A digital raw file is not an original image
kenrockwell.com
kenrockwell.com
Astronomers used to use huge glass plates to capture telescope images, but when CCDs became available there was a revolution in data quality. It's true that astronomers probably don't care about the same things photographers do wrt capture medium, but in astronomy, CCDs wiped film off the map.
The final killer for film: You can't back it up. Your slide may last forever, but if you lose it, or your house burns, game over.
Also, while individual digital recording devices have short lifespans, it is trivial to make completely lossless identical copies of digital information. However making a good copy of an analog recording is notoriously difficult.
No film is not forever, film stock goes bad too. One very good example are the dozen of famous classic movies for which no good quality print remain (i.e., many of the hitchhock films). Digital formats may change often but digital data is much easier to copy and transmit between formats which makes it much easier to protect and store without losing quality.
And film is not inherently more truthful than digital data. Film also depends on sensing and is subject to sensory errors. For example different types of color film can render the same scene very differently. Currently, digital sensores are not as accurate as film in terms of details but this may change.
One ironic thing is that one of the main reason raw formats are not "truthful" as Ken Rockwell says is actually partially Ken Rockwell's fault. See reviewers like Ken have been so obssessed with noise in digital cameras, that camera manufacturers have started to use noise reduction algorithms before the raw stage so that their images will show up better in online reviews (such as the ones that appear on Ken Rockwell's site). This of course is not good news because almost all noise reduction results in loss of detail, and a raw image should be free of noise reduction because it should be the image provided by the camera sensor as faithfully as possible. A user can always apply noise reduction later.
Any recording medium has limitations, and whether it's digital or analog is neither here nor there.
If you rescan your 35mm slides at high resolution in 10 years, you'll get a really nice picture of the individual grains in the emulsion, not a better picture.
This post http://forum.blu-ray.com/showpost.php?s=1e1304228adc42f80162... explains the resolution of film well (quoting here for posterity),
> Film resolution is specified in resolving power in c/mm (cycles per millimeter) or lp/mm (line pairs per millimeter). Diferent stocks have different resolving power. Normally finer grained (slow) film has more resolving power, higher sensitivity (faster) film has less resolving power, etc, all else being equal.
> A simplified triangle of image quality capability is made by grain-speed-resolution. If you try to get more speed, graininess usually increases and/or resolution decreases, etc. With advances in film emulsion technology the triangle gets bigger. You get higher speed with the same fine grain, equivalent sharpness, etc.
> If you have coarse big grain you get more speed (sensitivity) but the resolving power is decreased, while if you have finer smaller grain, packaged in a more uniform way into a thinner emulsion layer, you get better sharpness and the ability record finer detail per millimeter, but less sensitivity (you need more light) somewhat similar to having more pixels packed into a sensor.
> The lens on the camera also has a resolution limit and the combined resolution of the film emulsion and the lens resolution that ends up on the final image on the negative is less than each's.
> So having the resolving power of the final image (c/mm) and the size of the image (mm) you can multiply both and get what the resolution of the film/camera/lens system is capable of.
> Also what we perceive as grain on photographic images is actually grain clumps as the grains are randomly distributed in irregular patterns within the film emulsion. (The smoothing and more uniform distribution of grain in film emulsions is one of the ways film quality has improved over the years) We're not looking at the individual grains themselves when we look at images in normal picture and movie viewing magnifications. To see the real individual grains you have to use those microscopic enlargements where the image is blown up so much you can barely make any of it.
It's worth noting that the sort of optics common on a 35mm pro cinema camera are a bit large for a hobbyist still photographer to be toting around.
Film stock does not last forever either, it degrades with time. New film stock, according to Wikipedia, can last "hundreds of years" however. http://en.wikipedia.org/wiki/Film_stock#Deterioration
Interesting article, but more interesting is his usage of the trademark symbol. 28 times within 175 words. Calm down buddy.
He told me that a Portuguese tv station is deteriorating and the fumes it gives off spreads the process to the other film stock and even if they had all the 35mm film digitisers in the world running 24/7 they would still lose a significant portion so they are having to make some tough decisions about what to save.
oh and it was Bell-Labs that invented pixels.
You'll probably spend about $10 / roll for a contact sheet, and $40 / frame to get a print / scan that shows off this quality. Plus, your time, going down to a lab, dropping off, and picking up film. Then, you'll spend more $ on top of that archiving your film (and that expensive scan) perfectly for 20 years when ultra-super-drum scans are available at $2 / frame. In this fictional world, 20 years from now, if you're a pro your client no longer cares whether your image now looks better, they're done with it. If you're a stock photographer, you'll have to forgo the $0.001 iStockphoto is now paying out for full rights, and if you're an amateur your old photos are good enough.
Worse is better, digital is better at the things almost everyone (including 99.9% of pros) care about.
He refers to exposed film as if it hadn't been fucked with -- you can't go back and reprocess it differently (like you can with RAW).
He describes RAW files as being a fixed array of pixels with lossy compression -- that's anything but RAW -- the good formats are the un-demosaiced elements straight from the sensor, and that means that you can do things like rotate the image at subpixel resolution (especially for non-Bayer-filter sensors like the Foveon).
I don't shoot jpeg since I don't always trust what the how the camera is going to boil the image down and would like option of going a different direction later.
Raw just gives you a bit more latitude than jpeg and a hell of allot less hassle than film.
Film has a definite resolution. It's not as easy to measure as digital, but it's there.
I looked at film in a microscope, and the grain looks like colored clouds, i.e. instead of square pixels, you have roundish blobs, but they are certainly there.
Film has some great qualities, but durability is not one of them.
At a certain point "more detail" is just wasted effort. You're not going to stand right on top of a 15' tall blown up photo, counting film grain OR pixels.
I mean, where is my HDR 64bpp 300 PPI display? I remember hearing about HDR displays some 5 years ago, current bit depth is from 90's and almost 10 year old CRT can pull more pixels per inch than todays "highdef" screens.
edit: throw in factory calibrated too into my "requirements"
If our displays were better, it probably would expose the limitations of current digital capturing, and therefore push capturing tech forward, so that remaining gaps between film and digital could be closed.
I don't know what camera was used to take it, certainly not a D1H, but I do know it was of the studio, tethered-to-a-computer type, so I want to say we were looking at about a 10 megapixel image.
If the 'cost' of this is 'oh noes I need to buy Adobe Lightroom or Apple Aperture as well as my camera,' then so be it. The majority of people -- amateur or not -- that are shooting in RAW are doing it with DSLRs. They have already shelled out hundreds of dollars for the camera and possibly more for the lenses, who cares if they find out they need to purchase a ~$100 software program to fool around with their raw images.
{edit} Even if people don't know that they need to buy a separate program to process RAW image files, they get what they deserve. If you're on that much of a budget, why are you spending hundreds of dollars on a DSLR without doing even the basics of research into what you will need? Not only that, but you can do RAW editing in free apps like GIMP, you just won't have all the bells and whistles that the higher-grade apps have. {/edit}
> 20 years from now we can re-scan our film and get 2029-level image quality.
Unless your storage location burns down. Assuming that the film still exists -- and isn't lost in storage -- in 2029, blowing it up to 100 megapixels is just going to show you 100MP of grain from the imperfections in the lens you were using.... Or maybe he's advocating that amateur photographers should spend thousands of dollars on the absolute highest quality of lenses to take photos of their cats to upload to Flickr? Remember he started out this 'rant' talking about being 'consistently amused by innocent hobbyists who go through the aggravation of shooting digital camera raw files.'
> The Wizard of Oz was shot on film in 1939. Today it looks great on Blu-Ray High-Definition DVD.
Maybe could it be because The Wizard of Oz was stored in a high quality vault, temperature- and humidity-controlled, possibly in an old salt mine to minimize the possibility of bacteria as well? What about that 'cost' to amateur photographers? How many amateur photographers from 1939 -- or even professional photographers from 1939 -- have 100% of their collection sitting around today in the same pristine quality as the Wizard of Oz cans?
{edit} Also, the Wizard of Oz wasn't shot by 'innocent hobbyists {/edit}
> All we had was radio in in 1939, so how did they know how to make High-Definition DVDs back then? They didn't, but by having the forethought to shoot on film, they knew they always could scan the original raw film images with better equipment in the future.
Huh? So the film-maker said to themselves, "Hmmmm I know that we don't have the ability to shoot HD video presently, but in 2009 the ability to scan this back into these 'computer' things will make it possible to relive this film in HD quality on a 60" plasma screen!" They used film because it was all they had available.
> With film, we always have our original images.
Unless we don't (see: acts of god, etc). Digital images can be backed up.
> At best, digital raw files are a close approximation of what came out of a sensor, not what went in. Digital raw files have already folded, spindled and mutilated the original image with the limitations of that era's technology. Digital camera raw files have already limited the original image in at least three different ways.
The point of a RAW file is that it is the RAW data that came off the sensors. Not a 'close approximation of what came out of the sensor.' That's why there are a million different RAW formats, because the internals of the cameras are different camera-to-camera let alone manufacturer-to-manufacturer. That's why there can't be a standard unless we are standardizing on the way that the sensors are setup, and I would argue that the would slow innovation.
> Archiving film always allows you direct access to your original living, breathing live images at any time in the future.
Unless you stored it wrong. And what's up with the imagery 'living, breathing?' Are you saying that the photo would be less 'living or breathing' if I had used a high-quality printer to take my RAW images onto archival quality photo paper?
> If you demand the best quality for serious subjects, and don't mind investing a lot of time to get it, then step all the way up to film.
Wait a minute. I thought we were talking about hobbyists here. Have we now cross over into the realm of the 'professional?' Wtf are you talking about 'serious subjects?'
> Pixels were invented in the American space program of the 1950s and 1960s to allow computers to process images. The concept behind using pixels to approximate an image is that so long as you have enough of them, you'll have a close-enough approximation of the original image.
There's also mathematics behind this all too. Look up DFT (Discrete Fourier Transforms).
> Most digital shooters are wary of this, knowing that whatever they shoot today in digital may or may not be good enough to sell to tomorrow's market. Got raw files shot in 2002 on your then state-of-the-art $5,000 Nikon D1H? Enjoy going back to your 2.7 megapixel files! You may as well delete them now.
Might as well make fun of early-adopters of film technology where people always had stern expressions because they had to hold still for a minute or so for the film to capture enough light to produce a photo.
Look at those losers! They couldn't even take a picture with a smile on their face! Now they're lock in time forever frowning! Not like me! I waited until I could _smile_ in my photos!
> You can't go back to a raw file and get more resolution. With film, you don't have to make a resolution decision until you scan it.
You don't have infinite resolution with film. This is a fallacy.
> [...] because you still have your original raw image captured alive and well.
> With film, you always have access to your original live image.
> [...] or just look at the naked film in all its original glory.
He keeps using these terms like 'alive' and 'breathing' to describe analog film and prints. I'm beginning to sense a bias.
> I support my growing family through this website.
Ha! I've found your bias. Your family is analog!
The terrifying possibility that a potential client will pass over Ken Rockwell's RealRAW EX PRO™ and opt for a less expensive photographer who uses a digital (gasp!) camera.