snow and evergreens

[todd_west]

I've started working on this winter's project of shooting in the

snow. The subject matter is mainly evergeen trees and snow. I've

gotten the first round of slides back and, not surprisingly, the pos

film is having a hard time holding both the highlight detail in the

snow and shadow detail in the trees on sunny or lightly overcast days

(EV 14 or 15, ISO 100). Exposing for only highlight or only shadow

detail is not really an option; blown out snow with good trees or

blocked up trees with good snow both leave something to be desired

most of the time. I'm shooting RSX, which has the widest exposure

range of any E-6 (around 8.5 stops), and am already using a polarizer

to knock down the highlights as much as I can. Using an ND grad is

not feasible since snow and trees are mixed throughout the frames.

 

The only way I know to get more latitude is to switch to neg film.

Looking through photo.net, the preferred reversal film for this type

of photography seems to be Agfa Optima II; it has accurate color

rendition and the moderate saturation and contrast I look for in a

film for general nature work. What I'm seeing in the data sheets

supports this, so I'll be trying Optima 100 in 120 to see how it

does. Agfa has, however, discontinued Optima as a sheet film and I

would prefer to shoot 4x5 over 70mm. Unfortunately, the other 4x5

neg options are Portra and NPS, neither of which at all has the

qualities I'm looking for in a nature film.

 

My inclination is just to stick with RSX for 4x5 and go out and shoot

(probably with some Optima handy in a roll film back), but I thought

I'd put out a ping and see if anyone has any other ideas for dealing

with this situation.

 

Oh, and before someone points out the film doesn't matter because

anything which blows the dynamic range of film will blow the range of

a print, be aware the lab I use is now doing drum scans to a high end

Durst photoprinter as their standard print process, so compressing

film levels into what a print can handle is done more or less

automatically---I've already done test prints and gotten results

which are as good as what's on the RSX; the limitation's what I can

get onto film. Art Wolfe uses the same lab, so this shouldn't be a

surprise.

[scottfrey]

If it's going to be drum scanned and digitally printed anyway, expose a set of brackets (expose one for snow detail, and one for tree detail) and combine in Photoshop, like this one<div></div>

[david_henderson]

I've been thinking about this post since there are a few things about this problem that puzzle me.

 

To start, let me explain that I also have prints produced from transparencies after drum scanning, albeit on a LightJet, though my preference is for Velvia and Provia which sadly give me limited contrast ranges. I'm clearly familiar with the issue you face here. After spending a couple of autumns in Colorado, I have a number of images with trees, both coniferous and deciduous, in snow and these are correctly exposed with detail in snow and without undue blackness in the shadow areas. Obviously I try and avoid taking these shots in very contrasty conditions, but I do have several nicely exposed transparencies made in full sun, which means that the contrast range in these shots must have been five stops or less otherwise I'd have had real problems.

 

So, I guess my first thought is, have you measured the contrast in the scenes you're trying to take? It might be less than you think since snow can often reflect light up into the tree, lightening the shadows. Personally I'd as soon photograph conifers in snow as without it because of this effect.

 

You say you're having a problem even when using a film with a contrast range of 8.5 stops. That's one hell of a lot of contrast. I went into my garden today and took spot readings from a lit, recently painted white wall, and from the shaded undersides of some dwarf conifers and I got a 6 stop difference. I've never used RSX but I did look at quite a few threads on it and downloaded its tech spec. I can't find anything that tells me that it's contrast range is as high as 8.5 stops or indeed any particular claim for much higher than usual contrast capability. I think that if anyone was selling a competent slide film with 8.5 stops contrast ratio then

 

a. they'd be making a huge of a fuss about it

 

b. Lots of people would be using it and talking about it.

 

What people do say about RSX is that it's pretty neutral - and interestingly that it's very prone to react badly to processing that's less than 100% fine tuned. In short, whatever the contrast ratio is (and my own bet would be 5.5 stops not 8.5 stops but hey, I'm guessing) it might be that you're not seeing all of it and if so this could be a big cause of your problems.

 

If you have found a slide film that has a contrast ratio of 8.5 stops then there isn't a lot of point switching to colour print film.

 

So the following thoughts emerge to try,

 

First, measure the contrast of the scenes you're trying to capture and tie the film choice to the result. About five stops difference or less and use slide film. More than this use print film. More than about 6/6.5 stops either don't photograph, or accept the inevitable, or switch to b&w.

 

Second, if ( as is suggested on photo.net) RSX is prone to processing variations, give something that tends to be processed more reliably like Astia a try if you like a fairly neutral palette, or Velvia if it's dull and gloomy. You're going to be limited in contrast, you may as well see all you can get.

 

Third, I'd be careful with the polariser in these situations, and judge it shot by shot. In my experience you won't always gain more in reducing glare than you'll lose in blocking up the shadows. I know this isn�t the theory, but polarisers often seem to me to have a slightly stronger ND effect in areas where there�s not much light or colour.

[steve_thompson]

Pre-exposing does work. When I do it I set the lens to focus at infinite and aim it at something very close on the first exposure. Then expose the second time.

 

In this scenario I think cloudy or overcast light is the best option. Just remember to keep the sky out of the shot. Sunrise shooting too should help.

[todd_west]

<p><i>So, I guess my first thought is, have you measured the contrast in the scenes you're trying to take?</i>

 

<p>Pre-polarizer, the snow's mainline is two to three EV up and the trees two to three EV down, so most of the image is in the four to six stop range. Overall detail in the snow and trees is fine and the prints are certainly acceptable as they are now. The problem is in getting highlight detail on fresh snow faces angled for optimum reflection while also holding enough shadow detail in shaded trees to keep them from blocking up close to the trunk. I'm being something of a perfectionist here in trying to keep really bright faces and really dark trunks from washing out.

 

<p><i>That's one hell of a lot of contrast.</i>

 

<p>Not really; it depends on how much of the toe and heel you're using. With a 4.1 Dmax and levelization, it's all accessible. I figure Ektachrome at 10^2 (6.5 stops), most Fujichrome at 10^2.25 (7.5 stops), and RSX at 10^2.5 (8.3 stops). Most neg films are well over 10^3 (>>10 stops). I don't know of any non-digital way to compress this into the <10^2 range of paper.

 

<p><i>interestingly that it's [RSX's] very prone to react badly to processing that's less than 100% fine tuned</i>

 

<p>Yes and no. Take a look at the E-6 and AP-44 datasheets sometime; AP-44 is essentially a precise specification of Kodak's preferred E-6 processing parameters so, in principle, AP-44 films should process well on just about any E-6 channel. That notwithstanding, there are about 2.5 labs in the pacific northwest which I've found process them well. The one I'm currently using has significant problems with Precisa, but seems to do slightly better with RSX than the other two. I've not had much success ferreting out proprietary process details, but most of the problems at good labs seem to stem from the labs tweaking their processes to compensate for the high blue and depressed green response of current Ektachrome and Fujichrome emulsions. Since RSX is well balanced to start with, doing this screws up its color balance.

 

<p>At any rate, I occasionally shoot E100S and Provia 100F as controls and have found the data sheets for slide films to be very accurate predictors of E-6 film behavior. Astia, in particular, would not be a good choice, as it has slightly less of a contrast range than 100F (the situation's similar with E100S and SW).

 

<p><i>I know this isn�t the theory, but polarisers often seem to me to have a slightly stronger ND effect in areas where there�s not much light or colour.</i>

 

<p>I'm inclined to disagree. I'd not thought of this before, but shadowed or dark areas on sunny days are lit by singly scattered light from the sky while highlights are lit by direct beam radiation. So the fraction of polarised radiation off a diffusely lit surface is higher than one with specular lighting and more of it will be rejected by the lin pol grate in front of the circular polarizer's quarter wave plate. I will look at this more the next chance I get and will try to borrow a linear polarizer to see if not having the phase shifter makes any difference.

 

<p><i>In this scenario I think cloudy or overcast light is the best option.</i>

 

<p>For controlling contrast I quite agree. Unfortunately, it also makes capturing the relief in surface contours such as flowlines, small sastrugi, incipient suncups, frost layers, and the like much harder as the light is not nearly as sculptural. This is where a lot of the life and interest in snow is for me. Clouds also dramatically drop the relief in icefalls, crevasses, seracs, and nieve penitentes.

 

<p>Fill flashing a line of trees several hundred feet away is kinda tricky, too.

 

<p><i>If it's going to be drum scanned and digitally printed anyway, expose a set of brackets</i>

 

<p>Or just shoot Portra and Photoshop for saturation, neutrality, and the rest. Much less work than stitching images together, even with an assisting image overlay plugin---I spend all day sitting in front of a computer and go outside to take pictures to get away from it. Besides, paying $5000+ for a Dmax 4+ scanner which can handle 4x5 or paying $40+/scan isn't high in the priority list---the Epson 2450 would be adequate for neg films and low Dmax slide films (<i>e.g.</i>, not RSX) but only has around half to quarter the resolution I'm looking for (and the 3600dpi follow on doesn't look to be available any time soon).

[guytal]

Todd,

<br>

I shoot in snow quite a bit. Transparency film lattitude is indeed an issue, but some films are better suited to a certain spectrum. I find that despite the narrow lattitude, Velvia works very well in the type of scene you describe, because of its propensity to saturate greens (making the trees stand out against the snow). I make an effort to shoot in even light (low sun or overcast skies) and take special care not to blow highlights. Shadow detail, even if rendered too dark, is easier to correct.

<br><br>

Here's a couple of samples:

<br>

<a href="http://scenicwild.com/scenicwild/html/gallery/showimage.jsp?cat=b000623">Wasatch Mountains Winter Scene</a><br>

<a href="http://scenicwild.com/scenicwild/html/gallery/showimage.jsp?cat=c000031">Winter Blues</a>

<br><br>

An important consideration is to use multicoated lenses which really help with color casts and flare, especially in higher elevations.

<br>

On occasion I will also use Provia F, when more lattitude is required. Knowing it produces blue casts, I will usually use an 812 or 81B filter over it.

<br><br>

I'm really curious about your statement regarding 8.5 stops of lattitude on RSX (which I admittedly have not tried yet). Where is this fact documented?

<br><br>

Hope this helps,<br>

Guy<br>

<a href="http://scenicwild.com">Scenic Wild Photography</a>

[todd_west]

<p>It's unfortunate I don't have a low-cost 4x5 scanning ability to post some of the shots in question (I don't know anyone with a flatbed and transparency unit ;~). Guy's Wasatch shot is a pretty good example of the contrast range I'm dealing with; the sunlit mountains in back are way up and there are plenty of deep shadows in the midground trees. I'm trying to hold decent shadow detail without blowing the sunlit snow faces in compositions which aren't nearly as amenable to a grad. I would agree that, in general, it works better to drop the shadows than lose the highlights.

 

<p>With the exception of my 31 year old single coated 90 f/8.0 Schneider Super Angulon, all my lenses are multicoated. The main ones I'm using for the photographs in question are an EBC multicoated Fujinon 150 f/5.6 W and a 240 f/9.0 A EBC. I don't have any hard data, but their transmissivity looks comparable to that of my Hoya S-HMC filters. Definitely higher and less prone to flare than any of my NIC Nikkors, including the remarkably flare free 25-50 f/4.0. Certainly, shots in the f/22 to f/32 range (generally considered optmium for large format lenses, though I've not done any MTF testing) are quite sharp and have no obvious flare.

 

<p>I do have one shot taken pointing up close to sun in the middle of a snowfield which has a blue ghost in it. I believe it comes from sunlight reflecting off the top of the (matte black rubber) lens hood, bouncing down to the (matte black) bottom of the polarizer ring, and then hitting the lens and being imaged as a halo ghost by the pinhole formed by stopping down to f/45. It's subtle and I didn't notice it in checking the shot with the lens stopped down so it's also possible it's an internal ghost formed by image light reflecting off the off film and then scattering off the (matte black) bellows or, more likely, the back of the lensboard (also matte black). I'm planning to look at this more the next time I have a lot of fresh, sunny snow to stand in the middle of but at the moment I can't think of any mechanism by which internal scattering in the camera or lens could cause problems in the ghost free shots I have.

 

<p><i>I'm really curious about your statement regarding 8.5 stops of lattitude on RSX (which I admittedly have not tried yet). Where is this fact documented?</i>

 

<p>Same place as any other film; the exposure density curve. The usable latitude numbers I give above are based on processing and printing at the lab I'm currently using (ProLab; <a href="http://www.digitalimaging.com">www.digitalimaging.com</a>) and my personal comparisons of high contrast slides under the loupe and the resulting prints in the range of 5x enlargement.

 

<p>If you're looking for a specific number in the RSX data sheet which says 8.3 stops (or 8.5), you won't find it. Unless someone wants to loan me a densiometer and a <1 degree spot meter so I can measure various areas of a source image and the resulting slides and prints, you'll have to make your own assessments of workable contrast based on your films and image pipeline.

 

<p>As an aside, the 10 stop contrast range shown on most neg film data sheets is 60dB of dynamic range and the film is a pretty linear transducer from incident light intensity to halation density. In general, 80dB (14 stops/14 bits/4.0 Dmax) is fairly managable with good tools; it's at 100dB or so things tend to get tricky. Unless neg film does really wonky things above the upper limit of the data sheets, there's no engineering reason why it wouldn't be possible to pull 12 stops or more latitude out of a neg film and define some mapping which compresses that into the <40dB dynamic range of a print. Choosing a conversion which makes the print look good is the truly hard part; I've at times worked on color maps for rendering data with 100 to 120dB dynamic range on to CRTs (>50dB) and getting good detail over more than 80dB or so is tricky.

[canon man]

HMMMMMM. Snow and trees...why not try Black and White? I like the detail I can get with TMAX 100. Enlargents look tite all th eway up to 11x14, so in the medium format the detail should be awesome on that format! I'm using a Canon EOS 630 wit h a couple of USM lenses and getting awesome results.