This morning a technical article appeared on Luminous Landscape regarding image quality of what they call Compact System Cameras and we call Single Lens Digital. The article attempts to quantify image quality differences between Digital SLRs and CSCs. However, unlike most articles on Luminous Landscape, this one is riddled with vague statements which do not support the article’s conclusions.
By the definition they present, a CSC could be the same as an SLD or pehaprs a subset, if one excludes Sony Alpha SLT-series. Note point #2, unlike DSLRs, there is no flip mirror or pentaprism, does not exclude digital cameras with translucent mirrors. Point #1, [CSCs] rather employ an electronic viewfinder, excludes a number of SLDs, depending if the optional EVF counts, as offered by many models from Panasonic and Olympus. At least we can see that most Panasonic and Olympus Micro Four-Thirds cameras, as well as Sony NEX models, fall within their definition of a CSC.
When doing such an analysis, one must understand the separation of facts which are inherent to the system from those which are currently implemented. A simple example is the shutter-lag which presently favors DSLRs while having no technical reason for it to be slower on an SLD. A DSLR even has to move a mirror within the shutter-lag.
When describing the impact of using an EVF – or we would say of not having an Optical Reflex Viewfinder – the article correctly points the possibility of much smaller and lighter cameras but then they make some statements which appear to be based on a random sample of CSCs.
Particularly, they say that while vibrations caused by mirror-slap are gone [true, compared to SLRs], using a camera at arm’s length exaggerates shake. While true, this is completely artificial considering their definition of CSC requires an EVF and – by the way – DSLRs with Live-View can also be used at arm’s length. Just because they can be used this way, does not mean they have to and it certainly should not be a defining difference since both CSCs and DSLRs can both the used the same way. As a matter of fact, I am seeing more and more photography students who are used to shooting with a DSLR at arm’s length.
Then after saying that CSCs are more prone to vibrations, they say that this is minimized because of image stabilization which is available on most CSC lenses. This is a bogus claim since DSLR also have access to stabilized lens and because most CSC lenses are NOT stabilized. A quick search this morning shows 12 stabilized vs 16 non-stabilized lenses.
The claim that follows is that the speed and accuracy of autofocus is the same on both cameras. While we see no technical reason why it could not be the same, the present state of affairs is that top DSLRs focus faster than top CSCs while accuracy is equal with a properly calibrated phase-detect autofocus system. On the other hand, as we pointed out yesterday, contrast-detect autofocus does not suffer from calibration issues. Note that if we include all SLDs in the definition of CSCs, the distinction becomes arbitrary since Sony Alpha SLT-series use phase-detection autofocus just like all digital SLR cameras.
Even without using a translucent mirror, there is no reason CSCs have to use contrast-detect autofocus. Fuji has already implemented sensors with built-in phase-detection in their compact digital cameras like the Fuji Finepix F300 EXR reviewed here.
Something similar can be said about sensor size. There presently both DSLR and SLDs with the same sensor sizes and probably even the same sensor. There are no full-frame SLDs yet, but it is highly likely that Sony will produce one since they said all future DSLRs [sic] will have translucent mirrors.
The parts of the Luminous Landscape article about lens design are interesting. The shorter flange-distance required by CSCs has an impact on how lenses are designed and allows for wider maximum apertures for the same focal-length designed for a DSLR. On the other hand, lenses are more prone to vignetting if they take advantage of the shorter flange-distance, so it is a double-edged sword. Keep in mind the if part of this statement, since SLDs can use DSLRs lenses with a simple tube-like adapter.
There is a chart that summarizes Imatest’s findings. The problem is that if you treat this as the current state of CSCs, it is wrong. If you treat it as the potential of CSCs, it is still wrong but for different reasons.
The true difference between SLDs and SLRs is their potential for smaller size and weight, which is very well shown among current-generation models like the tiny Panasonic GF3 or Sony NEX-7. Size has more implications for differences then inherent properties of each system.