(commentary)
One additional point has been raised by the many emails I've received:
- Big files kill workflow efficiency
This is the typical wedding photographer comment. They want FX and low light capability but they don't want big files. The underlying problem is defined in two parts: (1) they don't need to print big; but (2) 36mp files take them longer to process the thousands of shots they take at a single wedding.
Okay, I buy that. But why do these folk need a D700 replacement? They may already be on their Last Camera Syndrome equipment if the D700 suffices for the time being. By the time they really need to replace that camera, we may well be on the next generation of cameras.
Basically, these shooters wanted a 16mp D800 (D800 body with the D4 sensor). That doesn't seem like a very substantive upgrade, frankly. Let me state it this way: using the D4 makes me think the D3s was a damned good camera and used ones in excellent condition are real bargains. Sure, I get a few more pixels at about the same noise levels, which allows me to print (or crop) slightly larger (more). But at the common wedding print sizes, I'm not sure that's really going to show up visibly. I then have to look at the features that the D4 added and my reaction is still not strong. Simply put: the D4 isn't a huge step up from a D3s in terms of still imaging. A D800 with the D4 sensor would have a similar relationship with the D700: small step.
I'm also a little suspicious of the "slows my workflow" claims (please note I wrote "a little"). If these shooters are taking JPEGs, just use JPEG Medium 1.2x crop (14mp). The D800 produces pretty remarkable downsized JPEGs, after all. The 1.2x crop gives them the "sports viewfinder" view of what's happening just outside the frame lines, which I've always found useful in candid situations.
Oh, it's those 75MB NEF files that scare them. Yes, if you shoot that way you're getting file sizes 3x what you get with the D700, which right away means 3x the ingest time to the computer. The more likely scenario is that you'd shoot the D700 at 14-bit lossless compressed (16.4MB) and the D800 at 12-bit lossless compressed (32.4MB), which is only a 2x difference, though. Saved you 33% in time right there, and you still have better image quality. Even more likely: shoot 5:4 12-bit lossless (27MB), trimming a bit more off the ingest time. But what makes me think that these folk are still using a single USB 2.0 card reader? ;~) Oh, and one more thing: a 16mp D800 would have bigger files than the D700 does, so we need to account for that, too. Bottom line: <2x difference in ingest time and storage space between the "optimal D700 replacement" for these folks and the actual D800 as you'd likely use it.
Batch processing the raws certainly will take longer, and I can't discount that. However, implicit within many of the wedding photographer comments I received is that the additional pixels didn't gain them anything because "we don't print that large." Frankly, wedding photography is a competitive business. I personally would want any advantage I could get. What most of the arguments I received said was "I can process images faster" is the advantage. But that's not a customer advantage, the same problem I rag on Nikon about from time to time. All else equal, a 14" print from a D800 should show more edge acuity and be perceived sharper than one from a D700. That's not a useful advantage? The bride couldn't care less about how much work I have to do; she cares only about the final results. Am I willing to upgrade my computer equipment and optimize my workflow and absorb some processing time to get an advantage in final results? Yes, I am. If I wasn't, I'd be perfectly happy with a D700 ;~).
So let me rag on Nikon instead of the wedding photographers for a bit ;~). There's a simple answer that would have stopped the "big file" complaint: sRAW. That's right, create an in-camera NEF file smaller in pixel size than captured, as do many of the Canon cameras. Now technically, Canon's sRAW isn't raw as it contains an interpolation. But you could bin the D800 sensor to create a true raw variation (3680 x 2456). Sure it would be 9mp, which for some may be too small. But boy would that be a darned good file to work with. (Opportunity alert: why we don't have any software that can convert a D800 raw into a correctly binned raw, I don't know. This alone might solve some of the wedding photographer's complaints about the D800 if done properly [i.e. on ingest]. You know, now that I think about it, I'll bet that photojournalists wouldn't mind binning the 16mp and 24mp sensors, either. After all, most time-sensitive publications don't need 300 dpi capable of printing at 18" or bigger. Note that a really well done solution would preserve the original raw file as well as creating the bRAW one.)
