some people have put forward that the image quality of current digital cameras can exceed almost anything. Some time ago I had read a number of pages which compared cameras like the 10D with Medium and Large Format film, such as:
Only one site found here published a test which I felt was more or less exactly what I was seeing.
I find that 4x5 is such an undervalued format that I think more needs to be brought out of 'hiding' and its advantages explored. Certainly in the past this stuff was so expensive as to be stratospheric, but as the prices of image stabilized zoom lenses rises (like over US$1000 for a 70-200 f4 zoom with IS) and the amount of second hand material appears it seems that the optics and accessories are becoming comparable and even cheap. My intention here is to bring attention to the fact that for less money you can get better images (if image quality is what counts).
I like photographing landscapes, and things which are static. In this type of photography the most important thing is detail that the camera can capture for making prints. So if you're the type of photographer that
So, this is what this page is about, it is my results of the test between a high quality digital and a large format camera to see if the digital really can compete.
Put simply, my findings from this test are that if you are fanatical about making high quality images that you will not get images from anything less than a montage of at least 2 images from perhaps a camera like the 1D III to rival the images you get from a Large Format image.
Even to make digital 'montages' from multiple images you will need to make your single image composed of many digitals to capture what you can with one frame of 4x5. Then you run into other problems and processing issues.
The overview image, no test targets here, but a 'real world' example :-)
First, lets look at the potential issues with my test:
Camera age: The 10D will be considered an older camera at this point. However, despite the wishes of many, things have not improved by double between this camera and the current model (such as the 40D). When you consider the differences found between the 10D and the 40D are really not as much as a 25% improvement. The 10D was a capable camera, and while the 20D improved upon its image quality by a margin, it is not 'chalk and cheese' with the latest models. Looking at DP review seems to indicate that noise between the 10D and 40D is comparable. This is consistent with the increasing of pixel density that has occurred between these cameras.
Comparing the resolution differences between these cameras, DP Review tests don't show enormous differences there either. Looking at the comparison from their site below it would appear that the 40D image begins to alias some time after the 20 marker anyway ...
| DPRreview 10D test | DPReview 40D test |
|---|---|
|
|
So, perhaps the camera isn't that much behind after all ...
Angle of view: This is not a straightforward comparison, as the differing formats mean that obtaining matching lenses to give the same angle of view on the image is not easy for a start.
The widest angle lens I have to fit my 10D is a EF24 f2.8, while the widest on the 4x5 is a Fujinon 90mm f8. Personally I have found that the Fujinon is not the sharpest lens in 90mm range, the lens is of an old design that is not even multi coated. but since I don't have a better one it will have to do
The angle of view between these two systems is different. When I took the images I was intending to make a montage of images with the 10D to equal the view of the large format camera. I have first placed the images side by side for you to see how each of the images look overall.
| Large Format negative | single 10D image |
|---|---|
| |
 |
The 10D covers less of the scene than the Large format, so to show you more precisely what the differences in coverage are, I've scaled them so that I could overlay them together (pixel for pixel) in the figure below. The images were over-layed and a rectangular frame placed around the edges of the 10D image
So, clearly the 10D has the advantage of not needing to cover as much area as the large format film camera.
The sheet film is big around 10 x 13 cm in size, in contrast the digital capture area is 2.27 x 1.51cm, this means that the digital system will really need it high precision and first generation capture to make up for its handicap of small size. Using my than perfect, low priced Epson 4870 flat bed scanner at 2400 dpi (certainly within its ability) I get a huge image. Much bigger than the one from the digital.
This is important, as the fellow who tested the 1Ds at the above mentioned test effectively biased the results of in favor of the 1Ds by only scanning the 4x5 at 1200dpi.
| 10D RAW file | 4 X 5 film @ 2400 dpi | relative comparison of pixels |
|---|---|---|
| 3,072 x 2,048 = 6.3 Megapixels |
12,000 x 9,600 pixels = 115.2 Megapixels |
 |
This we where we can see the sheer amount of area recorded in one hit in a large format exposure. This also effectively shows the differences in print sizes that can be had. Assuming the images are as good as each other, it means that if you printed an image from each (at any DPI you like) then the prints would be of the same scales as that if you put one over the top of the other.
Now, a trick that can be done with the digital is that you could turn the 10D on its side (making it a landscape when the orientation of this particular image is portrait) and make 3 images to be joined together, but you would still need to upsize the 10D image by more than 3 times to even keep the same amount of pixels in the area photographed. Still, this is doable (I've done it) but it requires some post processing and hopefully no changes in the scene (like wind moving anything around)
| advantages | disadvantages |
|---|---|
| digital: not required to cover the entire picture area (so that montage techniques could be used to make a single larger image from many smaller ones. Things would look much worse than this if I had used a lens wide enough to cover the same field of view as the Large Format camera image | film: lower magnification on film, thus more image detail is recorded by the film than the sensor. This allows the sensor to capture "more detail" per shot (if it can ;-) |
| digital: capture from RAW was used which allows the camera to yield its best | film: The scanner used was a flatbed design, which is not specialized to capture all the detail that could be captured on the film (NB the film image could do better still) |
| film: sheer size of film makes up for many imperfections of lens / film / scanner / technology | digital: by being compact digital needs to squeeze every last drop of technical advances in lens / sensor design |
So, keeping the theory and these size differences in mind lets look at a 100% segment of each...
So, what did we get?
The problem with the Internet is that I can't show you the prints. You've seen
the overviews (scaled down) so I can instead show a segment from both images.
By looking at the detail, you can see what will form the prints. The segment is
one of the palm trees you can spot in the middle of the picture, against
the white part of the cliff face.
| 10D image | Film Scan image |
|---|---|
 |
 |
Gosh ... its impressive isn't it. Remember, if using a light jet printer or other optical RA-4 laser process these represent small detail parts of the image. But just look how much more detail there is.
As you'd expect from the numbers that sizes are quite different! This means that printing each at the same DPI you'd get prints this much bigger from your 4x5 negative than from the 10D.
Clearly not only is the 10D image composed of far fewer pixels, it also holds far less detail than the scan of the 4X5 negative. To clarify this, I've up scaled (bi cubic sharper) the 10D image (working in 16 bit from the RAW file) to make it the same size in pixels as the crop from the Large format. If you hover your mouse over the image, you'll swap to the image from the Large format camera. .
While the image is slightly different (working from a negative scan its hard to balance the colors and contrast of a digital), you can see that there is clear detail in the fronds of the palm (especially of the upper right plan) and the leaves of the bushes above on the cliff
This confirms that the large format camera is still the top standard of the recording of the image world. While the digital cameras of different levels are meeting and exceeding the results achievable with 35mm film, Large format is still the only real alternative for high definition imaging. Of course, the professional photographers working away in studios making high quality advertising work know this. In fact, they're probably using stuff like the Better Light scanning backs (if they can afford it). Take a look at this page comparing them to DSLRs Interestingly, it also supports the calculated findings of Roger Clark on this page
But don't think of this as a film vs digital comparison. It is really only a comparison of capture media under ideal circumstances.
Sensors are VERY accurate, but film is cheap. So brute force wins over high technology when it comes to the outright sizes.
Depending on the subject photographed, the conditions and your technique if you are wanting the best in image quality for the largest of prints you should consider using 4 X 5 inch or larger format film. But keep in mind: Taking good images with an LF camera is really a dedication to the task. It is not an instant success by itself.
If you don't make the dedication then the capacity of the LF camera will be lost. There are many places out there in the wild where you can and will make less than ideal images with your LF camera. My current view is that if you are chasing the highest image quality for landscapes, and if you aren't taking action shots telephoto wildlife or sports and you are seeking the best images from your landscapes then you should look to Large format.
Especially if you have already been trying things such as
I have seen many people chasing image quality for their landscape photography spending thousands of dollars on the latest in 12Megapixel DSLR cameras, finding that their lenses aren't good enough when enlarged, spending more on better lenses and then finding they are still not satisfied.
Given the low cost of a simple large format set up, it is hard to understand why more people don't choose to try it. While it is possible that the current top of the range DSLR (such as a Canon 1D III) with the highest quality optics might match a 4x5 image, its just as likely that it won't. Either way, you'll have to spend something over US$10,000 for the ownership, and have to carry around as much weight as a typical 4x5 field camera setup.
Alternatively to spending these large sums, you could own a lesser DSLR and a few modestly priced lenses, plus a large format camera and scanner. You'd have spent less than half the money and be able to make better images over all by not making one camera do everything.
I have found that for many landscape picture taking situations that compact digital cameras are very capable choices. They are just good versatile image making tools as a DSLR, especially if you work with macro, and fit nicely into a day pack.
Have a read of my comparison with my Coolpix 5000 and my 10D to see if (like me) you too could benefit from a different strategy of mixing your digital camera tool set with a little bit of big film, saving weight in your pack and money in your wallet!
I hope that I've been able to provide you with some helpful information.
If you are interested in making any comments or provising any feedback on
this article, then you can e-mail me at
I'd be glad to have a chat ...
