May 2014

This is the first article in a two-part series on choosing a landscape camera.    This first article deals mostly with digital cameras, but in part two we'll also look at some good choices in a film camera for landscapes. 

Many cameras are said to be "good for landscapes", but what does that mean?  

There are probably many different answers you'll get from different photographers.  I'll cover one of the most basic ones-- resolution-- in this article.

Article Sections

Dynamic Range

  A.  Sensor Size
  B.  Detail Resolution

The Sweet Spot

More About Megapixels and Sensor Size!

Which Camera For Landscapes?


Any camera that's "good for landscapes" should have at least two qualities:

     1. Good dynamic range

     2. Good detail resolution.

Now, let's define "good".

Dynamic Range

I like slide film.  For artistic use I like it better than anything digital has to offer.  And yet, slide film is often said to have only about five or six stops of dynamic range (DR).  That doesn't sound like much, given that most digital cameras are claiming 11+ stops of DR.

Actually, slide film behaves as if it had considerably more DR than most digital cameras.  

Here's how I can know that for sure.  Let's say I'm macro capturing a slide with a DSLR that has 11.9 EV of claimed dynamic range.  And the DSLR capture is losing highlights that I know are good in the original slide.  If that occurs, then one of two things must be true.  Either:

1.  The DSLR doesn't really have as much DR as they say...


2.  The slide film actually has more DR than commonly thought.

Here's what probably is happening in the macro-capture example.  Glare and light-spillover can decrease the effective DR of a digital camera.  There might be other things happening, too.  Point is, even though a DSLR can get defeated by a picture of a log, that doesn't mean it's no good for landscape photos.  It just means you have to know the limitations and work within them.

I really like Canon DSLR's, but the benchmarks say they have even worse dynamic range than Nikon DSLR's.  My own experience shows the difference is not that huge.  But okay, let's say Canon DSLR's have less dynamic range.   Somehow, though, photographers manage to get pretty nice-looking pictures every day with Canons.   Again, it's about working within the limitations of the camera.

OK, enough about dynamic range for the time being.  Now I want to talk about the other important quality for a landscape camera.

That is,


What makes a digital camera have good resolution?  It's actually amazing how complex this subject is.  Even people with physics degrees argue about this endlessly on forums.   Sensor MTF, lens MTF, airy circles, diffraction limits, pixel pitch, quantum efficiency... it just goes on and on.  If you were to take the forums as any indication, the question is not settled and never will be. 

Hey, wait a minute.  If no one knows the answer, then how could the engineers design such good cameras and lenses?  Are they bickering about it too? 

Nope.  They know what they're doing, and they've already gone through the math.  Yes, it's extremely complex, but they've worked on optimizing the resolution within the design constraints (that's the kind of thing engineers are good at.)  They've already used the test equipment.  They've already used all the computer-aided design tools available to them.  So they've been squeezing the best performance they could from each new generation of camera sensor. 

That's great, but the rest of us want to know how to select a good camera for landscapes.  

A.  Sensor Size

You've already heard that megapixels are not all that important.   I still say that, because some of my favorite digital pictures were actually shot on low-megapixel cameras.

For portraits, you're actually better off having a bigger sensor than more megapixels.  A 6-megapixel APS-C camera is going to give better people photos than a 16-megapixel pocket camera with its teensy 1/2.3-inch sensor.  

All other things being equal, larger sensors make for nicer photos. 

But why? 

Smaller sensors mean smaller pixels.  This causes more crosstalk between pixels.  There's optical crosstalk and there's voltage crosstalk.  When pixels are small and close-together, basically you have both light and voltage spilling over into other pixels.  It's harder for the camera to sort the "good" signal from the "junk" signal.  The signal to noise ratio (SNR) degrades.  So, a 16-megapixel pocket digicam sensor (1/2.3") can't give as good a picture as a 16-megapixel APS-C sensor.   Processing technology can improve this, but if you keep all other things equal, a larger sensor gives a better-looking picture.

Here's how to minimize the crosstalk and image degradation.  Pick scenes that are well-lit but don't have extreme contrast or brightness differences.   Sounds easy, right?  Maybe.  Read my "film landscapes" article to learn more about that.

There are two conditions where small-sensor cameras stumble:   very bright light, and very dim light.   The bright, high-contrast scenes look extra-crunchy and harsh.  The highlights clip easily, and it looks really bad.  Meanwhile, the dimly-lit scenes get very noisy.  Either way, the image quality degrades.  To make matters even worse, most newbies like to crank the contrast way up because they think it looks more like slide film or something.  But it doesn't.  It looks like a bad Noritsu scan of cross-processed film, made by an operator who was in too much of a hurry.  (Send your film to a pro lab, or use this method to "scan" it.) 

Smaller sensors also hit the diffraction limit sooner.  A 1/2.3" sensor is already diffraction-limited at f/5.6.  Most zoom lenses can't go wider than f/5.6 at long focal lengths, and many of them are above f/6.  So right there, your small-sensor zoom camera is losing effective resolution.  The camera tries to compensate by doing various processing tricks.  These could include sharpening, but they'll never be as good as if you just started out with a larger sensor.

Here's another thing.  Smaller pixels do not generate as many electrons per pixel.  That's because they don't receive as many photons falling on each pixel.  Again, the signal-to-noise ratio degrades here. 

If you want nicer, mellower tones, go for a bigger sensor.  And up to a certain point, if you want clearer, more detailed images, also go for a bigger sensor.  Or, be very choosy about the scenes you photograph.

B.  Detail Resolution

Portraits don't need huge megapixel counts.  They do better with larger sensors.  

What about landscapes?  Well, it depends on what you're emphasizing.

Sunsets, clouds, & skies are about color and tones more than detail.  Megapixels don't matter much here.   You can even do OK with a smaller sensor if you know how to handle (or avoid) the extremes of brightness.

Just for fun... Tell me how many megapixels.

And tell me what brand or type of camera.

Email me with your guesses.

Detailed landforms (distant foliage, roof shingles, craggy rock faces, etc)  do better with higher megapixel counts.  And if you want the clearest detail rendition, it helps if those pixels are larger, which means you should go for a larger sensor if possible.

Below 12 megapixels, you'll notice detail dropping out of foliage.  Even at 12 MP, the distant trees can become green blobs when you look at them closely.  The magic number seems to be 16 megapixels.  That's just my own observation, no engineering math involved.   At 16 MP, the distant trees have a reasonable amount of foliage detail.  I notice a big difference in detail between this and the lower MP counts. 

A 6-megapixel camera can make great portraits and gorgeous sunset shots, but if your passion is photographing landscapes and preserving fine detail, then you'll probably want 16 MP or higher.  The difference will really start to become obvious in the bigger enlargements (starting around 16x20 inches).   At 40-plus inches along a side... well, see below.

I do landscapes for impression and color first, and detail a not-so-close second.  So I'm OK with a 12-megapixel Canon EOS Rebel T3.  In fact I love this camera.  Few people expect this cheap DSLR to have the performance.  Well, it may not have the raw megapixel count, but it happens to have a very good pixel pitch.  The pixel size is considerably bigger than the more expensive 7D and 60D cameras, which means better signal-to-noise ratio, etc.   So the Rebel T3 is actually a very capable camera.  It absolutely tromps all over the $4,300 DSLR's that people were gloating about in 2003.  

For someone else, who concentrates on fine details more, the starting point might be a Nikon D5100, with 16 megapixels, or a Canon Rebel T3i, with 18 MP.  Both are great all-around cameras.  Either could serve as your "just one digital camera".  We'll revisit these cameras, and their advantages, in just a bit.

But first, let's talk about megapixels some more.  Here's a comparison you might find interesting.   Study these two photographs carefully.

Photo 1.....

Now, here's Photo 2:

I sized them so the clouds & distant trees would look about the same size on-screen. 

Look at the sky.   The sky is all about colors, tones... overall impression.  It's not really about detail. 
And so, both pictures look pretty OK in the sky department.

Now, what about the foliage?   Look at the distant trees.  Look at the goldenrods.   What's different?

Aha.  Photo 1 lacks foliage detail. 

Let's make sure.  Here's a pair of 100% crops from the pictures.  These are what you'd see if these images were blown up to 44-inch posters and you were standing really close to them.  

First, here's a 100% crop from Photo 1:

Next up, here's a 100% crop from Photo 2:

Amazing difference.  So, which is which?

Photo 1.... Digital.  6 megapixels.  You can see there's a lot of missing detail in the trees, the foliage, everything.   

Photo 2...  Film.  35mm.  Look how detailed that is.  Every time I see the difference in resolution, it reminds me that we sold out our photographic heritage on the completely false premise that digital was a "superior" replacement for film.  People were saying stuff like this when cameras had 6 megapixels;  and yet, as you can see, an honest comparison makes the digital look like a bad joke.

I didn't even choose a pro film for this.  The consumer-grade 35mm film was already expired by about 5 years when this picture was taken.  Even more amazing, this is only from a 16-megapixel macro capture of the film.  A higher-res capture would probably yield more detail, since 35mm has anywhere from 18 to 24 megapixel equivalent (depending on film type and lens).  And by the way, I didn't use a tripod or mirror lock-up.  It's just an offhanded shot, probably at 1/60th or maybe 1/125th.

But as it stands, the 35mm film could make a 44-inch poster that would look pretty good.

I've often said that most people won't notice a big difference in detail unless you make big enlargements, but it doesn't take a pixel nitpicker to see that Photo 2 has more detail than Photo 1, even at 800 to 850 pixels wide on a web page.   So, megapixels don't matter for some things, and you can make really nice-looking enlargements of 4-megapixel images, but there really is a detail difference. 

At bigger enlargement sizes, 35mm film looks like 4x5 when compared to a 6-megapixel photo.

Ok, the point of this article wasn't to trash digital.  The point is to tell you how to avoid the situation we just saw.  But since we're on the subject... digital absolutely did not surpass film at 6 to 10 megapixels.  Not even at 12 MP, because even there, I can see the loss of detail.   In all likelihood, the 16-megapixel D5100 sensor is still acting as a resolution bottleneck for the 35mm color negative, and it's not even pro film.   This is Superia 100 that expired in 2007.  (Buy new film, though, and support real photography for future generations.)

Look at the 100% crops again.  And all this talk about lenses not being good enough for the sensor is just a distraction.  I could say the same thing about film... I often use cheap lenses on my film cameras, but they still tromp all over low-MP digital.  Every day of the week.  Now if someone were to say the Nikon D800 surpasses 35mm resolution, they'd have a point.  I'm about as "film die-hard" as it gets, but even I have to concede that the D800 meets or exceeds 35mm resolution (ha!  it took 'em long enough....).  This is why, if you can afford one, I would recommend a Nikon D800 for macro capture.  Then the camera won't be the limiting factor.

As for a 6 to 10 megapixel camera... this is less than ideal.    Forget those test patterns that everyone seems to use.  More realistic:  compare the foliage on distant trees.    I've found that 12 MP is just barely adequate for this, even with a very good lens.   A few extra megapixels beyond that, and the detail quality improves considerably.

I think what's happening is a lot of people are confusing acutance with resolution.   High-acutance, low-megapixel images look sharp and detailed, but when you look at them up close, the detail is actually not there.  (Refer again to that 100% crop of that 6-MP image).  Another thing that happened in these early comparisons was the scan bottleneck.  The digital guys were comparing crummy scans with photos they took on their 6-megapixel cameras and declaring the matter settled.  Scanning technology back then was not so great.  

It reminds me of a friend whose dad scanned all their family slides in like 2001 and then threw away the slides... ouch. 

Now let's talk about megapixels some more.

The Sweet Spot

I said before that a bigger sensor translates to better image quality.  

Well, sort of. 

Just because you build a camera with a large sensor, that doesn't mean it's automatically going to give nicer pictures than a small-sensor camera.  What if you gave it only 100 pixels?  (Not a hundred megapixels, just a hundred pixels.)   The images will look jaggy and pixelated.  Even down around 2 or 3 megapixels, something still doesn't feel right about it.

Each sensor type has a "sweet spot" for the number of megapixels vs. the sensor size.  It has to do with pixel density.  That is, how closely you pack the photosites together.  

I mentioned before that too many megapixels on a sensor will degrade the image quality.  Well, too few of them will also degrade it.

There's something called quantum efficiency (QE).  It's pretty technical stuff, but if you read through a lot of quantum efficiency data, you'll see that the highest QE for an APS-C sensor is somewhere between 14 and 16 MP.   And I've found that it's about 16 megapixels where 35mm film captures start to look really good.  What that also tells me is that 16 MP is a good starting point if you want landscapes with nice detail in the distant trees (etc.)

For an APS-C sensor, 16 MP gives probably the best balance between resolution and high ISO performance.  This puts the Nikon D5100 as one of the best DSLR's you could ever buy for all-around use (i.e., not just landscapes, but also indoor photos).  Get your D5100 through this link and it helps support my website.   You'll also need a macro lens.  Check out the macro capture article for some recommendations.

The great Fuji X-E1 / X-E2 / X-T1 cameras also have 16 megapixels.  Probably not a coincidence.  Something tells me the engineers at Fujifilm Corporation are quite aware of this.  Even now that 24-megapixel sensors are on the market, they chose 16 MP, because this is about the balance point between high resolution and good low-light performance.    (You will read that the X-series are not good for landscape colors.  This is a matter of personal preference.  I do see the difference, mainly in the greens.  Fuji renders green in a particular way, across all the cameras I've seen, even in the Finepix pocket cameras.  Some people like it, some don't.  I don't care, because it takes me about ten seconds to change the colors to my liking.

On a crop-sensor camera (APS-C), I wouldn't go lower than 12 megapixels for detailed landscapes.  As I said, 16 MP is preferable, because it's a good balance between detail preservation and low-light performance.  The Fuji X-series have no antialiasing filters, so they provide the sharpest, most detailed images that can be had at 16 megapixels. 

If you want more megapixels and no antialiasing filter, the 24 MP Nikon D3300 is the best deal around.  (I would rank the Nikon color palette a close second to Canon for landscapes, with Fuji and everyone else third.)  The D5300 and D7100, which have more features than the 3300, also have no antialiasing filters.  Great cameras, all of them, but because I like cheap cameras, I'd probably go for the 3300.

For a full-frame camera, it seems to me the images really start to look spectacular around 18 to 20 megapixels.  For daylight situations, you can pack 30 MP or more onto a full-frame sensor and the image quality improves even more.  Except in low-light situations.

Say, why not stuff 36 megapixels onto a crop-sensor?   Or, why not cram 60 MP on a full-frame sensor? 

Well, when the pixels get too small, the image quality starts to go downhill again.  Even in daylight.  Like I said, there's an optimal range of how many pixels should be on a camera sensor.  Exactly how many that is will depend on a lot of things.  Voltages, materials, image processing technology... probably a lot of stuff.    The newer Canon 5DS has proven that 50 MP is entirely feasible on a full-frame sensor;  it seems that improvements in technology have somewhat offset the drawbacks of pixel crowding.

More About Megapixels and Sensor Size!

So, let's suppose you have two different sensors, each with the same number of megapixels.  Sensor A is larger than Sensor B. 

That means Sensor A has larger pixels.

The thing is, if you select the right scenes and use the right camera settings, you'll probably never be able to see a difference. 

How big was the sensor that made this photo?   

Blue and White Sky

Mother's Day 2014
Camera:  you tell me which one.

OK, I realize it's scaled down for the Web, but here's the point.  This photo is mostly about that nice sky, not the fine details.  If you had an 8x10 of this photo up on a wall, it would be hard to tell if the sensor was 1.5", 1/1.7", or what.  

There is a little blurring at the corners, but that's from the lens, not the sensor (I should have used a narrower aperture.)   There's nothing else about the picture that gives any indication about the sensor size.  

There are many lighting situations where you won't be able to get around the limitations of a small sensor, though.  Small sensor cameras have tiny pixels that leak voltage into each other, clip highlights more easily, and so on.  They also have very high depth-of-field and generally very poor low-light performance.

This is why serious photographers tend to prefer DSLR's and large-sensor mirrorless cameras for their digital work.  But there are always exceptions.  If you can pick the right scenes and lighting, no one may ever know that you actually used a pocket digicam for that photograph.  And some of the newer small-sensor cameras have special settings and in-camera processing to try to extend the dynamic range and generally improve the picture quality. 

Tiny-sensor cameras really are terrible for some uses, though.  They are especially awful for indoor pictures of people, unless you use a flash, and even then you have to be careful.  But even a tiny-sensor pocket digicam can yield gorgeous pictures sometimes.

Which Camera To Get For Landscapes?

We've seen that if you want detailed landscapes with digital cameras, megapixels actually do matter somewhat.  Sensor size is also a factor.   So, when we say "Megapixels are unimportant," or "Your camera doesn't matter," it's with the understanding that your shots are more about composition, tone, and color than about fine detail.  Those, more than detail, are the foundations of good photography. 

Don't get stuck on the Megapixel Treadmill just for the sake of keeping up with the Joneses.  In the first few years of digital photography, "more megapixels" was a significant thing.  Back in 2000, any improvement was good.  Today, in many cases we've already passed the optimum megapixel counts for the sensors.  Going from 18 MP to 24 MP actually has as many drawbacks as it has benefits... unless we're talking about full-frame sensors.  Even there, it's diminishing returns past about 24 MP.

In other words, buy more megapixels if you know why you want them and how it will affect your pictures.   I didn't bother with megapixels for a long time.  I've sold photos that were taken with my old 6-megapixel bridge camera, which wasn't even a DSLR.  The buyers didn't care;  the pictures were nice looking, they fit the requirements at the time, and that's what mattered.)

If I shot nothing but low-ISO landscapes, I'd go for a high-megapixel camera with at least a 1" sensor, and preferably APS-C or larger.   But if you want to do a lot of indoor pictures in dim light, you'll do better with something in that optimal range of megapixels vs. pixel size. 

As of when I first wrote this article, the ideal ranges were:

- An APS-C (crop sensor) camera with at least 12 MP but probably no more than 18 MP. 

- A full-frame camera with at least 18 MP and probably no more than 24 MP.

Recent improvements have nudged these upward a bit.  (Consider the Canon T6S with 24 MP crop sensor;  review here.) 

Again, if you're primarily doing daylight photography, higher megapixel counts can be beneficial up to a point.  For APS-C I'd go as high as 24 to 25 MP (e.g., Nikon D3300, D5300, or D7100), and for full-frame there's the Nikon D800 at 36 MP. 

Theoretically these are not as good in low-light due to increased pixel crowding, but they are excellent for working at the lowest ISO values. 

In Part Two we'll look at some specific cameras, as well as some lens choices.  I've said earlier that my first choice is still a film camera, so we'll look at some of those, too.

As always, thanks for visiting my website!

Contact me:

3 p 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 letters.

Home Page

Article & Gallery Index


All photos on this site are Copyright 2010-2014.  All rights reserved.


Main Page

Back to Top of Page