What kind of post-processing does NatGeo do?
June 26, 2010 5:27 AM   Subscribe

What are the RAW-processing tricks used in magazines like National Geographic? For example. I can't figure out how such rich tonalities and color vibrancy can be squeezed out of the same RAW files I'm producing...

I've been taking photos for a number of years now, consider myself competent with Photoshop and a number of RAW-processors, regularly use pro-level cameras/primes like the Nikon D3s and Canon 5D, but for the life of me, I can't get similarly varied tonalities and color vibrancy like what I see in NG. I know that photographers submit all their RAW files, unprocessed to NG during assignments. I recognize that these are probably trade secrets, but I can't be the only one to be wondering about this...
posted by Muhim Nest to Media & Arts (24 answers total) 37 users marked this as a favorite
National Geographic seems to use high dynamic range imaging frequently. (Basically use auto bracketing, then merge the images taken to get a larger range of lighting.)
posted by anaelith at 5:55 AM on June 26, 2010

Nat Geo's Imaging site actually offers a processing service for RAW files.
posted by gman at 6:01 AM on June 26, 2010

No, those sample images are not HDR, the technique requires a still subject and those are portraits and live action shots.

I didn't see anything particularly unusual about those shots you linked to. What I see are photos taken with high quality photo equipment, particularly lenses that have excellent color saturation, an excellent photographer with an eye for subtle light and color, and good postprocessing color correction.

The service gman lists describes correcting for chromatic aberration and vignetting is a feature of the (very expensive) Photoshop plugin DxO Optics Pro which does some truly wonderful things. But that's about the most advanced thing I see in their standard services.
posted by charlie don't surf at 6:25 AM on June 26, 2010

National Geographic seems to use high dynamic range imaging frequently. (Basically use auto bracketing, then merge the images taken to get a larger range of lighting.)

I'd doubt that for any of the journalistic images in the magazine. I don't know if he still uses it, but he McCurry used to be one of the masters of slide film. That right there will get you a lot closer to great colors than a digital camera. The pictures in the linked gallery don't look like anything spectacularly out of the ordinary, color-wise. Solid exposures, good use of light, toned so there's a good black point and data throughout the entire tonal range.

With a digital camera, you'll find an underexposure will help out. Overexposure is hard, if not impossible, to recover from. Once the data has gone full white, there's nothing there. The blacks have a lot more data in them, in my experience.

Then, use a combination of levels/curves, the history brush, hue/saturation, and selective color, all with a healthy helping of the history brush to get different elements of the image just right. Here's a conversation with a couple of friends who shoot for newspapers and have a reputation for color. Go to the part that says "uses careful lassoing." Their portfolios: Chip and GJ.

Here's an example of darkroom printing instructions for a black and white photo. That looks really complicated, and the printer isn't even worried about color. It takes a lot to make colors look good in a photo.

One thing that really helps is, when you're in the levels dialog adjusting the white and black points, hold "alt" (windows) or "option" (mac) while you drag the sliders. The image will turn into a map of where full black are in the image. You'll want some full black in the image if you want the pictures to have that crisp color look.

You also need to be aware of how different programs/viewing situations affect the colors. National Geographic's pre-press people are probably among the best in the business. They are amazing at getting colors right for the magazine. But, it takes a pretty different approach to get colors right in a web browser. Some browsers have color management, others don't. It's really a gamble.

It all starts with a good, relatively flat exposure that has data both in the shadows and highlights. Experiment from there.
posted by msbrauer at 6:26 AM on June 26, 2010 [7 favorites]

What I see are photos taken with high quality photo equipment, particularly lenses that have excellent color saturation

Is this a real effect? Physically speaking, how can a lens possibly change the saturation of an image? Wouldn't that require something like fluorescence?
posted by teraflop at 8:05 AM on June 26, 2010

Physically speaking, how can a lens possibly change the saturation of an image?

There's a lot that goes into lens construction, and I imagine variety in quality of lens materials (quality of glass, plastic, number of elements) and lens coating could affect color.
posted by msbrauer at 8:11 AM on June 26, 2010

Best answer: Oh, and having looked at a few more pictures, he's sometimes exploiting what one might call "bad light" to give interesting color characteristics to the image. The picture of people around a fire, for instance, or of the family with the cart on a dirt road, are both taken in what is likely pretty dim, diffused light situations. If you wait until the blue of twilight, after the golden hour, you'll get a different color pallet in your pictures. If you shoot in a poorly-lit room, you'll get some interesting colors. All of this, mind you, will happen only if you're manually exposing, rather than relying on the camera's built in exposure modes. Spot meter something that you want to be about in the middle tonal range of the picture (if everything were grayscale) and expect that you'll get a few darker and brighter levels on either side of what you meter. You'll get some interesting colors that way.

As you play with exposure, you'll notice that a red shirt will look one way if underexposed, another way if metered for 18% gray, and still another if overexposed. You'll get another color character, too, if you use a flash. Or if the light is diffused or direct, or anything else.

Sorry this is all starting to get complicated. Quality of light matters quite a bit, but even if you're working with what one would call bad quality light, it might make for good colors. Tiana Markova Gold is one photographer who exploits bad light to make good colors (her images aren't always SFW...).

Bad light is harder to use with digital equipment....
posted by msbrauer at 8:35 AM on June 26, 2010

Joe McNally and Jim Richardson are both shooters from NatGeo who blog. Both talk about how they get the photos they do.

One trick is to Expose (to the) Right (but no further). As msbrauer points out, overexposure will kill your image in a hurry.

As for some lenses having better saturation than others, it's real, but it's more a matter of the lens not killing the saturation that's already there with lens flare (unless you want to be JJ Abrams).

The common thing in both cases is that once you've lost data, it's gone, and there's no getting it back. Capture the best data you can, which will often look flat straight out of the camera, and then process it to get the look you're after.
posted by DaveP at 8:40 AM on June 26, 2010 [1 favorite]

Solid exposures, good use of light, toned so there's a good black point and data throughout the entire tonal range.

I think that is key. Digital has less dynamic range than film (I guess, I've never done to work to prove it). So the key is to shoot the source correctly to begin with, never letting anything important "fall off the edges" of dynamic range, but while also making sure to use all of the dynamic range at your disposal. Easier said than done, of course.

In that first shot, in real life, that button might be white, but it also might not be quite as purely white as it appears. So I'd bet that the photographer picked that as a spot to limit the exposure on. Then, in the background in the upper left, the photo is very nearly black. In real life, you could probably see more, but it isn't necessary, so they didn't use it.

In other words, they looked at the scene, figured out where their arbitrary 100% white and 100% blacks were, and set up the camera to capture everything in between. (Proper lenses and aperture and exposure.) By not wasting any dynamic range, everything that is inside that dynamic range is much more dynamic.

Putting it in numbers, let's imagine the camera can capture 16 shades of light and dark. Every pixel in the sensor sees the image and chooses which of its potential values to report. If you can manage to both not overshoot or undershoot, that means every color has 16 potential shades to choose from. If you under or over expose, you might well reduce this to only 10 shades. That means everything looks flatter- where there are subtle shifts in brightness, less range means the camera ignores them.

Now, suppose the image is all black to 3, and all white after 13. You can take those 10 graduations and stretch them back to the 16 in post processing to improve the range, but the graduations are going to still be there. Worse, the 10 you have doesn't map directly onto 16. So your photoshop is going to have to guess. Is the source's 4 going to map onto the finished product's 2, or 3? You end up with just as little range (or steps), but now those steps aren't of consistent differences.

Another thing I noticed in those photos is that the photographer seemed to concentrate on getting the background exposed properly, and letting the subject fall where it may. In the shot where the guy is herding the animals, the undersides of the animals seem a little under exposed. But getting them exposed more accurately would have detracted from the shot, so they left it.
posted by gjc at 9:14 AM on June 26, 2010

To me, the NG photos just look like what a good photographer shoots with good exposure, good technique, and decent equipment. Plus, whomever does the post production work and raw file conversion is very good.

I don't see any particular magic evident.
posted by imjustsaying at 9:38 AM on June 26, 2010

Response by poster: Great to hear everyone else's thoughts on this. Yeah, that McCurry gallery is probably not the best example - this and this are more recent examples that have some of the qualities I'm curious about. With the second (redwoods) example, there's some flash work that's bringing out a 'popping' contrast, but even so, there's some really nice tonality and dynamic range in some of the images.

After proper white balancing, I suspect it's just careful layering of darks, midtones and lights extracted from the same raw file, a technique that's easily overdone.

That SShooter discussion actually did bring to mind a few things - paper quality matters hugely, especially in bringing out the darks, allowing for a sense of depth. My strongest impressions for most of this work is in print format, where there's also the benefit of having size and detail for proper viewing.
posted by Muhim Nest at 9:42 AM on June 26, 2010

The redwoods photo doesn't look like anything special to me exposure-wise. The cave shot with the guy floating in the water looks like the photog set up a LOT of flashes to get that diffused light throughout, probably spot metered off the swimmers skin and took bracketed exposures.

I think you are seeing the results of technique more than technology. I also think they are shooting with the newer Nikon cameras which seem to give very rich and warm colors for a digital [/slideshooter]
posted by fshgrl at 10:30 AM on June 26, 2010

A lot of it is related to knowing how a camera works, and how light works in the camera. This comes from experience and shooting lots of photos in different situations.

The photo you linked to is shot either at dawn or dusk (check the shadow of the child on the wall behind). As a result the light was probably rosy/golden at the time. Also, if it was shot just before actual sunrise/set, the light will be diffuse and softer. The light may also have been diffusing through dust/pollution. The light level was quite low, looks like it was shot on a telephoto with a wide aperture, which helps to bring out the detail in the subject and also to blur the background; and isolate the subject against the background. The color palette is also good - red/brown/yellow.

If I had been in the same room at the same time, I might have seen the same opportunity. You know it could be good, so you set it up and take it.
posted by carter at 10:32 AM on June 26, 2010

In other words ... you can figure out how to set up a shot that could be potentially good, without even pressing the shutter. Once you take the shot, you then have great material to further tweak in the darkroom, photoshop, etc. You can tweak a great photo out of an already good shot. It's harder or impossible to do so from an average shot.
posted by carter at 10:40 AM on June 26, 2010

Dramatic lighting, creative white balance, increased contrast, and added vibrance.
posted by kindall at 10:43 AM on June 26, 2010

In your question you talk all about cameras and tech and nothing about exposure and lighting. This is the problem--all the examples you point to undoubtedly have post processing, but a lot less than I think you think they do.

Think about it this way, photographers have been taking pictures like the ones you linked long before digital manipulation became as good as it is today--and if they were shooting for magazines, they were also shooting slides, not print, which meant everything had to be perfect (or almost perfect) straight out of the camera.

So my answer is this: learn to deconstruct images from an exposure and lighting perspective, not from a post-processing perspective.

e.g. Deconstructing the shot of the girl. From the shadows on the back of the tent, the main light source was low. The softness of the light came from a large and diffuse light source. There is very little or no light fall off between the girl and the background, this tells you the light source was very far away. The color is very warm across the frame, this tells you the light source was kinda orangy. Put all this together and you get a picture that was taken near dawn or dusk, with the sun low in the sky behind and to the right of the photographer.

No fancy camera was needed. The photog may have been able to take this picture with a point and shoot, and could have easily taken it with a cheap, two year old low-end DSLR.
posted by volition at 10:56 AM on June 26, 2010

Just to follow up, yeah, you would be surprised what a difference a great lens makes, especially in color saturation. And these guys can afford the best. For example, this is the difference between plain Canon lenses and their superior L Series lenses, they are vastly sharper and have better color saturation, partly due to reduced lens flare (as others mentioned) but also due to other complex factors (e.g. reduction in chromatic aberration, etc). And they cost a bundle. A Canon 50mm f/1.4 EF lens costs about $350. The legendary Canon 50mm f/1.2 L lens costs $1500. Nowadays, if you have a great DSLR, you're wasting it without a great lens. Modern DSLRs are almost at the point where sensor resolution exceeds the resolving power of even the best lenses.

Anyway, "volition" is right, in art school the slogan was "don't take pictures, MAKE pictures." The way you make a picture is by "seeing photos" not by taking them. "Seeing" is the elusive art of the photographer, he learns by experience to see the light that makes a picture, rather than see the subjects IN the picture. Once you get the light right, you can work with the subjects.

This is largely a classical art taught as "pre-visualization" in the Zone System school of photography. This is actually MORE important today than it was in Ansel Adams' time. The whole concept is previsualizing the result in the medium it will end up in, and setting up the recording medium to optimize results in that final output medium. Today, we call this "end to end color management" although managing color is only one of the required techniques. This is particularly important for high-end print magazines like NatGeo, which is well known for using a higher line screen (200 IIRC) than other color print magazines. This gives a superior image quality. I know of magazines that used to print 250 line screens, and would not accept any negatives or transparencies smaller than 4x5in (8x10 preferred).

Anyway, like others have said, there is nothing particularly magic about these photos, except that these guys are pros, they have great equipment, great subjects and locations, and know their job well. I even found one particularly crappy image in that water essay you linked, it had a washed out sky. Oh how many hours I spent retouching washed out skies, for a while it was my professional specialty. But it does not look so bad in context. One photog told me that "Photography is the art of making ugly things look beautiful." But that might just be because he did a lot of group portraits of ugly rock bands. Ha. Well anyway, photogs will tend to talk your ear off about techniques and equipment, because what they do, what we call "seeing" isn't really something we can describe. We show it in our photos, and it's the LEAST evident of our skills. Hiding it is what makes it look effortless. And it isn't effortless. Quite the opposite.
posted by charlie don't surf at 1:36 PM on June 26, 2010

Not to put too fine a point on it, but Steve McCurry or any of these guys could take a vibrant picture with a 40 year old Pentax and a roll of slide film. The issue isn't the post processing, it's the pre-processing. Finding interesting subjects, finding good light to put them in (right place, right time of day, etc). And then, as Robert Capa said "Liking people, and letting them know it". Making the subject comfortable.

That, and any photograph taken in India is better than any photograph not taken in India. The light there tends to be vibrant and warm. A lot of buildings have interesting windows and textures and the culture is fascinating.

Hard to find a guy with a sword sticking out of his belly in NY to take a picture of.
posted by sully75 at 1:59 PM on June 26, 2010 [1 favorite]

You know, maybe your problem is like mine. I just love film. I was shooting like crazy when it was slides....I loved everything about it....I loved having to adapt to the limits of the media, I loved the look, loved picking up a bag with 25 or 30 or 40 boxes of slides in it every week.....digital really just leaves me totally frigging cold.....sure, it's more versatile, it's this, it's that, I can make 100 arguments FOR it, but I don't enjoy shooting that way.

As I said, I was doing like 25 rolls a week for years, just for the fun of it. Since I went digital? I shot like crazy for the first year or so, and now? I have to think for a minute where my gear is physically located in the house. I dislike everything about it now, from the pressure to instantly review the shot and the attendant knowledge that I don't have to really be on the top of my game bc I can fix it in post, to the whole process afterwards...I can do anything in Photoshop, but it's such a DRAG for some reason....I just can't bring myself to do it anymore.

Sorry to go all ranty. Just a thought.
posted by nevercalm at 3:45 PM on June 26, 2010

nevercalm- you just have to relearn it the same way you did with film. Take notes, use the math and the rules of thumb, and then, if you are lucky, you won't have to do anything in post. That should be the goal- no post production necessary.
posted by gjc at 4:31 PM on June 26, 2010

Light is obviously the most important element and really great photographers, like they have at NG know how to master it. Photoshop processing is no substitute for being a master at photography. I think a lot of newer digital photographers should go back to some of Ansel Adams ideas about Zone Photography and try to pre-visualize then do precise metering to get the tones they want instead of doing several "hail Mary" exposures they can process in Photoshop. A good photographer doesn't want to try and get detail in every single shadow.

Also remember that NG photographers shoot primarily for printed publication. That requires a different way of processing than strictly for web viewing and even different than for ink-jet printing.
posted by JJ86 at 6:00 AM on June 27, 2010

You know what it is? I just hate all the technology between me and a good shot. I'm no luddite at all, but it's just not as appealing. I also made the switch from Nikon to Canon bc at the time, Canon D was where it was at, and I've never found any of their cameras to feel as good in my hand as my last Nikon....that thing was like another finger.
posted by nevercalm at 10:57 AM on June 28, 2010

Well, nevercalm, that is the crux of the issue. The Previsualization method (a la Zone System) requires you to know enough about the technology that you can use it optimally in any conditions. Then, essentially, the tech becomes transparent. Once you know enough about the theory of the equipment, and how it works in practice, you can look at the scene you're photographing, and know how use the equipment to produce the result you are previsualizing in the media you're targeting.

Just to give you an example.. I used to shoot a lot of B&W for crappy rag newspapers with really crappy printing. The line screens were really coarse, so it did not have a lot of tonal range. So I knew I had to produce fairly contrasty images to get them to look half-decent on newsprint. One approach was to look for scenes with contrasty lighting that would naturally look contrasty when printed, and would look good in a contrasty newsprint repro. Then the other approach was to shoot low contrast negatives so all the midrange tones were shoved together, that way, when printed in the contrasty newsprint, the tones would separate relative to each other, giving the illusion of more tonal scale than was possible. I'm not sure if that's too clear, but perhaps you get the idea. This is really hard to explain without visuals.
posted by charlie don't surf at 3:43 PM on June 28, 2010

I used to work at a photo agency, I know the tech. I just liked the older tech. I've never learned to resist the temptation to check the image on the back of the camera, which totally ruins the flow for me. And so on.....
posted by nevercalm at 4:09 AM on June 29, 2010

« Older Mobile internet in Spain for cheap?   |   zoom zoom...boom? Newer »
This thread is closed to new comments.