Mount Saint Helens

[Tom Rankin]

This is my favorite photo of our recent trip out west.

Picture was 'shopped' to sharpen, darken, and crop.

Picture is about 700K - dimensions - 2400 x 1870 (works for my screen ratio).

Comments welcome.

 

[MichaelJ]

Amazing to see it carving river canyons in its own mud from the eruption 31 years ago.

 

[Chip]

We had lousy views/clouds/rain when we were there and the destruction's path was still very obvious.

Tom, love the pic. I'd like to see 1 or 2 steps back to show a bit more sky, mountain and run off ?

 

[Tom Rankin]

We had lousy views/clouds/rain when we were there and the destruction's path was still very obvious.

Tom, love the pic. I'd like to see 1 or 2 steps back to show a bit more sky, mountain and run off ?

Ok, I think I tend to overcrop. But I really like the jagged edges of the rim and I wanted to bring that out.

 

[Billy]

My first reaction is that the power of nature is truly astounding. I think I read on a trivial pursuit card (and what higher authority is there?) that the mountain is 1300' lower than it used to be.

About the photo itself....everyone has their own style, but I like to expose for the highlights, at the expense of losing some of the darker detail. The snow in this photo seems to be over-exposed. But that's just my picky taste.

Overall nice shot, and thanks for tempting me to go there.

 

[DougPaul]

About the photo itself....everyone has their own style, but I like to expose for the highlights, at the expense of losing some of the darker detail. The snow in this photo seems to be over-exposed. But that's just my picky taste.

A histogram of the web version of the pic also shows some saturation both the bright and dark ends. However, this is the processed version--it does not say that the original has saturation (particularly if it was shot RAW).

I'm not judging the saturation--this image has a large dynamic range and sometimes the best trade-off includes some saturation.

Overall nice shot, and thanks for tempting me to go there.

Agreed--a nice shot.

Doug

 

[Paradox]

Wow, that looks great.

 

[Tom Rankin]

A histogram of the web version of the pic also shows some saturation both the bright and dark ends. However, this is the processed version--it does not say that the original has saturation (particularly if it was shot RAW).

I'm not judging the saturation--this image has a large dynamic range and sometimes the best trade-off includes some saturation.

Agreed--a nice shot.

Many of the shots we took out west of large mountains were in bright sunshine, and the snow was just pure white. If you try to point directly at it, you usually get everything else too dark. I tended to let the snow be very bright.

So here is the question: If you have to overexpose or underexpose, which is preferable?

I will try to post the original tomorrow.

 

[nartreb]

If you underexpose, you can generally get the details back in post-processing, but if you overexpose, those white pixels will never be anything but white unless you paint over them.

In this shot the snow being pure white doesn't bother me very much - snow is supposed to be white. The stream being pure white works great for me. There's just enough snow, though, that I wouldn't mind a little detail within it. I think it depends how big you want to print it. I want to see the foreground pretty big (those ARE trees!) but at my monitor's size, the snow on the right starts looking a tiny bit blank. It's bright enough to draw the eye, too. To me the most interesting part of the photo is the middle distance (I see lots of snowy mountains, but relatively fresh ashen desolation is rare), so I'd be tempted to mute the snow a little to focus more attention on the center.

 

[DougPaul]

So here is the question: If you have to overexpose or underexpose, which is preferable?

First of all, if you want details on the snow and details in the shade or on dark rock, you are dealing with a high dynamic range scene. (I'm not going to talk about HDR images (combinations of images made at several different exposures) here.)

Some cameras have a special mode for high dynamic range scenes. These modes have a "soft saturation" (reduced contrast for near-saturation parts of the image) to reduce the amount of saturation.

Graduated neutral density filters can reduce the range of some scenes (eg bright sky, dark land with a relatively straight dividing line). ND filters can be simulated to some degree in software if the info is in the original image.

If your camera supports it, shoot RAW (or RAW + JPEG). RAW images have more dynamic range than JPEGs. Printers and screens have even less dynamic range and you then have to process the image in a way that maps the dynamic range of the RAW image (generally >=10 bits) into a reduced range image for the printer or screen (8 bits).

Start by "shooting to the right" (setting the exposure to move the histogram as far to the right as possible without saturation) (See http://www.luminous-landscape.com/tutorials/understanding-series/understanding-histograms.shtml and http://www.kenrockwell.com/tech/histograms.htm) Color histograms are better than "intensity" histograms (which are often just the green channel). This maximizes the shadow detail (limited by noise and quantization) without losing highlight detail to saturation. Sometimes, of course, the range is too great for the camera--either you have to guess at the best trade-off or take several pictures at several bracketing exposures and choose later.

Once you have captured the image, you now need to process it in ways that map the original intensities into a range that looks good on the screen or in a print. I've mostly just altered the global density scale to do this, but one can also mask out regions of the image and alter the local density scales. I haven't done any regional alterations or anything more sophisticated, so I'll leave that to others.

Doug

 

[Tom Rankin]

Thanks for the feedback. In this case, it was a low end Point and Shoot. I have a bigger camera (older), with more advanced features. In retrospect, I really should have brought it along, the increased weight and bulk notwithstanding. Anyway, here is the original image (3.5 Megs)

 

[DougPaul]

Thanks for the feedback. In this case, it was a low end Point and Shoot. I have a bigger camera (older), with more advanced features. In retrospect, I really should have brought it along, the increased weight and bulk notwithstanding. Anyway, here is the original image (3.5 Megs)

A quick check of the original* shows some of the snow to be saturated, but nothing else. Large portions of the sky are just below saturation. It looks like the auto-exposure did a reasonable job.

* Of course, the "original" here is a JPEG, not the RAW image so I cannot analyze the true original. Of course a significant amount of processing occurred inside the camera to make the JPEG from the internal RAW.

However, the minimum pixel value is ~30 (out of 256), giving an overall dynamic range of ~3 stops (pretty small!). Perhaps some of the shadows were filled by haze or this is a camera limit. Your post-processing expanded the dynamic range, perhaps to remove the haze but mapping some of the shadows to 0 in the process.

Another, and perhaps better, way to increase the dynamic range and shadow detail would have been to reduce the haze with a polarizing filter, had one been available.

In 20-20 hindsight, it might have been nice to have bracketed the exposure or dialed in some (-1/2 stop?) exposure compensation. If your camera can display a histogram and/or flash saturated pixels, it may be worth checking it, dialing in some compensation, and taking another shot. (Some photographers put some compensation (typ -1/2 to -1 stop) in their standard setup to minimize the risk of saturation. However, this may reduce the effective dynamic range of the camera.)

Your choice to increase the dynamic range is an artistic decision--while I might have stayed a little closer to the original density scale, the net effect of your changes makes the image a bit more dramatic than the original.

Doug