HDR B&W, Digital vs. Film

Posted on October 22, 2011


Peck Hole HDR B&W

Author’s note: This post did everything I hoped, in that it covers a subject I haven’t seen elsewhere; it’s the most popular subject on my blog: despite it’s age it gets regular hits so people must be coming to it from search engines. It took me a long time to come to these realizations and a lot of research for this post. So if you read it, comments are welcome!

Time to get back to some more hardcore photography stuff.

HDR & The Zone System

I was thinking about HDR (high dynamic range) photography the other day. And I realized, HDR is not new. HDR is as old as photography and I use it all the time. We just used to call it the Zone System. In fact, most things we see are in “HDR.”

Let’s start with what dynamic range is. I’m gonna butcher it up a bit here. To be honest, I skimmed a PhD thesis worth of stuff on dynamic range in the last hour or so while contemplating this article. It’s deep and I’m gonna gloss across the top.  Dynamic range is essentially how many zones, or f-stops of data your medium (film, digital sensor, Shroud of Turin, whatever) can capture without losing information.  In other words, how much contrast can you capture?

Typically, shots that go beyond the range of the camera, or photographer, result in  “blocked shadows and blown highlights.” In other words, the shadows are black and the highlights are completely white.  On the one hand, this is really easy to do if you don’t understand what your camera is metering on, so you basically under or over-expose some part of the shot. On the other hand, some things just have such a range the camera just cannot capture it.

One of my first blog posts was all about this, I recommend reading it, but I’ll excerpt the part on dynamic range:

Dynamic Range

Okay, what does this mean? Say you are looking at a scene that has both shadows and highlights. Your black dog in front of your snow man, and you have an in camera meter set to spot meter (instead of some broad average, it will become clear why to do this in a bit). You point at the dog and take a picture. Then you point at the snowman, somewhere where the sun is shining on it, and take a picture. (This is where digital is awesome, you can do a similar experiment right now. In fact, you should do this right now, find something with high contrast and take these two pictures.)

Exposing for highlights and shadows

Here are two pictures that my friend Kalyx took of the Duvall train station in bright sunlight. In the first she metered on the clapboard, getting good detail on the cracking paint in the light; in the second she metered on the shadow bringing out detail in that paint while the paint in the sunlight is starting to get blown out.  These exhibit the typical hallmarks of point-and-shoot photography: blocked shadows and blown highlights.

You Cannot “Fix this in Photoshop”!

This is just bad photography that comes from not understanding the tools. Whether it is a negative or a digital sensor, if you don’t record the information correctly, you cannot ever retrieve it. We’ve had the Zone system for about 100 years, it’s time to learn how to use it!

And, by the way, it shows exactly why HDR came about: because there is such discrepancy in the contrast between the light and dark areas in an image, you have to choose whether you want to capture the shadow details, or the highlights. The camera cannot do both in one capture. (I used to have this problem all the time with my Rottweilers.)  In HDR you take the first picture exposing for the shadows, and the second exposing for the highlights and then layer them in your favorite post-processing program for a composite image that has both.

NOTE: Typically people would say you “underexpose, take a ‘correct’ exposure, and over expose.”  But what you are really doing is metering on the highlights (since these are brighter you need to close down the f-stop or they will blow out), metering on middle gray (trusting the camera has it figured out), and metering on the shadows to make sure you get the correct detail (opening up the f-stop to let in extra light so they do not come out black). You can choose how much to open or close the lens by how much contrast is in the scene. Generally this is done by using aperture bracketing (f16, f11, f8), which I’ve already suggested should be the default setting on your digital camera. But you can use shutter speed (1 sec, 1/2 sec, 1/4 sec). Heck you could use ASA on a digital camera (400, 200, 100)

Some people take 3 photos 1 f-stop apart, some take 5 photos one f-stop apart. So lets say a digital camera has a dynamic range of 5 f-stops, shutter speeds, or zones. These are all equivalent because each time you open up the aperture by one stop, or double the time the lens is open, you double the amount of light, or change zones by one.

You meter on a scene and it is more than five zones, so you take one picture where the camera meters, one picture by opening up the f-stop by one and one picture by closing it down by one. Now you have covered seven zones in three pictures so you have both the highlights and shadow detail.

(Of course, I’ve been doing similar things on film for years. For instance when we were in Alaska on Prince William sound I was trying to capture the sunset over the ridge across the water. The clouds were luminous, but the backlit islands were blobs.  So I set my camera up in the afternoon, metered on the island, took one half of the exposure I metered on (one less f-stop, or in this case 1/2 the time), waited for the sun to light up the sky in the evening, metered, and double exposed the print at one half of the exposure again. This way I had a shot that was both forelit and backlit, the same way you might use fill flash to lower the contrast of a backlit subject.  By taking 1/2 of the exposure each time, I made sure that I did not over-expose the film.)

Seattle at night single exposure

Seattle at night double exposure

I often also do the same thing with city scenes, taking the image in the daylight while I have plenty of detail, and retaking it at night with the lights on. This way you can get the exposure without the lights in the scene blowing out the image, like at the top of the Needle.  (Sorry about the hairy scan.) With film you layer “in-camera,” with digital, it’s post processing. The idea is as old as photography.

Here is an example from  fly-fishing guide and photographer Mike Kinney. He does a lot of HDR and was very kind in providing me with examples for this blog.

Mike Kinney Links
Mike Kenney Fly Fishing

Image of the Peck hole exposed for shadow detail

And looking at its histogram. You can clearly see that the bottom of the histogram (Zones 1,2,3) have a lot of information. You can also see that at the upper end Zones 9 and 10 do too, with Zone 10 being “off the chart.” The shadow details were captured quite adequately to the detriment of the highlights in the sky, which are completely blown out.

But in a clear case of previsualization, Mike knew there was more to this scene, which frankly looks pretty pedestrian at the moment. So he took the picture with one stop less exposure to get some of the highlights. In his case he used film speed using f11 and 1/25, 1/50, and 1/100th of a second for the three exposures.

Peck at the “Corect” exposure as metered by the camera

And finally, one with the sky metered for its information (1/100s).

Metering on the sky

Two stops more  closed, the sky still has some issues and the foreground shadow detail is nearly gone.  So we know that the original shot encompassed at least seven zones, if his camera has a dynamic range of 5 zones.  I wouldn’t be happy with any of these images. Yet combined and turned into B&W…

Peck Hole HDR B&W

Peck Hole HDR B&W

And we have a very supra-natural shot where the strong composition of the image is apparent in a way that never was in color, while the subtle palette converts perfectly. I particularly like the specular effects in the trees. Look how the eye is drawn to the center of the image! This is one of those shots that makes you go “I wish I was there.” But you can’t be “there” because “there” only existed in Mike’s mind. It’s his artistic impression of that time and place that moves you. This is the endearing aspect of B&W, and I want to be very clear about it: the Zone System is all about post processing. Digital workflow  is just a new tool for the same old job.

Some people hate HDR, some people love it. For some it’s and end in itself – lurid, for others it’s a tool so subtly applied you don’t know it’s there. In the shot I posted last week that Bernard took of me:

Secret spot, photo by Bernard Hymmen, bernardsphotos.com

This image is highly post-processed. Not only is it stitched together from several shots to make the panoramic, but he used HDR to balance out the lighting in the stitched images (see Comments, below for more details). Yet this image is not at all unrealistic and does indeed capture the situation the way I remember it. (Minus the big fish I will tell everybody I caught there. ) There is nothing that strikes you as unnatural about this, but this scene couldn’t even have been caught realistically without HDR. So often, the image fails to live up to the memory because we see in HDR.

Hey, Jon, I Thought We Were Going To Talk About B&W Film HDR?

Last week I mused about how my “backup” digital photos from my last Elwha trip, converted incredibly poorly to B&W.  I was a little surprised about this. But in thinking about it a little more, I realized a lot of it has to do with my personal style for B&W prints, which is pretty contrasty and would generally exceed five zones of dynamic range.

For instance,  just taking Mike’s “straight shot” and converting it to grayscale:

Straight conversion
Straight conversion

I am unmoved by this shot.  No wonder my conversions didn’t move me. When I shoot B&W film, I generally shoot HDR!

That’s because one advantage, the defining advantage, that film still has over digital is that it’s dynamic range is many times greater than digital. If  this image was on a negative, I would know I could burn in the sky and bring out detail similar to what Mike did, see below, but a single digital image simply wouldn’t contain the information I needed.

Bruce Barnbaum maintains you can record 18 zones on B&W film (color slide film is about the same as digital). That is 2 to the 13th times more information than a digital sensor that can hold five zones,  or about 16,000x the information. In fact analog B&W is so superior to digital, that many photographers shoot film , and then convert them to digital to post-process or print. For instance Nick Brandt, on my blogroll to the right.

Wait just a darn minute, you say. How can you record 18 zones when Zone 10 is pure white? Great question. Paper can only hold ten zones, they eye can perceive more, and B&W film can hold much of what the eye can perceive. It is constantly balancing out a scene. So the trick is to capture all of that information, and then somehow compress the tonal scale to fit on the paper’s scale.  This has lead to the famous maxim “Shoot for shadows, develop for highlights.”

HDR B&W Film Example

Let’s take an example. The Zone System basically has three parts:

  • Expose correctly
  • Develop correctly
  • Print correctly

Where “correctly” means to get the image you want or “previsualize.” In a scene with too few or two many zones, you can use chemistry to make the negatives’ overall contrast greater or lesser, and then you have various local controls you can use when you print the image.

Snoqualmie, Twin Falls

Snoqualmie, Twin Falls

I actually took this image to explain the Zone System to my sister who was standing next to me. It is an interesting image, with a lot of problems you might typically find in an image. The sun had just broken over the ridge and was reflecting off of the top of the falls, and there was this cave on the bottom right that was full of flotsam. I had metered on the cave and placed it in Zone 5. That put the white of the falls in Zone 14, a 9 zone difference.  When the sun popped out it shown right at the top of the falls, putting it in Zone 18, for a total dynamic range of 13 zones.  (Had it shown directly on the falls, I would not have even taken the picture.)  In addition to the range, I wanted to make sure that I had definition in the water, not just a big white stripe down the middle of the image, and the wind was blowing. So I needed a pretty fast shutter speed.

I had to make some decisions. I decided I would let the top of the falls blow out and become pure white so that I could get the detail in the cave. I think the exposure was about 1/15th of a second. You can see the branches which overlap the falls are slightly in motion. I can live with these things. This was back before Kodak reformatted their 400ASA TMAX film to have the same grain as the 100ASA. Now I shoot the 400 exclusively to have more options  in these situations.

Traditionally, I would’ve shortened my development time to bring the 9-14 zoned highlights in a bit, then used local controls in printing. I still have not had much luck with changing development times and TMAX has so damn much latitude, I’m not sure what I would gain from it. I only even started metering my highlights recently, before I entirely ignored where they landed.

Thinking Negatively to Control Your Contrast

You have to think negatively to understand contrast controls a bit. The whitest parts of the print are the darkest parts of the negative, requiring the longest times under the enlarger.  So, the dark cave details pop up right away and will continue to get darker and darker (as they are in the lightest part of the negative allowing more and more light onto the print). The highlights can take 2,3,4, 20x as long to come in. (I have one print where various areas of the print get between 30 seconds and 30 minutes of exposure.)

To help the highlights come in, I pre-exposed the paper to light, so that any additional light would add detail. Then I put my hand over the cave to withhold light (dodged) . Finally I used a piece of paper with a small hole in it to add extra light (burned) just at the top of the falls to reduce the glare area much more than it is in the negative scan here.  Because I had previsualized it, I got the final print in just three tries, a major milestone for me. In the final print there is tremendous definition in the water, you can count the logs in the cave, and the big rock center right glows like a ’49 Buick grill.

'49 Buick grill, the "Million Dollar Smile"

’49 Buick grill, the “Million Dollar Smile”

What if I had done this digitally? Well assuming I could capture 5 zones on my camera, with even the 9 zones I would have to have had my base exposure, one two stops more, and one two stops less. Some people faced with such a range would actually take five, or more, exposures – one for each f-stop. But then the water would not have been frozen, the branches would’ve moved, and I still would not have had any information at the top of the falls to burn it in.

In a master’s hands, like Mike an Bernard, HDR is a great tool in the digital workflow, and maybe even a required one. But in film, HDR is inherent in the image itself.  Here is one of my most successful images, a straight scan of a 10 zone image (it has pure blacks and pure whites).

Marymere Falls

Marymere Falls

When It Comes to Digital vs Film, It’s Not All Black and White.

It’s the best negative I ever made, and I don’t think I could’ve done it digitally. I’m sorry, but for me, film just kicks ass for B&W HDR and that is the defining reason I still shoot it.

P.S. Who Are You and Where are You From?

This post gets a couple of hits every day. I want to hear from people who read it!

Who is visiting me from the Ukraine?

WordPress gives me these awesome reports and I can see how people find me, where they live, how they found the blog, what pages they visited. I see people from Europe, Asia, South America, and I always wonder who they are. I would love to get comments from  you!