Category Archives: Discussion

I want a camera that takes better pictures!

I hear this, or sentiments like it a lot:  “I want a new camera so that my pictures will be better.”

I’m going to tell you a very special secret.  In fact, it’s such a big secret that even professional photographers, camera reviewers, and photography journalists don’t know about it.  Don’t ask me how I learned about it, and don’t tell anyone I told you because if they find out I’ll be in big trouble.  OK?  Promise?

OK, here’s the deal.  Every single camera ever manufactured contains a secret ingredient.  The name of that ingredient is a closely-guarded secret, and even I don’t know it, so let’s call it Magic Crystals.  Magic Crystals were first discovered by Ansel Adams when he was a young photographer trying to take photos inside of a cave in Yosemite.  Young Ansel quickly realized the potential of these crystals and added them to his camera, and that’s how he became successful.

You see, Magic Crystals are the secret ingredient in every good photograph.  Nobody really knows quite how Magic Crystals work– science has tried to understand them, but they’ve come up blank.  The only thing that we know is that Magic Crystals make good photographs.   Big expensive cameras contain more Magic Crystals than small, cheap ones, and so expensive cameras take better pictures than cheap ones.  The cheapest cameras can only make out vague shapes and colors, much like when you wake up without your contacts.  Expensive cameras have so many more Magic Crystals that they take great photos.  At the far extreme, there’s a Hasselblad dSLR that has so many Magic Crystals that it takes great pictures without you even having to touch it– you just think about the picture and the camera does the rest.  Before you whip out your credit card you should know that it’ll set you back about $30K.

OK, you in the back.  I can see you looking skeptical.  You don’t believe that Magic Crystals really exist.  You think I’m just making this all up.  You know what?  You’re wrong.  There really is a $30K Hasselblad… and if anyone wants to send me one I’d be happy to put it to good use.  The stuff about the Magic Crystals?  Yeah, you’re right.  I was making it up.

The truth is that the camera body is one of the smallest factors when taking good pictures.  Every dSLR ever made  has the capacity to take stunning, award-winning photos.  Most people think that upgrading their camera body will make their photos better, and they’re shocked and disappointed when I tell them that’s not true.  Either that, or they spend a lot of money upgrading their camera only to be disappointed when the results look just like they did with the old body.

In my not-as-humble-as-it-should-be opinion, there are three primary factors that determine the quality of photographic results:

  1. The skill of the photographer
  2. The quality of the lens
  3. The camera body

Those are listed in decreasing order.

By far, the skill of the photographer is the most important factor in creating quality images.  Imagine finding the best, most expensive camera you can imagine, and handing it to a beginner.  What do you think the results would be?  Mmmmm hmmm.  You’d get exactly the sorts of crappy snapshots that you would grow to hate if you worked in the photo lab of your neighborhood pharmacy.  They’d be badly-exposed, out of focus, abysmally composed, and completely uninteresting for anything other than horror value.

Similarly, if you were to round up a half dozen talented expert photographers and send them off with disposable cameras, you’d likely get some excellent and fascinating results.

Photographic skill comes in lots of different flavors.  First you have to be able to visualize the shot that you want to take, which requires skilled seeing and composition.  Once you’ve done that, you need to be able to translate that image into technical factors– aperture, shutter speed, etc.  And finally, you need to be able to use your camera to maximum effect, handling the intricacies of autofocus, metering modes, and the rest of your camera’s rich and complex feature set.

Which Lens On the equipment front,  improving the quality of your lenses is the best way to improve your image quality.  If you lens isn’t sharp, then no amount of skill will ever overcome that limitation.  Similarly, some lenses are better at rendering colors than others.  You probably won’t notice things like that without doing a side-by-side comparison, but once you do the difference is obvious.  You can click on the image at left to get a larger version, but you probably don’t need to in order to see that one of the lenses looks a lot worse than the other two.

See the muddy colors in the bottom picture?  No amount of skill will overcome that.

Lens quality often comes with a high price tag, but not always.  Of the two lenses that took pretty good shots in the comparison, one of them is the Canon 50mm f/1.8, also known as the nifty fifty.   I’ve raved about this lens before, and for good reason– it has excellent image quality and carries a price tag that’s right around a hundred bucks.

And finally, the camera body does have some bearing on image quality.  For example, some cameras are much better than other at shooting at high ISOs– older and less-expensive cameras will generally have more noise than newer and/or more expensive ones.  They may be better at resolving detail or rendering colors, though in my experience this difference is very slight and is generally hidden by other factors.

More importantly, a new camera may give you new tools that you can use to take better photos.  A spot meter can be a fantastic tool, if you know when and how to use it to best advantage.  Newer and more expensive cameras may have better autofocus mechanisms.  My 5D has a better layout of the controls, which means that I can change the aperture and shutter speed in manual mode more easily than I could with my old Rebel.    The bigger viewfinder on my 5D means that I can manually focus more effectively.    All of those things are tools that I can use to improve my photos, but they don’t help much on their own– I still have to have the skill to use them well.

Now you know that there are no Magic Crystals that will help you take great pictures– it mostly comes down to skill, and the quality of your lenses.  A new camera can help, but it won’t be a silver bullet.

I’ll talk more later about how to know when it’s time to upgrade, and how to choose a new camera when you decide it’s time.

Leave a comment

Filed under Discussion, Equipment

The Advantage of Practice

One of the things that I always encourage photographers to do is to experiment with new techniques and practice the things that they’ve learned.  You’ll see experiments and exercises scattered throughout this blog, and there’s a very good reason for that– doing something and seeing the results will always give you a better understanding of something than just reading about it.  When you practice, the techniques sink into your brain and they’ll be available in the future if you ever need them.

The Roman philosopher Seneca once said, “Luck is what happens when preparation meets opportunity.”  As far as I’m concerned, Seneca speaks the truth.  You never know when some interesting and unusual photographic opportunity will fall into your lap, and if you know what to do you’re more likely to get the shot.

This evening I was at a shoot that I do regularly– it’s dress rehearsal for a friend’s stage show.  I’ve done this shoot every couple of months for the past few years, and I set up the lighting for it myself, so the technicals of the shoot are easy for me.  I know from past experience what works best, so before the action starts I drop the cameras into manual mode (no surprise there), pick an ISO, aperture, and shutter speed that I know works for the situation, and I’m then free to think about more important things.

Tonight, however, the friend threw a spanner into the works.  “During the first set, we’re going to want to kill all of the lights for a couple of minutes and just use the glowsticks that we’ll have on the stage.” I had a brief moment of thinking that he’d lost his mind, but then I thought about it for a moment and realized that it was no problem.

Colors In Motion
You see, I’ve already done a lot of very similar work, like this image. This was done with LED glow poi, but I estimated that they would have a very similar characteristics, but that the poi were probably a little bit brighter. I knew from experience that I could get good but slightly-underexposed results with glow poi at f/4 and ISO 400, and whatever shutter speed was right to get the amount of motion I wanted.  (One of the interesting things about shooting moving lights is that longer shutter speeds don’t make the image brighter, but instead give you a longer motion trail.  Do you see why that is?)

I was shooting most of the show at f/4, ISO 1600, and 1/60 sec.   I estimated that f/4 and ISO 1600 was probably about right for the glowstick shots, and that I’d probably want exposures of about one second.  I also knew that autofocus wouldn’t work very well, and that I should pre-focus and keep the lens in manual focus mode.l

We discussed what the cues would be for cutting the lights, so I knew when it would be happen.  When I saw the first cue I changed the shutter speed to one second, switched to manual focus,  and waited.  We cut the lights, I got a few shots of pretty moving lights, and then we brought the lights back up and continued with the set.  When the lights were back on I changed the camera back to 1/60 and autofocus.

I got one good shot from that segment, which is about what I would have expected.  The exposure I picked was just about spot on, and because I was shooting in raw mode I had the latitude I needed to make final adjustments on the computer.


If I’d never practiced this sort of shot I would have had no idea what to do with the unfamiliar situation.  However, tonight practice and experimentation sereved me well– when an unusual opportunity arose, I could make my own luck and get the shot.

Now it’s your turn– go find an unfamiliar situation and play with it, so that the next time you run into something similar you know what to do.  What are you waiting for?


Filed under Discussion

DPI: When you DO care

In my last entry, I said that you usually don’t care about DPI.  You should definitely read that and understand it before you dive into this entry.

Now that I’ve told you why you don’t care about DPI, I’ll tell you the one time that you do– when you’re printing an image.  That’s the one time that DPI comes into play, and when you get to it you’ll really care.

For print resolution, there are really two things you care about– dots and inches.  Dots are easy– that’s the number of pixels you have in each direction of your image.  If your file starts out with a resolution of 3600×2400, you have 3600 pixels/dots on the long side, and 2400 pixels/dots on the short side.  That’s pretty obvious, right?

The second number is inches.  If you’re going to print your photo at 8×12, there are 12 inches on the long side and 8 inches on the short side.  I know, I know, but I’d really rather state the obvious and get it over with rather than having some poor guy in the back scratching his head.  Besides, if you’re not American you probably grew up with the metric system, and you think an inch is something that should be scratched.

Ahem, OK.

We were talking about DPI.  Right.  As I said earlier, DPI stands for Dots Per Inch, and it’s a measurement of how much to stretch the image out when it goes on a piece of paper.   Working with DPI requires some really simple math.  Duck, because I’m going to hit you with a formula.  Ready?

Pixels / DPI = Inches

That wasn’t so bad, was it?  That’s really the core of what you need to know about DPI.  Let’s walk through it.

Remember our 3600×2400 image?  Let’s figure out what happens if we print it at 300DPI.  For the long side, we have 3600 pixels.  If we divide that by 300DPI, our printed image will be 12 inches.  For the short side, 2400 pixels divied by 300DPI will be 8 inches.   See?  That wasn’t so hard.

Now, try it yourself.  What happens if you print it at 200DPI?  Go ahead.  I’ll wait.  Here’s a pretty picture… you can think of it as the equivalent of on-hold music.


You’re done?  Did you figure out that it would be 18×12 inches?  If so, give yourself a gold star.  You did the math correctly.  If not, let’s walk through it.  3600 pixels divied by 200DPI = 18 inches.  2400 pixels divided by 200DPI = 12 inches.

Most printers will have some native resolution that they want you to send them files in, often 240DPI or 300DPI, but check the doicumentation for the device you’re printing on.  If you just send the image to the printer at that resolution, you now know how to figure out how big the print will be.

Often, however, you want to go tthe other way– you have an image, and you want to print it out at a fixed size.  In order to do that, you just swap the equation around a little bit:

Pixels / Inches = DPI

It’s the same equation as before, only we’ve rearranged the terms ala Algebra 101.  You do remember Algebra 101, right?  If not, just trust me.

What this means is that if you have a 2400×3600 image and you want to print it out at 4×6, you can do a little bit more basic math.  In particular, 3600 pixels / 6 inches = 600DPI.  Since DPI is essentially always the same horizontally and vertically, it works out well that 2400 pixels / 4 inches = 600DPI.

(Note:  Everything I’m writing today assumes that the aspect ratio of your image is the same as the one that you want to print at.  If you have a rectangular photo and you want a square print, you’ll need to start by cropping or padding the iamge to the right aspect ratio.  That’s not a topic I want to cover today.)

So what happens if you have a 2400×3600 file, you want a 4×6 inch print, and your printer really insists that you send it 300DPI files?  In that case, you have to resize your image.

The mechanisms for doing this will vary greatly depending upon what image manipulation software you’re using, and resampling is an art unto itself.   One way to do this is to do the math yourself and figure out how many pixels you need (in this case, 6 inches * 300 DPI = 1800 pixels on the long side, and 4 inches * 300 DPI = 1200 pixels on the short) and then resize the image to those pixel dimensions.  In Photoshop you cantake the easy way out and let Photoshop do the math– resize the image, set the DPI and the inches on each side, and let it figure out how many pixels the resulting image has to be.  Other image manipulation software is probably similar, but I’m not familiar with the exact mechanisms.

Most consumer-grade printers have drivers that are pretty smart about resizing images on their own.  My Epson R1800 will let me send any size image to it, and I can change one setting to tell it to fit the image on whatever size paper I’ve chosen.  That’s simple!  You still want to understand the basics, though, so that you know if your image will turn out OK. The lower the DPI, the fuzzier your image will look when it’s printed.  200DPI or more should certainly give you acceptable images, and for very large prints you can go lower– people stand farther away from big prints, so they won’t notice if the print isn’t completely crisp and perfect.

If you’re resampling the image, you can generally make it smaller without losing a lot of image quality.  There’s a limit to how much you can make it bigger, though.  When you make an image bigger (called upsampling, among other things) the software has to guess at what information it should fill in.  Newer programs are pretty good at this, but they’ll never produce anything as good as if you’d captured more pixels to start with.

Does this all make sense?  I hope I’ve left you with a better understanding of DPI.

Leave a comment

Filed under Discussion, Equipment

DPI: Why You (Probably) Don’t Care

I see this question, or some variation on it, over and over again:

“I just got a new camera, and it shoots at 72DPI by default.  How do I change it?  Won’t 300DPI give me higher-quality images?”

It’s a common question, and a very common misunderstanding about DPI.

First, let’s start with the camera.  Your camera’s sensor has some native size, some fixed number of recording points (remember the sensor gnomes?) on its sensor.  Each recording point records one pixel of the image.  Every image that’s shot with the camera has that size no matter what.  Using the Canon Digital Rebel XTi as an example, the sensor is 3888 pixels wide and 2592 pixels tall.  If you multiply those numbers together you get 10,077,696, which is why it’s called a ten megapixel (ten million pixel) camera.

Remember that.  As long as you shoot at the XTi’s maximum resolution (and by that I don’t mean DPI, I mean L or L+RAW) you’ll get an image that is 3888 pixels wide and 2592 pixels tall.  Some cameras let you shoot smaller images, by changing the size from L to M or S.  If you do that, you’re shooting at whatever size that particular setting gives you, but it’s still independent of DPI.

72DPIFlower So what’s DPI?  It stands for Dots Per Inch. For our purposese, a dot is the same as a pixel, so it would be pixels per inch. (There is a technical difference between dots per inch and pixels per inch, but for purposes of this discussion we’ll treat one pixel as being the same as one dot– it’s one unique spot of data in our digital image)  DPI is not really a part of the image itself, and doesn’t change anything within the image.  Rather, it’s a number that’s stuck onto the side that tells a printer how big it should print out the image– imagine someone handing the file off to be printed, and sticking a post-it note on the file so that the printer will know how big to make the image.  If the image is 72DPI, the printer will print out 72 dots in every linear inch.  At 300DPI, the printer will print 300 dots in every linear  inch.

If you’re working with an image on the computer, and never printing the image, DPI is completely irrelevant.  A 3888×2592 image at 72DPI is exactly the same image as a 3888×2592 image at 300DPI or 1DPI, or a million DPI.  Really.

The thing about DPI is that you can change it arbitrarily, and it won’t affect the image.  As long as it stays 3888 pixels wide and 2592 pixels tall, you can rip off the 72DPI sticky note and change it to any other number, and you still have exactly the same data in your image.  That means that you don’t really care what DPI the camera sticks onto the side when you first take the picture– changing it in-camera won’t change the information that you capture, and it’s easy to change it later if you want to print the image out.  Plus, if you’re printing the image you probably will want to resize it, crop it, or make other adjustments.

And that, in a nutshell, is why you probably don’t care about DPI.  Unless you’re printing the image out, DPI is a completely irrelevant concept, and one you basically just shouldn’t worry about at all.

Next up, I’ll explain how DPI works when you’re printing.


Filed under Discussion, Equipment

Slightly Off-Topic: Photographers’ Rights

This isn’t really about how to take better pictures, but it’s important enough that I want to share.

Yesterday, while walking through San Francisco’s SOMA neighborhood, I spotted a sign in a bookstore window. The sign reminded me of something, so I pulled out my phone and took a picture of it. Just as I finished taking the picture, one of the store’s employees came up to the window and gestured to indicate that I wasn’t allowed to do that.

I didn’t push the issue, but he happened to be dead wrong. In the US, I have every right to photograph the exterior of a commercial building, so long as I am doing so from public property. If I’m on private property, the owners of the property have every right to ask me to refrain from taking photos, and I must do so.

Every photographer should read the following information on photographers’ rights.  It’s even provided as a printable PDF, so that you can keep a copy in your camera bag.  In our current social climate, photographers are being hassled more and more in the name of security, and it’s important to know your rights, responsibilities, and restrictions.


Filed under Discussion

Two interesting photos

I’m a big fan of the Day in Pictures on SFGate — the photos are generally quite interesting, they’re well-captioned, and the editor likes to play amusing games with proximity.

Today’s set had two photos that are excellent illustrations of the things you can do if you understand exposure. This one shows an elderly man praying in the middle of the road, with cars whizzing by on either side of him. The blur of the cars is worrisome– you’re sure he’s going to get hit at any moment– but it shows his conviction.

The second one shows a landscape with a closeup of poppies in the foreground. The poppies are crisp and sharp, and the background is reasonably sharp as well.

Take a moment to think about how you might shoot those two photos.

Got it? OK, here’s my take.

The first one requires a slow shutter speed in order to capture the motion of the cars. If we look closely, we can see that everything is sharp as far as the eye can see. This suggests that the photographer used a very narrow aperture and a slow shutter speed, and probably a low ISO as well. We don’t know how fast the cars are moving, but let’s guess, oh, 30MPH. We can ask Google and find out that’s 44 feet per second. I’d guess that the shutter speed was somewhere between 1/2 second and 1 second. I’d guess that the photographer also used a neutral density filter (something I haven’t talked about, but it’s basically a piece of grey glass that reduces the light coming into the camera) to get a slower shutter speed.

The poppies were clearly shot at a fairly small aperture, since the background was mostly in focus. The shutter speed is indeterminate, but was probably “whatever it takes”.

Whenever you see an interesting and unusual picture, take a minute to see if you can figure out what the photographer did to get it. This is excellent practice for figuring out your own shots when you’re trying to get a specific effect.

1 Comment

Filed under Discussion, Exposure

Why are there spots on my picture?

Dusty Flower You might notice it in the blue of the sky or the white of a wedding dress. Maybe, as often happens with me, it’s the petals of a flower. Whatever the image, someday you’re going to look closely at a picture and see that it’s full of little spots. These often look like little dark spots ringed by a color that’s slightly lighter than the background, and they show up in the same place on every picture you take.

These spots usually come from dust on your sensor. You won’t see them too often at large apertures, but when you stop the camera down they start to become visible. They particularly stand out in areas that have large expanses of a single light color, but you can often see them in other areas of the image as well.

This flower was taken at f/32, which as we all know is a very tiny f-stop. It’s not a great photo, but because of the small aperture it shows every little piece of dust on the sensor. You can’t really see it at the small size, but if you were to look at the full-resolution image you’d see lots and lots of little bitty spots all over the photo.

Sensor DustOn the right is a collection of full-size crops taken from the same image. These spots become highly-visible and quite annoying when the image is enlarged.

There’s an easy way to test to see if you have noise on your sensor. Find a brightly-lit white or light-colored wall, or a piece of plain white paper. Set your camera to manual mode, shutter speed of about two seconds, and the smallest aperture your lens will do. Point the camera at the white surface and take a picture, moving the camera around just a little bit so that the background becomes blurry.

Transfer the image to your computer and look at it full-size. Chances are good that you’ll see at least a few little specks, and if you’ve never cleaned your sensor before it may be pretty bad.

This image originally came from Flickr user djwudi, and I thank him for allowing me to use it. He works in a camera shop, and a customer came in complaining about spots on his photos. This is what the sensor looked like at f/32 before cleaning. I’ve personally never seen one this bad, but dust can start making your pictures ugly long before this point.

If you change lenses, and sometimes even if you don’t, getting dust on your sensor is inevitable. However, there are a few things you can do to reduce the amount of dust that your sensor collects:

  • Never store the camera without a lens or other protective cap
  • Minimize the number of times you change lenses, and minimize the amount of time the lens is off the camera
  • When you change lenses, make sure the body of the camera is pointed down and the camera is turned off
  • Try not to change lenses in dirty environments

My procedure for changing lenses is to turn the camera off (though sometimes I forget), then take the rear cap off of the new lens and get it ready to go. As soon as it is, I point the camera straight down, remove the lens, and put the new one on as quickly as possible. I then put the rear cap on the old lens and store it in my camera bag.

Do you have dust?  Don’t worry!  Here are SSA!’s instructions for cleaning your sensor.


Filed under Discussion, Equipment

Shooting Through a Fence

In the last lesson, I talked about shooting at the zoo under difficult circumstances and how to break the problem down into manageable pieces. One of the problems that TJ faced was that there was a fence between him and the subject. I promised to take a few sample shots through a fence and show what effects different apertures had on the image.

The following images were all shot with a Canon Digital Rebel XT and the Canon 50mm f/1.8 lens, also known as the nifty fifty. I chose it because I wanted to be able to open up the aperture in order to blur the fence as much as possible. Here are three sets of images that show how the fence looks from different distances at different apertures. I apologize for not finding more interesting subjects, but these were taken during a quick stop at the Port of Oakland when I was late for work.

In the first set, the camera is just a couple of inches from the fence. The camera was in aperture priority mode, and the four shots were taken at f/2, f/4, f/8, and f/16. As you can see, at f/2, the fence blurs so much that it’s essentially invisible. At f/4, you can start to see the fence blurring parts of the image. At f/8 the fence is quite visible, and at f/16 it’s almost sharp.

Close f/2 Close f/4
Close f/8 Close f/16

This time, I’m standing about an arms length away from the fence.  At f/2, the fence is just a blurry grid.  At f/4 it starts to look like a fence, at f/8 it’s clearly a fence, and at f/16 it’s fairly sharp.

Medium f/2 Medium f/4
Medium f/8 Medium f/16

In the final set I’m standing about five feet away from the fence, and also at a slight angle.  At f/2 the fence is a blurry mess, at f/4 it starts to sharpen up, and at f/8 and f/16 it’s sharp enough to work as part of the image rather than being a flaw.

Far f/2 Far f/4
Far f/8 Far f/16

Conclusion: if there’s a fence in your way, figure out what you want to do about it and adjust accordingly.  If you want to make the fence disappear, get as close to it as you can, and open up your aperture as wide as possible.  If you want to use the fence as part of the image, back up a bit and stop down.


Filed under Discussion, Lesson, Question

What do you want to know?

I need to go shoot some examples before I write the next couple of lessons.  Unfortunately, I’ve been spending painfully long hours in the office this week and haven’t had time to get out and shoot.

In the meantime, what do you want to know?  My plan is to fill out the basics of aperture and ISO, and then do… something.  What should I write after that?  Do you have questions about what I’ve written so far?


Filed under Discussion