Pounds per pixel
I had cause recently to muse on the rather variable cost of a pixel. It's like this.
Pentax has a new digital camera out. It's called the K-7, it is apparently quite nice, and the body will cost roughly £1,200. It will produce images with 15 million pixels. This is fairly typical for a good quality digital SLR.
Now consider the large format camera I bought the other day. It cost £200, and it will take 5x4 film. Even a medium quality flat bed scanner can produce a three hundred million pixel scan from a 5x4 negative. Lenses for the LF camera are roughly the same price as high end digital lenses - a Rodenstock Apo Sironar S 150mm is comparable in price to the Pentax 31mm limited. So I am getting a pixel for 0.000067p, vs. 0.008p for the digital photographer, a factor of over 100 better.
Now of course digital is free once you have the camera, and large format isn't. But the £1,000 I have saved will buy quite a big pile of Fuji Velvia 5x4. Sometimes old technology is cheaper and better than new...
Pentax has a new digital camera out. It's called the K-7, it is apparently quite nice, and the body will cost roughly £1,200. It will produce images with 15 million pixels. This is fairly typical for a good quality digital SLR.
Now consider the large format camera I bought the other day. It cost £200, and it will take 5x4 film. Even a medium quality flat bed scanner can produce a three hundred million pixel scan from a 5x4 negative. Lenses for the LF camera are roughly the same price as high end digital lenses - a Rodenstock Apo Sironar S 150mm is comparable in price to the Pentax 31mm limited. So I am getting a pixel for 0.000067p, vs. 0.008p for the digital photographer, a factor of over 100 better.
Now of course digital is free once you have the camera, and large format isn't. But the £1,000 I have saved will buy quite a big pile of Fuji Velvia 5x4. Sometimes old technology is cheaper and better than new...
Labels: Cost Benefit Analysis
posted by David Murphy at 09:27
4 Comments:
Im a big fan of the blog. I sent this post to a friend who is a designer and uses a digital SLR. I have flirted with the idea of buying one myself so your post was interesting to me. Here is his reply.
"
I can shoot 250 - 300 frames on my DSLR, and could up that to 1000 for about £20. I can pull all those images into my darkroom ready for printing in seconds, and only print the keepers instead of having to expend time and effort developing and printing each frame, only then to decide which to store digitally via a manual scan.
My development workflow (and I'm not a practised professional) is typically 30-40 minutes long from import to development and publication on flickr with metadata intact. I don't have to tag photos manually with dates, and some newer DSLRs will also tag images with location lat-long automatically.
A 6 megapixel image is enough for most mid-format printing, and a 12 megapixel image can happily be blown up to A0 or greater. Cost per pixel is not really a good measure of "value" - where's the cost per darkroom hour? The cost per curation? The cost per storage litre? Digital photography has *never* been about quality. It's only now that quality is starting to rival film. It's about convenience, opening up flexibility, and conserving effort.
Sorry - that was a harsh rebuttal, but I get a bee in my bonnet when people fix on a single measure and then trash something based on it, without even qualifying why the measure is important in the first place.
"
No, not harsh at all - comment is appreciated...
I don't disagree with your friend in the slightest, except in his assertion that 12M is enough for an A0 image. One of the cheapest good UK photo shops, Peak, has a reasonable printer which is 254 dpi, so let's use that. A0 is 33.1 × 46.8 inches, or 8400 x 11900 pixels, roughly. That's about 100M pixels.
Now, it may be that some 12M images look OK at A0 size. But if you are interested in making each pixel count - something I agree wholeheartedly is irrelevant to most professional designers and photographers - then there is no digital camera available that will produce an A0 print at 254 dpi, because there is no digital camera that is rated at 100 mega pixels. A large format camera, on the other hand, can produce a 100 megapixel scan without breaking a sweat. Yes, taking one shot with it takes me ten minutes. Yes, a single image costs three pounds or so. And yes, there is no way of telling if that image is good until you get the neg back from the developer. But it does kick any digital camera's butt for pure image quality, even a Hasselblad H3 (which costs more than ten grand).
His response:
"
A good response, and highlights his "straw man" argument - nobody in their right mind is buying digital cameras for pure image quality". They're buying them for all the other wins, and the quality is "good enough". I agree about the A0, but I also know professional photographers who have shot in digital and had their images blown up to A0 and beyond for marketing purposes. The key is that, as soon as you get into larger prints (say for billboards etc), you're into offset, which means that "dpi" is no longer the measure. It's a bit like trying to compare your LCD screen with precise pixels with a cinema screen. Even digitally projected cinema screens don't really rely on "pixels" for quality - they rely on your eye's resolving power, which is not 300dpi at x00 yards away.
Pulling a 100 megapixel image through a digital printer and on to an A0 is going to result in bleed, and you're not going to see the pixels - that's the point. Even printing at half the resolution (50 megapixels, and still outside the bounds of high-end digital cameras) is going to be acceptable to the human eye.
So, yes, he's right - the "cost per pixel" is very high for digital cameras. But nobody's selling them "per pixel" at the professional level, and at the consumer level, it's just another number that marketers can point to and say is bigger than the competitor. It's meaningless. The best camera, as they say, is the one you have with you, and I suspect there are as many interesting and "good" shots taken with shitty iPhone cameras as there are with large format cameras. Probably more, as you're more likely to have your iPhone with you than your large format behemoth ;-)
"
I read this blog for finance so this is the last reason I thought I would post a comment here. Its the only finance blog I still read. :-) Keep up the good work.
That's all very fair - thank you for the comment.
My interest in photography comes from artists who happen to use photography - people like the Bechers, Dijkstra, Ruff, Wall, and so on. All their best work (and pretty much* all their work full stop) is taken large format because they want to be able to print between 4 by 3 and 20 by 12 (feet). Even at the smallest of those sizes, in gallery conditions, you need LF image quality. But of course this kind of photography is irrelevant to 99% of professional photographers.
*The exception is Ruff's p0rn pictures.
Post a Comment
<< Home