B&H Photo has Adobe Lightroom 4 for only $99 when ordered with the Canon 5D Mark III. Thank you for ordering using any of the links from the site.
I will be testing the Canon 5D Mark III (and Nikon D800E and D800 and Canon 1D X and Fuji X-Pro1 just as soon as they arrive). A big thank-you to my trusted vendor B&H Photo for making these cameras available for review purposes, though I will need to acquire the D800E and 5DM3 and 1D X for long-term testing.
Perspective on the 5D Mark II predecessor
I am not a video shooter, but I understand that the older Canon 5D Mark II was extremely popular for video, for good reasons. And the video I have seen from it has been very good indeed.
However, I was never particularly happy with the still photo quality of the 5DM2 in terms of color-speckle noise in dark tones, a strained look under harsh lighting, color rendition, dynamic range— areas on which I’ve commented on and shown for several years now. The Nikon D3x and D3s easily outperform the Canon 5D Mark II in all these areas. Perhaps the 5D Mark III will bring parity, or near parity. And perhaps not, since Nikon’s D4 and D800/D800E might well prove another leap forward also.
The 5D Mark III
The Canon 5D Mark III (and 1D X) represent another generation of sensor development. The most exciting thing about the Canon 5DM3 for me is the expectation of much improved still image quality. That remains to be seen, but I think it’s a realistic expectation.
What might also be interesting is whether the Canon 1D X offers superior image quality at lower ISOs (not just higher ISO). Since I don’t need the high speed features of the 1D X, the 5DM3 could be very appealing with its lower price point, so long as image quality is as good, and the extra 4 megapixels over the 1D X won’t hurt either.
My eyes glaze over when I read about how many AF sensors and zones there are or DIGIC 5 or whatever gobbledygook is noted; I just want to know how well it works in my hands.
And so, the most interesting changes from my perspective with the Canon 5DM3 are:
- The 14-bit image processing pipeline and the new sensor. Even ISO 100 was noisy on dark green foliage with the Canon 5DM2, so I hope the Canon 5DM3 solves this.
- Claimed 100% viewfinder coverage (I was always off a little on the 5DM2, which caused mis-framing for me; I frame tightly). The non-interchangeable focusing screen is of some concern (see Reader Comments below).
- Dual-axis electronic level.
- The 1.04 million pixel rear LCD with “Clear View II” technology— I use the LCD a lot to check focus and sharpness, so any improved quality is a useful feature when evaluating focus, etc.
- One-touch image zoom to check focus on rear LCD (I do this all the time, so this will be a time saver).
- (claimed) Lower mirror-bounce— for more keepers at low shutter speeds.
- The locking mode dial— the 5DM2 was very prone to having this dial change unexpectedly— one more hassle that is gone.
- GP-E2 GPS receiver. I would not use this a lot, but for certain types of remote shooting, this would be of some interest over time, and with mapping software. I will test this with 5DM3 if I can get a unit from B&H for evaluation.
- Dual cards— might be handy as a backup, but having two types of cards is a dubious choice that I thought Canon had finally come to its senses on.
- HDR— with apparent RAW support (not just the silly JPEG-only support as in most cameras).
Ben W writes:
Lloyd, I was reading your thoughts on the 5D MIII that you posted on your blog recently and, while I agree with your thoughts, I'm curious if you noticed that the focusing screen is no longer interchangeable. For me, this is an absolute deal breaker.
Through your blog I've come to love Zeiss lenses and manual focusing at very wide apertures, which absolutely require a different focusing screen than the stock one in most Canon cameras. With the 5D MIII using essentially the same screen as the 7D, the DOF seen through the viewfinder will be the equivalent of around f/4, making it impossible to accurately focus manually using wide apertures (as is my experience with the 7D). Huge disappointment! I was really looking forward to this camera, but this is a deal breaker.
DIGLLOYD: I have not had difficulty focusing with the stock screen on the 5D Mark II. How the 5D Mark III will fare is something to be seen.
I don’t know what is meant by “DOF equivalent of around f/4”, but it is true that the focusing screen can make it more or less difficult to focus accurately. As for the Canon 7D (with its inferior viewfinder), it has much smaller photosites, and therefore accurate focus with the 7D requires significantly higher precision than on the 5D Mark II or III. I think there might be a mistaken premise here, though I cannot be sure.
Another solution is the use Live View for even more accurate focus, and this can be done handheld with an LCD viewfinder like the Zacuto. It’s a bit awkward compared to the viewfinder, but more accurate.
Norman M writes:
I saw your recent post about the 5DMkIII and the reader comment about the focusing viewscreen.
The reader is correct. Most viewfinders do not actually display the "true" DoF for very fast lenses (1.8, 1.4, 1.2, etc). I don't 100% understand the reasoning but from my research I think the cause is twofold:
1) optimization of the "ground glass" surface of modern viewfinder focusing screens for high brightness, which limits the DoF seen in the viewfinder (due to the microlenses on the focusing screen being optimized for light at certain angles)
2) exit pupil effects affecting the light path to the viewfinder, which are not present when the mirror is moved and the image exposed on the sensor. This ultimately limits the maximum f-stop viewable through the viewfinder optics.
#2 is evidenced by the fact that putting an f1.4 lens on does not result in an accordingly linear increase in brightness through the viewfinder (compared to an f2.8, you would only say get a 3x increase and not 4x, because the max. effective aperture of the viewfinder is limited to somewhere between f2 and f2.8).
Here was a good article I found
I don't know about the f/4 DoF in the 7D viewfinder specifically, as it seems a little low. I have seen quotes of around f2.2-2.5 maximum effective aperture for Nikon bodies. I also don't know if the camera being FX or DX makes a difference (maybe FX cameras just have better viewfinders in general...?).
Anyway Canon actually sells (sold?) replacement viewfinder focusing screens that were designed specifically to aid in manual focus of wide aperture lenses. I believe it was called the Ee-S. It specifically states that it is for optimum focusing with fast lenses, as the cost of brightness (because the screen surface is optimized for accurate DoF and not brightness).
Switcing to live view mode gets around this problem because it does not use the viewfinder optical path or focusing screen. Also, split prism focusing screens solve this by allowing you to visualize focus in a different manner.
DIGLLOYD: Interesting commentary. If I understand correctly, the referenced article says that certain types of plastic focusing screens use microlenses which (like the AF system) effectively “see” only the more central rays, thus excluding off-axis rays, and generating a discrepancy at wider apertures that is especially problematic for lenses with spherical aberration (the above referenced article did not seem to put those facts together). Effectively a focus shift error a subject on which I’ve written about a lot (see Making Sharp Images). As I’ve previously advised, such lenses are best focused at the shooting aperture up to f/4, either by stopping down manually or by using Live View at the shooting aperture. If the focusing screen is effectively only seeing the central rays equivalent to f/4, then exposing at f/2 (for example) will be recording far more rays than those central rays, and with many lenses, that means a different point of dominant focus.
In general, the better corrected a lens for spherical aberrations, the less of an issue this ought to be. But it doesn’t explain the Zeiss 100/2 which is all but free of spherical aberration, so microlenses must interact in other ways also.
There are without a doubt focusing screen interactions. For example, the Zeiss 100mm f/2 Makro-Planar is difficult for me to focus by eye on Nikon (and I’ve heard this from others also), but I have been spot-on with Canon— so far. We shall see how the Canon 5D Mark III fares.
The whole issue also assumes that the optical paths match— they often do not; the distance to the sensor is slightly different than that to the eyepiece. One is lucky if the DSLR is within 20 microns, which is a significant error.
As far as alternative screens, it seems that the new Canon 5D Mark II offers none, so users might consider the Live View along with the Zacuto for pinpoint focus. I have used the alternative Canon matte screen on the 5D Mark II. I did not find it particularly helpful. It was darker, and I did not focus any more accurately.