120studio.com - updated 9/2014
(If you want to cut to the chase, read "The Test" below.)
article is not about saying one medium is "better" than another,
generally, but here's my reasoning. Given the incessant stream of
"film is dead" articles that seem to keep clogging the Internet, there
has to be some counterbalance. Rumors and hype are jeopardizing
the possibility that future generations will be able to enjoy film
photography. That would be an art tragedy of epic proportions.
Digital cameras have some strong points... but they also have at least one, major technical
flaw. This flaw represents a huge step backward in technology.
Despite all the continuing hype about the "death of film", the fact remains that film pictures
better in some important ways. I think others realize it,
too; the photo person at a local department store tells me
they're selling their film almost as fast as they can re-stock the
shelves. I'm also watching the prices on used film cameras slowly increase.
When I've taken pictures of people, they've seen the pictures and said "how did you get them to look
so good?" I think it's that they've spent so long gazing at cold,
flat digital pictures that they've forgotten what was missing. When they see it, it's kind of a
Anyway, I wanted to do an easy, factual test that would underscore the
biggest limitation of digital. If you want to see what that
scroll down and look at the pictures.
"Dynamic Highlight Range"
is the range above correctly-exposed neutral gray. In practical
terms it's the maximum difference in f-stops between the
areas and the brightest areas that aren't blown out (clipped).
If the picture's brightest area is five actual stops above neutral, but
the highlights start
clipping at only three stops, then the DHR is only three stops or less.
The traditional test of DHR involves something called a Stouffer
Wedge. It's basically a series of gray scale tones. These
are useful, but I
wanted to do a test that's more like real life. You don't go
around photographing grayscale tones or Stouffer Wedges in real
right? Another thing about those kinds of tests is
they can give better results than the camera can actually do in real
That's where "Camera versus Log" comes in.
There's a log sitting on a piece of metal.
The log is in the sun. It's also very light-colored.
The metal is in the shade. It's also dark-colored.
Using my camera's light meter, I check the exposure
difference. Today it reads five full stops between the metal and
the log. Digital camera or film camera: same reading.
If I meter for a correct shot of the corrugated metal, the log will be five stops overexposed.
But let's go even better. Forget the corrugated metal. I mean, let's not use it as the reference
point. Instead let's make sure the test is reliable no matter how
bright the sun is on a particular day. So, let's just meter on the brightest
area of the log. Whatever reading we get, we'll set the camera to
overexpose that by five stops. Then it won't matter if we
have hazy sun, noonday sun, morning sun, or what have you.
(By the way, my test setup is pretty consistent... the shaded metal and the
white log are usually 5 stops apart).
Two different cameras, two different light meters... same reading. Five stops overexposed.
That should be no problem, right? I mean,
everyone says digital
cameras have tons and tons of range, right?
No, wait. Five is kind of much. How about
this: Let's make it even easier for the digital. Let's
only four stops of highlight range. Now it should be really easy, shouldn't it?
The light meter says f/4 at
1/2000th on the log.
If I wanted five stops overexposure, I'd set the camera to f/4 at
I wanted four stops, so I set the camera on f/4 at
It's not the absolute settings that are important,
it's the difference. Whatever the brightest area would meter on a
particular day, I would choose four stops over.
I did the first run with a Canon digital camera of recent make.
Canons are known for excellent image quality, some say the
best in the industry. This camera is no exception. It got excellent reviews
for picture quality everywhere I looked. I say this because I want you to understand
it's no slouch compared to other digitals. In other words I didn't use a $49 bubble-pack special.
So I tested the Canon digital at +4 stops, and here's what happened:
Oooh, not good. The corrugated metal is about right, but the log detail is pretty much gone.
The awful detail loss isn't from messing with the contrast. You might
have been hoping that, but no. (I would hope that, too.) The
highlights were gone when the picture was taken. This is
hard shoulder, the biggest flaw in digital cameras. Detail looks
good up to a certain point, then it suddenly falls off a cliff.
When you pay all that money for a good digital camera, and it gets defeated by a log... that just ain't cool. Not even a real log; a picture of a log.
Say it along with me: My digital camera got defeated by a picture of a log.
At +5 stops, the whited-out area was even bigger and nastier than you see
here. Even a dime-sized area of that would be too much as far as I'm concerned. If
the area was like Texas at five stops and Nebraska at four, I'm
thinking at three stops it will still still be bigger than dime-sized. (And it was, as I found out.)
So, that was the Canon.
Nikon owners, let's don't be too smug just yet. The D7000 digital camera fared just as badly in the test. Take a look at how this DSLR performed:
That was the mighty D7000... defeated by a log.
The brightest area of the log metered f/4 at a
1000th. I set the camera for f/4 at a 60th. You can see the
I have even worse news. I tried the D7000 at +3 stops and it was still awful. In
fact-- and I promise I'm not trying to make you mad here-- the picture
was unacceptable at +2 stops. Two stops! I've read
Stouffer wedge tests that said it should do 3.5 to 3.8 stops of highlight.
Well, Camera vs. Log showed it wasn't even two stops, and probably not that much more than one stop. That's real life, not a Stouffer wedge.
There it is. A mere two stops, and the picture is
unacceptable. The corrugated metal is looking pretty dark,
because now it's a couple stops underexposed.
Two stops. The D7000 got trounced by a log.
No, not even a real log: a picture of a
log. Say it together with me. "My digital camera got defeated by a picture of a log."
Look, buddy, I'm not too thrilled about it either. I have a digital
camera, too, you know.
The D7000 results became acceptable when the log was only one stop above
the correct exposure setting, but by that time, the corrugated metal
was starting to get lost in shadow. At +0 stops (correctly
exposed), the log details were all there (you'd kind of hope so), but
the metal was almost blacked out. I might post all the
pics later on a separate page, but for now, we have to move on.
Round 3 of Camera vs. Log is coming up, and you don't want to miss this
Someone is surely going to say: "A bright white log in the sun and some metal in the
shade are probably more than 4 or 5 stops apart." First of all,
that's not what the meters said (5 stops range). Second, that
wouldn't matter anyway. The corrugated metal is a secondary
reference only. Remember, the test is really based on the brightest area of that log.
Whatever reading you get, set the camera to overexpose by four
stops. Or three, if you want. Or two. The shaded
metal is there to provide the darker end of the range, just for reference.
Besides, we haven't reviewed the results for the film camera yet.
Now Compare With Film
The next photos were taken with Fuji Superia 100 print
film. It may not look it, but this was hazy sun. The log was actually pretty bright to look at. There was some noticeable shadow boundary, but overexposure has washed it away:
+4 stops, and the highlight details are all there. Once again this is Superia 100 film.
I know. Some of you may be thinking "Not fair, the sun
wasn't bright enough!" But no: whatever reading
I got on the brightest area of the log, I set the camera four stops
over. Once again, the log in the picture above is four stops
overexposed on the highlights.
You thought I was just going to throw you a softball, there?
Nope, just to make sure there's no ambiguity, let's redo the test in
afternoon sun on a cloudless day, with the harshest shadow possible...
Four stops of highlight overexposure. Again. I used an
all-manual camera just to make sure. The highlights are still
closely and compare to the photo before it. Even in the
brightest white areas, the details are still there. That's film for you.
Oh, but wait, there's more. I wasn't content with four or five
stops of highlight overexposure. I wanted to see what film could
really do. So I took a photo with SEVEN STOPS
overexposure. In the sun on the whitest part of the log, the correct setting should have been f/5.6 at
1/2000th... so I set the camera at f/2.8 at 1/60th. (Whoa,
is he really gonna do this?) When I was taking the picture,
I even thought "there's no way this one is going to come out."
The highlights are still mostly there at +7 stops. The parts of the log that are just starting to go... are gently washing out. The detail didn't fall off a cliff the way it would have with digital.
That's seven full stops of highlight overexposure, people.
Just remember: the digital cameras were out of the running at two stops.
If someone were to come out with an improved dynamic range
camera (like the Fuji S5, which many did buy precisely because of that), it would have to be solidly in the running at a minimum of +4 to +5 stops to
make me want to use it much. I'm not seeing that. I don't have to go out of my
way to find examples with this kind of range; the sunlight on the
side of someone's face could easily be that many stops above their
The film retains nearly all of the tone information even at +7
stops. With some tweaking in Photoshop, you could probably
enhance the 7-stop picture to look a bit better.
And another thing... the film photos might be of a boring old log on a
piece of corrugated metal, but they're actually pleasant to look
at. Even the badly overexposed one isn't that bad. The
four-stop picture in the bright sun is actually kind of pretty (ignore
the lettering for a moment).
The digital photos, on the other
hand, look like boring, harsh, ugly pictures of junk. See how tonal range
and color can influence how much you like a picture?
The results of Camera vs. Log are so unfavorable to digital
that I know some people are going to claim the test was rigged. Nope. Or, they'll say
it's such an unusual setup that it doesn't matter. Wrong
again. Four stops
of highlight range is very common in real life.
In every photo, I checked and double-checked with spot metering.
If I wanted to rub it in, I'd mention that this isn't even fresh
film. It expired like four years ago. (Buy new film,
though, because it encourages them to make more!).
I want you to understand that I was just as hopeful to see digital
succeed here, especially the D7000. (Mainly because the person I
borrowed it from is eventually going to read this article.)
But I'm going to tell it like
it is, and here's the way it is: film has much better range than digital. Especially highlight range. (And maybe you haven't even seen this yet. )
This is just one reason why I still use film in 2014. It's why I
switched back almost 100% to film after using digital for a
decade. Let's stop trying to be politically correct.
Film is better in some ways. Digital has its own set of
advantages, too; it's just that dynamic range is
not one of them. There's no reason you can't use both
media. Use digital for quick and easy pictures, but when you want
incredible tonal range and richness, reach for the film.
See? Different tools for different purposes.
Unless their brains fell clean
out, I can't see the manufacturers abandoning film. But seriously: let's turn off the hype
machine. Can we stop acting as though it's
the conclusive end of film? There's just no need.
I for one could do without a constant stream of Associated Press
articles that keep trumpeting "the impending death of film". It's
as if Ford owners got together and decided to ensure that nobody could
ever drive a Chevy again, let alone even buy fuel for it. I find
that wholly unacceptable.
If the digital camera industry's product is so inferior that they feel
the need to keep floating "hit pieces", then what they really should be doing is
spending their energies making better products.
Besides, digital can't even begin to replace medium and large format
Some More Thoughts
Look, I'm not trying to make anyone feel bad. What I am
trying to do is challenge those who would dismiss film photographers as
"stubborn holdouts" and "hangers-on". Those kinds of
characterizations are unhelpful, ignorant, and irresponsible.
media intend it or not, their articles have come to resemble
thinly-veiled "advertorials" hatched in the back rooms of digital
camera companies. They're contributing to this strange
"good riddance" attitude. It's almost as if journalism has been refined to
the point of communicating on non-obvious levels, as if someone were deliberately trying to sway readers. But I would
never suggest that. (wink)
If we're going to say "good riddance"
to the old, let's at least come in with a better technology.
That's not what's happening. Show me a camera with the dynamic
range of film (and lose the creepy data-mining and face-recognition
technologies) and maybe I'll consider it. Oh, and it has to cost under $500.
The digicams I used in this test were examples of what's out there. I don't
care if it's a D3, a D90, a Coolpix, a Digital Rebel, or what...
they're going to fail badly in highlight range. I chose the
D7000 and the SX120 because they are two of the better cameras in
respective class, in terms of image quality. One is an enhanced
point-and-shoot with some pro features; and
the other represents the latest and greatest of Nikon's prosumer
Digital cameras are great tools for certain jobs. I used the SX120 to take this gorgeous photo of a day lily:
This is What Sells Digital Cameras
camera, same basic settings. The sharpness is
amazing. Once you get beyond about four to six megapixels, it's
not a function of sensor resolution anymore. The sharpness is a
function of computer-designed aspheric lenses.
Even in the lily picture, there are some small highlight
are starting to get clipped, but they're hard to spot. If I
don't point them out, most would never even notice them.
See, if you're very careful in your choice of
subject and background, digital can make some great pictures. I
just prefer not to work within the severe limitations. A
14-megapixel camera is not very useful to me if it can't even do four
stops of highlight range.
Hey, Wait a Second! Why Not Just Underexpose With Digital?
It's been said that digital cameras have a lot more shadow range
highlight range. I've decided to update this article to
address that in a bit more detail. Yes, you can underexpose
everything by a couple stops and then "bring it up" in Photoshop or
some other software. Problem is, for that to work really well,
you have to shoot in RAW mode, which eats up memory cards fast, and
also requires proprietary (expensive) software to handle the
files. The drastic shifts required to "bring up" a dark
image tend to cause really bad data gaps in a JPG image. The more
tweaking you have to do, the worse it's going to look.
Posterization, or "tone banding", can be one consequence of this.
Worse, even if you use RAW or 48-bit TIFF, the end result still doesn't
quite have the nice tones that a film picture would have had from the
start. Even after all the work done on it, there's still
something not quite the same about the digital picture.
When a computer simulation claims to emulate the "look" of film, you're
relying on the assumption that the programmers know exactly what it is
that makes that "look". Beyond that, you're relying on the
implementation. Thus far I've never seen one that was quite
right. It can look great, but it doesn't look exactly like
film. If you're doing journalism or something, this might not
matter, but in art photography the subtleties can be everything.
As I've said before, the chemical medium of film yields a continuum of
tones. That's very difficult to emulate with a digital
system. Each improvement may be a finer approximation, but at the
end of the day, it's still an approximation. If you remember
anything at all about limits and integrals from calculus, envision the
area under a curve. Now picture that area filled up with
rectangular sections, made as thin as possible. No matter how
thin we make them, they still have a stair-step appearance if you look
closely. Digital photography is built on that.
Here's another problem. Yes, you can underexpose the pictures, but
if you don't underexpose quite enough, the picture still looks
nasty. I have seen many otherwise-excellent digital photos, but
they had the telltale "digital blowout" in some small but noticeable
areas of the picture. Once that nasty-looking region of blow-out
is in the picture, it's there and it can't be recovered. To me,
that's still a deal-breaker.
and let's throw another reason on there: portraits!
Underexposing portraits is a good way to lose customers, because it
brings out every wrinkle and line. Using curve adjustments will
not make the photo look right. With film, you can shoot regular
or even high-key without that harsh clipping.
It's 2014 and when I shoot digital, I'm still using a cheap one for most things. That's because in late
2014, even the expensive ones still blow out highlights in a nasty, computerized,
unacceptable-looking way. Full-frame digital gives slightly more leeway, but it's still not slide film.
At least when highlights get blown out with film, they look
natural. As we've seen, Fuji Superia negative film was able to
tolerate seven full stops of highlight overexposure
and still yield a pleasant, smooth-looking picture. It
could be ten or twenty stops overexposed, and the worst that would
happen is that it would yield a smooth-edged transition to pure
Digital failed at just two stops, yielding a nasty, ugly, unnatural appearance.
Even slide film,
which many claim to be similar to digital,
still looks smoother when there are overblown
highlights. Slide film also tends to deal with in-frame sun a lot
better than digital does. You can even leave it stopped open enough that
you're not seeing a disc. Here's a photo I made with Elite Chrome 100 slide film:
Look how well the slide film handles the bright areas near the sun (and the sun, too!)
How many stops above that snowy field do you think they are?
Hint: a lot.
Even the sun shining directly into the lens looks natural
It doesn't have that harsh look of digital.
Kodak Elite Chrome 100, once again.
Sunsets are a great way to test the response to lighting extremes, so
lately I've been looking at a lot of sunsets from the best digital cameras.
To me they just don't have this natural, warm look that I get from
film. Digicams have a tendency to do one of two
things: nasty, hard-edged clipping, or smoothing algorithms that
turn the sun into a freakishly big hazy area. Slide film has a
narrower dynamic range than negative film, but in a "digital vs. slide film"
shootout, I choose slide film without hesitation. Make that
120 slide film (they're not really "slides" anymore, but whatever) and
it's a no-brainer.
Nobody's telling you to stop using digital if you get enjoyment out of
it. While we're on the subject, though: maybe journalists
should quit trying to tell me to stop using film. Really, why not pick
up a film camera and use both? (You can pick up a used Pentax
K1000, the classic all-manual 35mm camera, through this link, or browse the 'bay for 35mm cameras through this link.)
Even if and when
digital cameras have the bugs worked out of them, digital will still be
digital and film will still be film. To me, that counts for
If you haven't tried film photography yet, I encourage you.
I hope you enjoyed this article. Thanks for visiting this site!
P.S. If you pick up your film, camera gear, or anything else through the links, it helps me keep this site going. Thanks!
o.t o .1 2 0 s t u d i o.. c o m
This won't directly copy and paste. Please manually type it into your mail program.
No spaces between
Article & Gallery Index
Articles and photos are Copyright 2010-2013. Copying or distribution for any purpose is prohibited without express written permission.