Enthusiastic about iPhoto 5‘s ability to handle RAW images from digital cameras, Amy shot a short roll in RAW, only to find that iPhoto 5 refused to import them. Eh? I probably should have known this, but it turns out that RAW (known as NEF in many Nikon cameras) is not a file format as such, but a representation of the actual bits coming directly off the image sensor, with no in-camera processing whatsoever. Because image sensors are different from camera to camera, so is the layout of the raw data. So not only do RAW formats differ from one manufacturer to another, but even from camera to camera by the same manufacturer. In fact, the RAW “format” sometimes changes within a single product line! iPhoto’s list of supported cameras (lower right of that page) shows that iPhoto supports the PowerShot G5, but not the PowerShot G2, which we have.
The idea of RAW is that the photographer gets a “digital negative” — the most possible data available for post-processing and archival purposes. The downside is that image manipulation applications need to know the specifics of the data layout on the CCD to be able to do anything meaningful with the images. To overcome this compatibility problem, some cameras do perform a minimal amount of processing on the image before storage (e.g. “RAW+JPEG”).
So why don’t cameras simply support one of the many uncompressed image formats available — say, TGA or BMP or uncompressed TIFF? No data loss or compression would be necessary, and the application-level compatibility issues would go away.
Photoshop supports RAW for a wide range of cameras, and there’s the dcraw.c library for Linux, which supports some 87 cameras — so the code is out there and we can expect to see better iPhoto RAW support in the future (hopefully there’s a RAW plugin architecture at work here). But it bugs me that application vendors should even have to worry about this kind of thing. I’m sure insiders can come up with all kinds of historical and technical explanations for how we got to this point, but it’s a classic example of an industry working at cross-purposes to itself.
i haven’t investigated this thoroughly, but i suspect that what they’re calling “raw” data is quite a bit more information than would fit into a standard tiff file. the ccd data might very well be no more a picture than a pile of transistors is a cpu, if you get my drift. so saving such data in a tiff file would completely defeat the purpose, and if that’s what you want, why are you going with raw format anyway?
Walther, I’m not sure I understand what you’re getting at. Are you referring to EXIF metadata — data that stores time of capture, exposure info, resolution, etc.? EXIF is supported by TIFF, so it would seem to be a fine substitute for RAW.
A professional photographer wants uncompressed image data and the associated metadata. If camera manufacturers can give them that in a standard file format, why deliver it in a non-standard format?
That is why Adobe has developed the open standard DNG format, and has gotten lots of big name photographers and imaging pundits pushing for its broad acceptance by camera manufacturers. Read all about it at: http://www.rawformat.com/
I don’t use iPhoto any more as I only shoot RAW on my DSLR, and I find Photoshop’s built-in File Browser to be far more suitable for my needs. Besides, IPhoto 5 doesn’t read my camera’s RAW format either.
iPhoto 5 doesn’t work with my Canon Powershot S50, which is a pitty. I had to revert Photoshop too work with the raw images.
I found some additional info on this site: http://minimal.cx/2005/02/24/ilife-05-iphoto-501-and-canon-powershot-g3-raw-files/
Larry, it sounds like DNG is going to solve a lot of problems – excellent to hear. But after browsing around those sites, I still wasn’t able to figure out what DNG offers that any of the several extant uncompressed file formats (with EXIF metadata) don’t already offer. Why are they developing a new format rather than using, say, Targa?
DNG preserves the benefits of using a RAW file format, i.e. the ability to produce versions of a digital photo with differing exposures, white balance adjustments, etc. while preserving the original photo in an archival format which won’t run the risk of being eventually unreadable if, say, a camera manufacturer goes belly-up and the original proprietary, possibly encrypted RAW file format is no longer supported by any modern software.
Here’s another link:
http://www.luminous-landscape.com/reviews/software/dng.shtml
RAW image data doesn’t store pixels and color data in the way TIFF or some other uncompressed or non-lossy format does. This is why it allows you to do things like:
* Change the white balance afterwards. In photoshop you can recolor a JPG photo, for instance with the curves tool, but this is nowhere near as good as changing the white balance on the camera would have been. With RAW it doesn’t matter. White balance processing is done in your software, instead of on the camera. The results are far better then playing around with a JPG in photoshop.
* Change exposure compensation afterwards. That right, you can actually lighten and darken image using exposure compensation on a RAW file and the results will be the same as they would have been, had you set the exposure compensation on the camera before shooting. Again, lightening a JPG image can only do so much, as all the information it has are the pixels. Under exposed areas in a JPG tend to contain very little information for instance, so a whole area will simply light up, instead of gradually exposing the shadow detail. RAW usually has far more information to work with.
Also, because the image has to be reconstructed from the actual read-outs of the photosites (the CCD basically) there is a lot of room for the software to improve upon the way the camera does this internally, when it creating JPGs.
If you compare the camera’s RAW processing software that shipped with it to software like C1 or DxO Optics Pro you’ll find that the quality difference is huge. The commercial software really beats the pants off of (for instance) Canon’s RAW software.
Of course when the software improves you can simply regenerate JPGs from your RAW files for a “free” quality improvement.
All this said though, it all really depends how far you want to take things. When I say there is a huge improvement in quality, it is not something a lot of people will see immediately. It is like pro audio equipment: Most people simply do not hear (or care) about the difference in quality. RAW shooting and processing also takes a lot of time. The processing is heavy and to get the best results require you to tinker with every photo. In many cases the camera pick a lot of settings for you in “auto” mode. Now you’ll have to do all those settings manually.
Of course the flexibility of RAW formats will allow you to “save” photos that would otherwise be unusable (wrong exposure for instance), but for the majority of your shots you really don’t need it. An offshoot of Murphy’s law: The chances a picture could be saved with RAW processing inversely proportional to the chance that you had the camera set to JPG. And vice-versa.
My advice: Unless you plan on selling prints of your photos or selling photos directly, that need to highest possible quality (portraits and studio shots for glossies, say) don’t bother with RAW, unless you like tinkering endlessly with JPG photos in photoshop a lot already.
iPhoto 5 will import certain RAW formats (and hopefully one day DNG), but it does the entire RAW conversion for you, giving you a jpeg that might or might not be satisfactory, but sort of negating the reasons for shooting RAW in the first place. Any image adjustments you then make to the image in iPhoto are happening to an already converted jpeg.
When I bring an image in Adobe Camera RAW, I’m the one making white balance and exposure descisions before it becomes a tiff, jpeg, whatever, and not the software. If you’re shooting with RAW files, Photoshop (or another advanced RAW converter) is really the way to go, not iPhoto.
If you’re interested in tinkering with RAW, “Real World Camera RAW” by Bruce Fraser is an excellent starting point. The book is specifically about Adobe’s Camera RAW program, but gives you a great idea of the advantages/disadvantages of shooting withraw.
http://www.amazon.com/exec/obidos/tg/detail/-/032127878X/qid=1110813776/sr=8-1/ref=pd_csp_1/103-5165574-5734204?v=glance&s=books&n=507846
-n
Larry, Tao, Nick –
Thanks for all the great info. So it sounds like RAW (or DNG) provide a lot more flexibility than you would get with a simple uncompressed image. So now I “get it,” although I’m still having trouble wrapping my mind around exactly how it’s possible to, say correct white balance on a RAW image in a way that’s different from the process of changing it in another bitmap format. I mean, the RAW image still has pixels, and those pixels still have values. I’m having trouble imagining how RAW differs meaningfully from an uncompressed TIFF or TGA or BMP, and none of the sites I’ve looked at seem to address this.
Not sure whether I’m being dense or the difference is really subtle/arcane, but it doesn’t make intuitive sense that a RAW image could somehow let you adjust exposure post-facto in a way that differs from what you could do with standard formats, unless somehow the logic of the camera’s circuitry was itself embedded in the image, and the software could actually “recompile” the image from different parameters.
Just expanding on that thought, it seems like if you encode the exposure setting, the white balance setting, the flash intensity, etc. into the EXIF data, which is already possible with TIFF, you’d have all the metadata you need to come as close as possible to being able to emulate the initial conditions under which the picture was taken.
Scot, here’s a link to an excellent white paper entitled “Understanding Digital Raw Capture” by imaging/Photoshop pundit Bruce Fraser, found on Adobe’s website:
http://www.adobe.com/digitalimag/pdfs/understanding_digitalrawcapture.pdf
It’s 4 pages, written in laymen’s English, with illustrations, and explains how a RAW capture is different from a plain old image file format.
As soon as I make a change to an image in photoshop, whether it’s a tiff, jpeg, psd, etc, I am changing the information in the image file and, most likely, discarding certain information that might be useful to me at a later date (adjustment layers can help minimize this, but that’s for another post). By shooting with RAW and doing some adjustments using a RAW converter, I am giving Photoshop this best information I can, reducing the need to toss out information with photoshop adjustments.
This is not to say that one shouldn’t do any image adjustments in photoshop. It’s just that, like any piece of media software, the program is a good as the image/information I give it. RAW helps me give it “good” information.
Beyond adjusting the white balance, I find shooting RAW helps especially when a picture has been under or over exposed. When I shoot with negative film there’s quite a bit of dynamic range (the range of how dark and light I can make the image, sort of) compared to, say, chrome (slide) film. By being able to control some exposure issues with greater range after the picture has been shot, RAW allowes for an increased dynamic range compared to jpegs and tiffs.