I started this thread. Certainly I got a lot of ideas from those of you who contributed. Thanks again!

 

Using spot metering on a white bird will result in a grey bird. Also, remember that keeping the spot on a flying bird is quite difficult. If you use spot and meter on a white bird, then you probably need +1EV or +2/3EV adjustment, to get the bird white and still not blow out highlight.

 

BTW, this is a VERY COMMON error. I see postings all the time, showing what should be a white bird and it's grey. Using Spot metering demands that you understand what the meter is trying to do. The meter assumes that what you've selected to meter is actually 18% grey, so it'll try to make that selection look 18% grey, underexposing white things and overexposing black things (think of a black labrador).

Using spot metering on a white bird will result in a grey bird. Also, remember that keeping the spot on a flying bird is quite difficult.

A: It's all in How you use the spot to get what you want. B: No harder than wingshooting - at one point, in saner days, there were camera gunstocks. I have a spare stock, and have considered building one.

A: It's all in How you use the spot to get what you want. B: No harder than wingshooting - at one point, in saner days, there were camera gunstocks. I have a spare stock, and have considered building one.

 

I've never seen a survey about photographers' understanding of how metering works, but my casual observation would have me betting that well over half don't understand it, including those that use spot metering.

 

I don't know anyone personally that uses a stock, but I've seen reviews of one or two that are out there. I hold my camera with 500/f4 attached, much like I did an M-16 in the army, EXCEPT there's no stock against my shoulder. My right hand a arm take a good bit of load, that I've now trained. My face is against the camera and my hand. Starting with a stock is a good idea.

 

That said, putting a spot on a bird and shooting and putting a spot on a bird and following are two different things. With a heron or egret, it's no big deal, but most diver ducks and puffin are different matters, because of small size and velocity in the 50-mph range. I'd rather set my exposure Manually and not have to worry about my spot meter being on exactly the right part of the bird.