Late last year a good blog friend wrote asking whether I could help with some questions about photographing with a digital SLR. Some of his questions I'd encountered before from other people, and it's likely many others share a similar confusion. I know I did!
I've paraphrased his questions (sometimes to the extent of taking liberties with them) and answered them as best I can in this post, trusting at least some of my responses will prove useful to at least someone out there.
Why it is better to shoot in RAW?
RAW files are the data the camera's sensor collects. They're compressed, but with no loss of data, whereas JPEG files discard some of the data (that's why you should never resave JPEGs more than is absolutely necessary—each time you save a JPEG, you lose more data). Moreover, RAW files are not processed—no sharpening, no saturation adjustment, no noise reduction, etc. They're just the information the sensor records. You apply those corrections later using computer software that's much more flexible and powerful than the in-camera processor. Therefore, many settings on the camera don't matter because you can change or apply them later; moreover, if you don't like the result of the conversion, you can try again with different settings. Colour balance is a good example; I leave it set to “auto” (or, very rarely, might choose a fixed colour temperature) and set the colour balance later on the computer.
When using RAW files, don't you need to go through a complicated process and use specialized software to work on the images?
No, and yes. The process isn't particularly complicated, but you do need specialised software to convert RAW files into editable formats. Very simply, all you do is upload your RAW files from the card onto your computer and then into a RAW converter program, apply the tweaks you want (sharpening, colour balance, exposure corrections, etc.) and transfer the files to an image editing program like Photoshop.
Most (all?) SLR cameras come with a RAW converter program, but Photoshop (including Elements) can also do the conversion for you. The other option is a stand-alone RAW converter. I use Adobe's LightRoom (v. 1.2) and find it excellent. Some photographers swear by Breezebrowser or Bibble, and Capture One Pro seems to be the top-of-the-line converter if you can justify the expense (I can't). LightRoom has several modules, but the essential, RAW conversion functions in LightRoom seem identical to those in Photoshop so you could get away with PS alone. I do like LightRoom, though.
I've heard that LightZone is great for RAW processing.
It looks excellent but I've never tried it and can't afford it, particularly now I've bought Photoshop CS3. LightZone might be a better option than Photoshop for photographers, particularly those who don't want to invest an inordinate amount of time learning the latter. LightZone seems to be based on Ansel Adams' Zone System; moreover, unless I'm mistaken, its major advantage is that you can edit selected areas in the RAW image itself—with LightRoom, you can only change the entire image; selective edits have to wait until after the conversion. Photoshop is always likely to offer a greater range of more powerful editing tools, but at the cost of a long, steep learning curve.
Photoshop is apparently being redesigned to make it task-oriented (modular), and given the success of that approach with LightRoom, such a redesign would probably tempt many undecided customers to opt for the industry standard Photoshop rather than LightZone.
What are "embedded or non-embedded JPEG files"?
Digital SLR cameras allow you to shoot RAW files with the option of including a “bundled” version of the image in compressed format; this has the “.jpg” extension. Thus, when you upload the file to your computer, you have two versions of the same file: a RAW file and also a compressed, processed file you can view with any image viewer, including, for example, Windows Explorer. Most programs (including many image editors) don't recognise RAW files—you need a special program to convert RAW files to an editable format. The embedded JPEG files are useful for easily sorting through your photos to see what you've got, and also if you don't have access to a program that can edit RAW files—you can work on the JPEGs instead.
I've heard about using the histogram before, using it for processing in PhotoShop, but have no idea how to read this while looking at the camera. May I ask what looking at the histogram does for getting the image right?
Put simply, if the histogram on your LCD screen truncates abruptly at either end, you've lost details in the shadows or highlights. Truncation on the left side of the histogram means the shadows are blocked—you have solid black and you can't retrieve any detail. Truncation on the right side of the histogram means the highlights are pure white, i.e. no information, no detail. This is usually a more serious problem than blocking the shadows. RAW converters, like LightRoom, do allow you to “recover” some of this “lost” detail, apparently by extrapolating from other colour channels, but it's best to avoid the problem in the first place. If you see your histogram's truncated on the right, take the photo again, with decreased exposure and see if the histogram now falls entirely within the limits. This is called “chimping”. My LCD (and I assume those of most cameras) shows highlight truncation (clipping) as areas that blink off and on, warning you that you have apparently no detail in those areas.
In theory, you should try to keep the histogram as far to the right as you can, without truncating it. On the LCD, the photo might look overexposed, but theory says ignore that—it's the histogram that's important. However, in practice, I find this “expose to the right” rule can lead to clipping one of the colour channels, so if the photo has bright, saturated colours (yellow or red flowers, for example), I try to keep the histogram centred or only slightly to the right. Some of the most recent displays actually show the red, green, and blue channels separately—a good feature which helps you avoid clipping one of those channels.
Of course, if the contrast range is too high, preserving highlight detail by shifting the histogram to the left (decreasing exposure) will block up the shadows. There are fancy ways of solving this, but I don't want to confuse you any more than I already have... ;-)
In short, look at the histogram after you've taken the shot; if the right side's chopped off, decrease the exposure; if the left side's chopped off, increase the exposure.
At what size do you usually shoot your images?
At the largest setting; i.e. maximum dimensions, highest resolution. It's easy to shrink the images, but even the best interpolation algorithms struggle to enlarge very small images. If you find you need to reduce the size or resolution to save space on your memory card, buy another, larger card. Even the best quality cards are cheap, especially if you buy them from somewhere like Flashcards.co.nz (highly recommended—and that’s an unsolicited recommendation for which I get nothing other than knowing you’ll get excellent service).
I want to have the highest quality images with the best resolution for the media that I show them on.
For computer monitors, a resolution of 72 pixels per inch (ppi) is fine; 96 ppi is the best you can get; anything higher is wasted and simply increases the size of the file. Resize your photos to the dimensions you want (for my blog I don't go above 700 pixels wide) at 96 ppi.
For printing, you need a much higher resolution: other things being up to par, 240–300 ppi will give you a beautiful print. Of course, if you try to produce an A2 size print at 300 ppi, you'll have to enlarge the photo (i.e. interpolate) substantially (“up-rezzing” is the jargon). If you do want a print that size, let a good, professional lab do the up-rezzing and printing for you (unless you’re an expert, in which case you’ll get nothing from reading this).
For projection, you're again limited by monitor resolution. Use a resolution of 96 ppi and set the dimensions to fill the screen on the computer you'll use for projection (1024 pixels wide is a good choice if you don't know what computer you'll be using). This is the one area where, in my opinion, film still has a significant advantage. I've yet to see a projected digital image that approaches the quality of a good, projected transparency.
Perhaps the most important thing to do is make sure that your monitor is properly calibrated. If your monitor has, say, a green cast (you probably wouldn't know, because you're used to it), or if its contrast is set too low, your photos on other monitors (especially correctly calibrated monitors) will look weird. It's possible to calibrate by eye (approximately), but hardware like the Huey Pro (which I use) and other systems make it relatively easy and reliable.
Do you use the noise reduction setting or the sharpening in-camera setting?
Sharpening in the camera only affects the embedded JPEG file, and as I work on the RAW file, I don't bother with this setting—it has no effect on the RAW file. On my 20D the only noise reduction option is for long exposures; I've turned this on, but I'm not sure if it affects the RAW file. I don't think it does; I think it affects only JPEG files (embedded or standalone).
However, when I was overseas I had to use the embedded JPEGs because no Internet cafe had a RAW conversion program, nor any useful image editor. Therefore, I set the sharpening to medium, because unprocessed digital files are inherently soft because of the anti-aliasing filter in front of the sensor.
My suggestion is to ignore these settings. Leave them on the factory default settings and use the RAW files. (You might come across recommendations, including some from high-profile photographers, to turn off the sharpening. This shows that even high profile photographers sometimes don't understand everything, even if they think they do. Nor do I).
What camera settings should I use if my photos are intended to be similar to the kind of photos you take? (Some photos I'd like to one day print up on paper, too).
Here's a summary of how I use the important settings on my camera:
ISO: The ability to change the ISO is one of digital's major advantages over film. No trying to change films mid-roll, no fiddling with “pushing” or “pulling” films—just dial it in and carry on photographing.
Choice of digital ISO is similar to choice of film speed. The higher the ISO the more noise you'll have in the shadow areas. This shadow noise is exaggerated if you increase the exposure or add fill light during image editing, but image noise is generally low in most digital SLRs (particularly Canon cameras) so don't fret over it. My inclination is to err on the side of a high ISO—I'm more likely to ruin a photo with a too-low shutter speed than an slight increase in noise. I guess it also helps that I don't wail and gnash my teeth over a little noise; in fact, I often like it.
I generally leave my 20D set on ISO 200, sometimes 400. If I can use, or want, a slow shutter speed I'll use ISO 100 (the lowest for this camera); if there's little light, I'm happy dialing up ISO 800, 1600, or even "Hi" (3200), but as I've said, I rather like noise (my mum reckoned I did, too).
The one minor disadvantage of being easily able to change the ISO is you have to remember to change it!
Program modes: I never use any of those fancy portrait/macro/night scene/fireworks/blahblah modes. I use two modes on my 20D: aperture priority (“AV”) and, occasionally, manual (“M”). I use the manual mode when I want flash as a main light (e.g., for high magnification macro work) or when I want a consistent exposure over several frames (e.g., if you want to stitch several photos for a panorama you need the same exposure for each frame or the joins, particularly in the sky, will show).
Colour balance: Leave it set to “auto”. Sometimes, setting a fixed colour temperature makes the post-processing (editing) a little easier and faster, and it's useful for consistency over several frames (as I mentioned above for “Manual” mode), but I've never found the other settings (“shade”, “cloudy”, etc.) useful. You'll need to correct the colour balance in your RAW converter anyway.
Aperture and shutter speed: no different from a film camera. Try to get these right, and if you're not confident and the photo's important, bracket your exposures one stop or more either side of your (or the camera's) best guess.
Finally, a very important point. When you convert the RAW file and send it to Photoshop, make sure you convert it as a 16-bit file. I won't go into the theory of it; I'll just point out that it allows you to edit far more aggressively without the photo beginning to fall apart, i.e. showing artefacts like “banding” or posterization. Y(ou can use most of Photoshop's tools in 16-bit mode now, but not in PS Elements; in particular, with a 16-bit file in Elements you can't use layers; you must convert it to 8-bit mode.) Often, this doesn't matter, but if you want to alter contrast or saturation, etc., substantially, you're safer in 16-bit mode.
Hope this helps rather than confuses. If something's not clear, just ask, and if you think I'm mistaken about something, please, do query it. I, too, have an enormous amount to learn.
1. Sometimes you don't even need to edit the photos. For example, all but a few photos I posted of this year's Baring Head Bouldering Competition were processed entirely in LightRoom and exported from that program, resized for the blog.
2. However, Lightroom cost me nothing because Adobe distributed it free to Rawshooter Premium users (they bought out Rawshooter and the free distribution was part of the deal). If I had to buy Lightroom, I wouldn't. I'd use Photoshop's converter (Adobe Camera Raw (ACR), which is incorporated into Adobe Bridge.)
3. LightZone 3 currently (11 March '08) sells for NZ$198 (c. US$150) for the Basic version and NZ$331 (c. US$250) for the Full version. During March '08 you can get a discount of 20% on these prices. Still unjustifiable for me, regardless of how much I'd love to run it through its paces.
4. Apologies if this is second nature to you, but for the sake of those for whom it isn't: To decrease the exposure, increase the shutter speed, decrease the aperture (i.e. increase the f-number), or lower the ISO. Note: these work only if the camera is set to “manual” mode; if it's set to, say, AV or T or (heaven forbid) P, you must use the exposure compensation dial.
5. The number and range of quality of websites discussing these and related issues is astronomical. One of the better starting points (probably the best known) is The Luminous Landscape, although some of its earlier articles are now substantially dated.
Photos (click to enlarge them):
1. The Pourangaki River, Ruahine Range, on 3 March, the day we walked out after a great couple of days in a magical place.
2. This was what the Pourangaki looked like less than 24 hours before. A good lesson — like most New Zealand rivers, those in the Ruahine rise fast but fall fast too, so don't try to cross when they look like this. Wait a few hours and enjoy the place.
3, 4. Another river—the Pohangina—a few days later. We came upon these two male whio (blue ducks; Hymenolaimus malacorhynchos) at the upstream entrance to the small gorge between Leon Kinvig and Ngamoko huts. The bird at the back is whistling ("whio" is also the Maori word for "whistle"). Females have an entirely different call, a kind of rattle. We saw 4 or 5 birds on this last trip, all males.
5. Tuatara (Sphenodon punctatus). CAPTIVE, at Pukaha Mt Bruce. As much as I'd like to say I photographed this gorgeous reptile (not a lizard) in the Ruahine, it's a sad fact they survive in the wild only on some of New Zealand's offshore islands. This is Rewa, a female captured on Stephens Island in the 1980s. She still trailed a few scraps of recently moulted skin. I photographed her through the glass of the cage she shares with her mate, Tahu: glass well smeared with the enthusiastic finger and nose prints of many small humans. The trick with this sort of photography is to get the lens as close as possible to the glass and to shade the glass to eliminate reflections or light scattered by those little human prints.
6. On the other hand, I photographed this native land snail (Powelliphanta sp., I think — I'm trying to get an i.d.) at the top of Shorts Track on the Ngamoko Range as we walked out last Monday. The shells, usually empty, can be seen fairly readily, but it's rare to find the living animal and even less common to find one active. Yet another of our wonderful endemic animals in need of active conservation. [Update: Kath Walker, snail expert from the Department of Conservation, confirms it's Powelliphanta marchanti.]
7. Coprosma sp. seedling growing on the river bank below Ngamoko hut.
8. Pohangina River between Leon Kinvig and Ngamoko huts. The gold is sunlight reflected from a grassy slip just downstream; the blue is light reflected from the shadows. Photography is often a matter of seeing the light.
Photos and words © 2008 Pete McGregor