It seems as if every other month some camera magazine article has some hapless debate about the end all discussion of the full frame debate for Digital SLR cameras (DSLR). It’s about as silly as the predictions newsweek made 15 years ago about how the internet would fail. I am not here to tell you how sensor formats will shake out for the next 50 years, but only how things are now. It could change in 3 months, likely in 3 years, definitely in 30 years. I don’t want to have some a-hole coming back here in 2025 telling me that I was wrong. If you think I’m wrong, it better be before 2012.
I will be focusing on the full frame debate from the Canon perspective. I have the most experience dealing with Canon but the same applies to Nikon and pretty much every other DSLR manufacturer. There are really four major things that should influence your decision.
- Cost.
- Field of View.
- Physical Size.
- Photosite Size.
Some of these things can be remedied with research and some of them are just limited by the laws of physics. I will give you a brief overview of the breakdown of these categories and what they mean for both crop and full frame. I will then not declare a clear winner, and force you to make up your own mind.
Cost
Full frame cameras cost more to produce. It is not simply demand or marketing cache driving this fact like the case of LED based LCD TVs. The sensors are produced on wafers and because of their significantly increased area, fewer full frame chips fit on a given wafer. This would lead directly to a 1.5 or 1.6 price premium (the surface area ration) but things like manufacturing defects further compound this problem.
A defect that causes a chip to be rejected would have a larger impact on production yields. For example: If we assume that a given wafer can produce 180 APS-C chips or 100 Full frame chips (because smaller blocks pack more efficiently on a circle)and that the process will have 10 randomly distributed defects across the surface of the platter, this will yield 170 out of 180 for APS-C and 90 out of 100 Full Frame chips. For our hypothetical process the same chip manufacuring run produced 94.4% yield and 90% yield. These statistical yields are very important with such valuable chips. Additionally full frame chips also cost more because they are physically larger. This will be discussed more in the section on physical size.
Ultimately, no matter what happens with the R&D tech, Crop cameras will always be cheaper than their full framed brethren because they simply cost more to produce.
Field of View
Field of View, also called crop factor, focal length multiplier and any of a number of confusing terms is the thing that most people recognize as the difference between a crop camera and a full frame. Effectively it’s the ratio of the size of the sensor to a standard full frame camera that psuedo crops the image creating ‘magnification.’ This is mostly true. For a 35mm camera a 20mm lens has specific characteristics that do different things with shapes through distortion and other principles but it is most important to know that for the average consumer you can just multiply the focal length by the given ratio (2, 1.6, 1.5 etc)
This becomes an issue at the fringe focal lengths. A 24mm f2.8 is ~$350 to have a crop camera with that field of view it would require a ~15mm lens. This means that the 14mm 2.8L is required which is about ~$2000! That sucks! Forget about finding anything that goes 14mm wide on a crop camera in rectilinear fashion. Additionally the 50 f1.8 prime that is so great in just about every system – a lens that has been the standard of budding photographers since time immemorial – requires a 30mm lens. The 50 is about ~$100 the 30 is $300-$500 depending on the make and model (there is no 30 1.8).
On the other end, the crop gets their first good news of the day. A 300mm f2.8L costs ~$4400. If one used a crop camera they would get roughly the same coverage with a 200 2.8L. A lens that costs ‘ONLY’ $750. This is an amazing bonus for the crop shooter. It also allows the photographer to only use the center of the image circle where the lens usually performs the best. Using a crop camera is significantly advantages for any focal length above 200mm on a full frame camera and still has it’s value above ~110.
Physical Size
Not every one sees this category the same way. Some people look at a bigger camera as durrable, sturdy and with more innertia to resist accelleration. Some people look at a bigger camera as heavy, bulky and intrusive on actually bringing with them. No one can take a picture with a camera they don’t have with them so this will depend a lot on personal preference.
The only fact of this category is that a full frame camera will be larger. The imaging circle is larger, the mirror box is larger, the viewfinder is almost always brighter/larger, and the accompanying circuits tend to be larger. This category is open to interpretation but with the larger view finder I tend to think of this as a benefit to full frame.
Photosite Size
This is the one that people talk about least, but has likely the most influence on the quality of images coming out of the camera. New lenses can be designed that take advantage of crop characteristics and reduce the full frame advantage. We have already seen this with lenses like the ultra-wides of 10-22mm from Canon or the 11-16 2.8 from Tokina. Also we have seen crop-only primes from Nikon and Sigma as well as others. Optics can be redesigned, but the wavelength of light is fixed by the laws of physics.
The manufacturing process of camera sensors have crammed so many pixels onto such a small area that the recently released 7D from Canon is only approximately 5 times larger than a single wavelength of red light. What this means is that no matter how sharp the lens is, it can’t be focused any tighter than 20% of the pixel size. This creates diffraction limitations at only F6.3. Additionally the electrical properties of the underlying silicon have trouble differentiating signal from noise at certain exposure thresholds leading to ugly red, blue and green dots as well as other image artifacts. The 5dmk2 on the other hand has photosites roughly 4 times that of the 7d yet has 30% greater overall resolution than the 7d. The difraction limitation remains up around f11.
This leads to better ISO sensitivity and dynamic range despite being introduced over 1 year later. They are both amazing cameras with different purposes, but in terms of pure image quality, a full frame should always best a crop at almost any price point.
Conclusion
The crop will always be less expensive than its full frame cousin. Full frames are becoming more affordable as manufacturing processes become cheaper. The glass selection limitations of crop cameras are less and less a problem every day. The ultimate conclusion should be different tools for different tasks. Anything wider than ~30mm on a regular basis would be best served by a full frame. Anything narrower than 200 could benefit from a crop. Studio work and high sensitivity natural light work will benefit from the more expensive tools available to full frame shooters. (Amateurs, Sports & Birds: Crop; Studio, Photo Journalists, Landscape, Weddings: Full Frame)