RailPictures.Net Forums

RailPictures.Net Forums (http://www.railpictures.net/forums/index.php)
-   Railroad Photography Forum (http://www.railpictures.net/forums/forumdisplay.php?f=9)
-   -   My Latest Rejected Photo: Overexposure (http://www.railpictures.net/forums/showthread.php?t=17272)

dnsommer2013 10-30-2014 11:36 AM

My Latest Rejected Photo: Overexposure
 
4 Attachment(s)
Hello everyone,

Still attacking that ole learning curve here...

The link below will take you to my latest rejected photo. It was rejected for being overexposed. This is troubling. To me it looks fine. The histogram shows a shallow mountain in the middle and two spikes to the right, which I take to represent the clouds.

I expected the photo would be rejected primarily just because it's unremarkable. If it's overexposed, then there really must be some greater problem.

1) Pentax K-5 metering is inaccurate or it sucks and I should have bought a Nikon
2) My monitor needs calibrating
3) Color space is wrong RGB vs sRGB (?)
4) I am just plain incompetent (!)

The way I figure, if I can't take a correctly exposed photo of a stationary locomotive generously bathed in mid-afternoon sunlight, I've really got some learning to do! And I guess I'm feeling pretty discouraged right now. It looks fine on my screen!

I started with a DNG RAW image. I added a hint of extra blue to the sky. I added a tiny bit of sharpening with Smart Sharpen. I reduced it to 1200 x 795 pixels. I saved it as a high quality jpeg. That's it.

Here is the link to the photo:

http://www.railpictures.net/viewreje...86&key=3205188

I've also attached the same photo below if anyone wants to look at the photo in question after the link expires, plus three other shots taken along with this shot. I chose the photo I submitted because the others all look too dark to me. If this rejected photo is overexposed, but looks fine to me, while these others look too dark, then I'm guessing the dark ones are actually the ones that are okay! Which means my monitor brightness is incorrect. And/or I have to buy a $200 Spyder Colormunki or something.

Ergh. Meh.

Anyway. Thanks for your feedback / advice!

JimThias 10-30-2014 12:10 PM

You're overanalyzing it. You probably should have shot at 1/160th instead of 1/125th. Or maybe 1/200th. Simple fix, just bring down the exposure slightly in your editing program. Resubmit, continue on with life.

Also, the easiest way to calibrate a monitor is to get a print made, hold that print up next to your monitor and adjust the brightness until it matches. Boom goes the dynamite.

bbrant 10-30-2014 01:07 PM

I agree with what Jim. I shoot with a K-5 and have never had an instance where the camera made the mistake. Sounds like you're going through a learning curve and the best thing to do is listen to the advice and learn from your mistakes. We've all been in the same boat at some point.

JRMDC 10-30-2014 02:02 PM

The fact that you are looking at alternative shots, and choosing on exposure, means that you are missing a step 5), which suggests not so much that 4) is true :) but that you need some more skills. Exposure adjustments can easily be done in software and any of the five shots can be adjusted.

JimThias 10-30-2014 02:03 PM

Quote:

Originally Posted by bbrant (Post 182079)
I agree with what Jim. I shoot with a K-5 and have never had an instance where the camera made the mistake.

Actually, cameras (or lenses) do make mistakes occasionally. My Canon 17-40 is sick right now and seems to be wanting to set its own exposure. I was out shooting the other day and with the camera set at 1/500 and machine gunning, it took three different exposures. :-(

dnsommer2013 10-30-2014 04:29 PM

Thanks for your replies.

Well, with the K-5, you adjust shutter and aperture until the thing-a-ma-bob is centered over the fat tick mark in the middle of the scale beneath the scene in the viewfinder. The train was standing still, so I had no need to rush. Metering was set to multi-segment. The sun was behind me. I can't imagine why I'd use spot or center weighted metering here. Seems very straight forward in this instance.

If the photo is overexposed, why is the histogram mountain so shallow? It's centered, and rounded, but not very tall. Isn't overexposure indicated by the peak or peaks hitting the ceiling?

Anyway. To frame my consternation another way...

My father had his 80th birthday in September. The family celebrated the event with a big reunion way up in Maine. So I took a week off and drove all over Quebec and New England on the way there and back, visiting all sorts of railfan sites I don't really ever get a chance to visit. And so far not one of the photos I took on that trip has made the cut. So yeah I'm wondering how I blew it here. This shot seems like an easy one, exposure wise. Not really a remarkable subject or composition, sure, but a shot that should have been easy to nail correctly. (Which can only mean one thing: I am an absolutely horrible despicable incompetent railroad photographer! The worst! No more than a shameful footnote to the avocation! Maybe there was a time when I coulda been a contender, but let's face it, I'm a bum, that's what I am! Probably the only railfan who ever blew a cakewalk photo op like this one! (LOL!)

(*Sigh*)

Sure wish I could do trips like that more often...

Maybe I also should have put K-5 in "P" mode to see what exposure selections the K-5 thought best. At this point it still probably knows better than me.

miningcamper1 10-30-2014 09:05 PM

Quote:

Originally Posted by JimThias (Post 182078)

Also, the easiest way to calibrate a monitor is to get a print made, hold that print up next to your monitor and adjust the brightness until it matches.

:confused: How can you calibrate a backlighted display using a frontlighted photo? :confused:

I did not trust my new LED monitor until I had viewed hundreds of presumably correctly processed online photos.

Even so, variations are rampant. I opened the original above rejection in 3 different editors, and got 3 very different looks. Is PanAm blue a pure blue, or more purple as it looks in some photos?

JimThias 10-30-2014 11:18 PM

Quote:

Originally Posted by miningcamper1 (Post 182087)
:confused: How can you calibrate a backlighted display using a frontlighted photo? :confused:

You have brightness controls on that backlighted display, right? Process your favorite image and have it printed by a company you trust. If the print looks too dark, your monitor is too bright. If your print looks too bright, your monitor is too dark. Adjust accordingly and VOILA! :D

JimThias 10-30-2014 11:20 PM

Quote:

Originally Posted by dnsommer2013 (Post 182086)
So I took a week off and drove all over Quebec and New England on the way there and back, visiting all sorts of railfan sites I don't really ever get a chance to visit. And so far not one of the photos I took on that trip has made the cut. So yeah I'm wondering how I blew it here.

How did they not make the cut? What were the rejections? You're going to have to elaborate a little more than that.

If they didn't make the cut because of color or exposure, your processing skills probably need to be worked on.

bigbassloyd 10-31-2014 12:47 AM

The print versus display test works well as a double check for me. I make sure my monitor is calibrated to my print company before hand, but I always verify prints in hand when I get a shipment in. I only had an issue with one print out of the last bunch of over 300 or so this month.

Loyd L.

Chris Z 10-31-2014 04:08 AM

Quote:

Originally Posted by JimThias (Post 182081)
Actually, cameras (or lenses) do make mistakes occasionally. My Canon 17-40 is sick right now and seems to be wanting to set its own exposure. I was out shooting the other day and with the camera set at 1/500 and machine gunning, it took three different exposures. :-(

Sounds like your camera is set for bracketing.

Chris Z

dnsommer2013 10-31-2014 12:07 PM

Quote:

Originally Posted by JimThias (Post 182089)
How did they not make the cut? What were the rejections?

If they didn't make the cut because of color or exposure, your processing skills probably need to be worked on.

The reasons have been:

Rule of Thirds
Contrast
Saturation
Sharpness

I would say I *definitely* have a ways to go as far as learning the software, not probably!

As I stated already though, this particular shot seems properly exposed to me, both to my eyes and as indicated by its LR histogram. So I did not expect it to get flagged for an exposure issue. I more expected a rejection along the lines of it being an unremarkable or uninteresting subject/composition. Which is why I am now questioning the accuracy of what I'm seeing on my monitor.

Did the other three photos look rather underexposed to you, as they do to me? This will give me an idea as to whether I do indeed have a calibration issue.

And yes, I know I can make exposure adjustments to these photos in post. I just figured I should start with the image that seems best, straight out of the camera.

Thanks for your reply!

miningcamper1 10-31-2014 12:58 PM

Quote:

Originally Posted by dnsommer2013 (Post 182094)
The reasons have been:

Rule of Thirds
Contrast
Saturation
Sharpness

I would say I *definitely* have a ways to go as far as learning the software, not probably!

As I stated already though, this particular shot seems properly exposed to me, both to my eyes and as indicated by its LR histogram. So I did not expect it to get flagged for an exposure issue. I more expected a rejection along the lines of it being an unremarkable or uninteresting subject/composition. Which is why I am now questioning the accuracy of what I'm seeing on my monitor.

Did the other three photos look rather underexposed to you, as they do to me? This will give me an idea as to whether I do indeed have a calibration issue.

And yes, I know I can make exposure adjustments to these photos in post. I just figured I should start with the image that seems best, straight out of the camera.

Thanks for your reply!

First one: too light
Second: exposure looks good IMHO.
Third: too dark
Fourth: even darker

Why not create a page of accepted RP shots (like what Flickr does) that you can paste your shot onto? Then see if yours is similar or different. This side-by-side-by-side method would eliminate all variables.

JRMDC 10-31-2014 01:45 PM

Quote:

Originally Posted by dnsommer2013 (Post 182086)
If the photo is overexposed, why is the histogram mountain so shallow? It's centered, and rounded, but not very tall. Isn't overexposure indicated by the peak or peaks hitting the ceiling?

You misunderstand the use of the histogram. Overexposure is not high peaks but rather the entire distribution shifted to the right, toward the bright end. A high peak can mean a lot of things. But a high, narrow mountain can mean lack of contrast - contrast means a range tonality, of darks and lights, so a narrow mountain means your tonal range is poor. Increasing contrast results in that mountain being wider and, necessarily then, less tall.

At the same time, don't obsess about the exact shape of the histogram. There is no one ideal shape, not even close. The histogram is useful for understanding what is going on and looking for adjustment ideas.

JimThias 10-31-2014 02:20 PM

Quote:

Originally Posted by Chris Z (Post 182093)
Sounds like your camera is set for bracketing.

Chris Z

Nope. Bracketing would show three different exposures. My camera was set at 1/500th and was showing different exposures. It only does it with the 17-40 lens, and I'm assuming it has to do with the connection it's making to the body. Sometimes when I put the lens on, it gives me an error message when trying to shoot at f8. If I dial it down to f4, take a test shot, and then dial it back up to f8, it works. Very strange. It's actually been doing something odd like this for about 6 years (even after it's been to Canon for servicing). Prior to this current issue, once in a while when taking a test shot, it would capture an image that looked about 10 stops over exposed. (could have been less and that may be an exaggeration). Then next frame would then be fine.

f8, 1/500th, ISO 200:

http://i974.photobucket.com/albums/a...psd969f520.jpg

Same settings:

http://i974.photobucket.com/albums/a...psbe5a6db5.jpg


Histogram for both:

http://i974.photobucket.com/albums/a...ps9b56aecb.jpg

JimThias 10-31-2014 02:24 PM

Quote:

Originally Posted by dnsommer2013 (Post 182094)
The reasons have been:

Rule of Thirds
Contrast
Saturation
Sharpness

I would say I *definitely* have a ways to go as far as learning the software, not probably!

All fixable rejections, including the RoT rejection IF you have enough image to work for cropping.

Processing will come with experience. Just keep working on it.


All times are GMT. The time now is 03:26 PM.

Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.