Antarctic Photography Techniques
PART 4: A COMPARISON BETWEEN FILMS.
A few years ago I took some slides using Agfa. The skin tones came out great but ice had a disturbing greenish tone, so I've sworn off the stuff for polar photos. Sorry Agfa. But if I wanted to do portraits it would be a very good choice.
For a long time I used Fuji Velvia almost exclusively. Its rich blues and exquisite detail are wonderful, but I started seeing waaaay too much contrast in my ice shots, thus the impetus for this film comparison. I don't normally shoot with negative/print film, but it has a much wider dynamic range so I thought it should be part of the test too.
The pictures shown are nothing special, just a few representative
samples. In all cases I used the following
conditions:
- No sharpening
- No color or contrast enhancement
- RGB colorspace, no conversions other than film color correction
when scanning negatives.
- "Light" dust/error correction
- A single scan pass; no multi-pass scans
- I used a Canon FS4000US scanner with Hamrick's VueScan version 7.6.64
Because these images are somewhat raw, the colors will appear less vibrant than usual, and not quite as crisp as they might be with a little manipulation or sharpening. The intent was to make the images as comparable as possible, even though the subjects and conditions may have been different in each photo. In each case I present the histogram from each scan; ideally I'm looking for a gentle bell-shaped curve, indicating a broad range of exposure, without any severe spikes or sharp peaks at either the dark end or light end of the spectrum.
I was attracted to this film because of the "High Definition" title. I'm not sure how it compares to other 400 speed films, but I was not impressed. Even so, you need speed if you're going to be taking pictures from the deck of a vibrating, pitching ship in high winds and low light. Some of the pictures I took might not have been possible without a high speed film, but given the graininess I don't know that I'll ever do anything with them except go "ooo-ahhh". Since I doubt I'll ever enlarge these, next time I'll either stick with a lower ASA or just do digital. A 4000dpi scan is a waste of time and disk space with this film; 2000dpi would do just fine.
The colors are distinct. The sky appears very true to its original lighting and color, but the snow is a tad greenish in places and the bow of the ship (low, near the water) is an unnatural greenish too. But to be fair, VueScan didn't offer a color matching profile for 400HD film so I used the nearest match (Vericolor 400), so that may be a cause for some discrepancies. Color matching like this is not anywhere near as much of an issue when scanning in transparencies, of course. In the mass-market prints that came with the negatives the bow appeared reddish, which I tend to think is more accurate. So the color's off, but it's probably not the film's fault. Given the overall grayness of the image it seems strange that the histogram is skewed towards brightness. The graininess shows up most in the darker areas; this is consistent with how film chemistry works: some grains of emulsion are more sensitive than others, and the difference is most pronounced in low-light areas. Regardless of the chemistry, however, I think it's pretty darn grainy for a "high definition" film.
A different picture with the same Kodak 400HD film. Admittedly, this one is overexposed. The white highlights in the detail image are burnt out and the color under the arch is a screwy neon green. The burnt out whites are disappointing because I was hoping print film would be more forgiving on exposure problems. Hmmm... Less latitude than expected. Note the vignetting on this, and to a smaller extent, the previous picture too. This is actually on the film, not an artifact of the scanning process. Just why it's on the film is a whole 'nother question... It's almost unnoticeable on the prints, partially because the prints don't encompass the entire extent of the negatives like the scanned images do. Additionally, the gamma appears to be greater on the scanned images and that's exacerbating the vignetting. I really need to get to the bottom of this as I've seen it frequently on other images as well. It's probably a certain combination of filters and/or lens cover causing it. I wonder how long this has been going on? (Review rule #5: occasionally look at your results!)
Based on this image alone, this is great film. The color is beautiful and, yeah, it's grainy, but it's not bad for a 200 speed film. This was taken off the deck of a ship in low light, without a tripod of course. (Ever try to use a tripod on a ship? Forget it. Engine and wave vibration telegraph up through the legs. Help me make my fortune: let's design a kind of steady-cam tripod for ships.) Even though you can start to pick out the grain in the first detail zoom, the sharpness remains. Click on the second, closer detailed image (118Kb) and you'll see a screen capture showing a 1600% blowup of the raw tiff file, wherein the individual scanned pixels are plainly visible. To me, this looks like the pixelation is right at the grain of the film or that there may even be more detail that the scanner can't pick out. It's not, however, a very good test of contrast issues, so I also picked another image to examine (below).
This image is a better example for comparing contrast between films. At first glance, some of the white highlights appeared to be totally burnt, but when I zoomed in on them I found that there was still detail, if a bit bright. This seems acceptable to me; you're never going to get rid of all burn-out -- we're just trying to minimize it. In my opinion, the crispness, speed and color of the film out more than make up for a tendency to burn, especially since I haven't found any fast slide film that does much better when it comes to high contrast situations. Note the turquoise ice face in the detail image: that's a hard color to capture and this film did a good job. I would use this in low light situations or when I can't use a tripod, but would preferentially use a slightly less grainy film when shooting with more light.
I couldn't hold the camera steady enough on this image, so it probably isn't the best for sharpness comparisons; but for color, contrast and graininess it makes a good example. The color is okay, although it's just not as vibrant and saturated as the slide film (this should come as no surprise to most experienced photographers). It does, however, appear to have a really good dynamic range for high contrast shots. As usual, the darker areas show greater graininess. For a better comparison of sharpness see the following photo...
Keep in mind that these are unsharpened images, raw off the scanner. I think this gives some indication of what the film's potential is. As far as detail, with a cooperative subject, a good lens, and a steady tripod... pretty incredible. Also in this case, like the 200 speed slide film, I'd say the grain of the film and the resolution of the scanner are about the same, 4000dpi. I still have some issues with the color saturation, but I am happy to say that I took this picture and I think a big enlargement would look great. The bird, by the way, is a giant petrel chick, about the size of a basketball, on Hermit Island, Antarctic Peninsula.
This film is a bit strange and I'm not sure what to think of it. By the name, it sounds as if it's designed for portrait photography. Judging by the results I'd say that's the case as well: most of the pictures had a creamy, sometimes satiny or even milky look. It does good with skin tones, but puts a bit of a greenish caste on everything. (This color rendition is very similar to my experience with Agfa slides.)
In this series you can see how the detailed the film is and how smooth the grain. The creamy aspect appears to be in the grain: even though it's definitely grainy, most of the grains are similar in color. The detail goes down below the range of the scanner, greater than 4000dpi. But it also has an all-or-nothing appearance in the dark areas, as if only the blackest shadows get saturated. This effect is not unique to this photo; all of mine show the same thing. It might be an artifact of the scanning process, but it's suspicious enough that I don't intend to use this film again for polar pictures (if I wanted to take soft portraits, it would probably be my first choice).
It handles the highlights very well, it isn't too contrasty (except in the dark areas), and it's extremely detailed. But the funny color and black saturation just doesn't cut it. Darn.
This film has excellent detail. It's not very grainy, and the detail inherent in the film goes down below my scanner's resolution. If you look closely at the last detail photo above you'll see the blocky pixels in the reflection in the eye (this is a crop from a 1600% zoom in Photoshop). Compare the grain of this film with the 200 speed slide film above! If I had a higher resolution scanner I could probably pick out a reflection of the landscape in the bird's eye. Color in this photo seems fairly true, if a bit subdued.
This is a tough shot, with the bright glacier in direct sunlight and the berg in shade: the classic dilemma with polar photography. The glacier is a bit more burnt than I would like, but the shadow details did come through nicely. This film is not perfect, I would still like more dynamic range. But the color and detail are very good, and I find it less contrasty than Velvia (below).
For sheer detail and color saturation Velvia is incredible. It loves blues, which is why I started using it for underwater photography and how I got hooked for land shots too. Unfortunately, it is also very contrasty and doesn't have the dynamic range I need for most Antarctic shots. This is not to say that there aren't lighting conditions and situations where it would be perfect on the Ice, but that for a general-purpose film it's not the best choice. Also consider that the sun in polar regions is often at low angles and glare can be extreme, exacerbating the contrast problem. The above pictures show burn-out whites (too much for my tastes) although the exposure is fairly correct. I've attached a couple extra photos below for additional comparisons. These are probably not the best examples of the color saturation, but they do show the detail possible and the typical contrast. The picture above was taken on Sea Lion Island in the Falklands, and the photos below were taken at the South Pole.
For a truly complete comparison between films I should also have included some examples from Fuji Provia (100ASA), the new 100 speed Velvia, which is also supposed to be extremely detailed, and maybe some other 200 and 400 speed films. But of the films reviewed above, I'll have to go with the Kodak E100G for all-around quality, with the Royal 200 a close runner-up for it's contrast handling and the Ektachrome 200 for it's color saturation.
Since I don't really need resolution beyond 4000dpi, the Velvia is probably overkill. I'd rather have a higher speed and a more forgiving film (for contrast and poor exposure errors). I was going to be exclusively on a ship I would probably look for a good 200 speed film that I could push to 400.