I’m a little confused by the luke warm response to the new Fuji S5. For those that hadn’t heard the news, here’s the skinny:
As depreview already noted in their S3 Pro review, the extendended dynamic range function of Fuji’s sensor really works. Furthermore, the S3 Pro had “Good resolution, better than the average six megapixel”. The S5 promises all that and the improved AF, metering, ergonomics and durability of the D200. What’s not to like?
- D200 body, AF and metering
- Same extended range sensor as the S3 Pro with improvements to the low pass filter
As depreview already noted in their S3 Pro review, the extendended dynamic range function of Fuji’s sensor really works. Furthermore, the S3 Pro had “Good resolution, better than the average six megapixel”. The S5 promises all that and the improved AF, metering, ergonomics and durability of the D200. What’s not to like?
- A certain amount of noise can improve the performance of non-linear systems. This sounds like nonsense, but it’s not when you stop to think about it. For example, B&W photographers have long known that a grainier image can often look sharper than a grainless one. If you are interested in the science behind it, this guy has written some articles about noise that are entirely over my head. His book, which was discussed on Science Friday recently, sounds like its more my speed. He also has some challenging thoughts on the diminishing importance of the comma.
- Although a lot photographers seem to be pursuing noiselessness, noiseless photos fail to satisfy me on some deeper level. I can appreciate the technical mastery, or least the amount of money, involved in creating a noiseless image, but I don’t find them particularly interesting. I’m not the only one, but I’d argue that as a group, photographers don’t know nearly enough about the roll noise plays in the success or failure of a photograph. For example, can you to a certain extent hide blown highlights by adding the right amount and kind of noise?
- I’m pretty happy with the noise qualities of the D80. You can check out a whole bunch of test shots and comparisons over at Dpreview. I haven’t used the D80 at 1600 all that often yet, but I’m pleased with the results so far.
ISO 1600
Crop of ISO 1600 image
Great B&W work lies nearer alchemy than science. The arcana of developing (time, temp, solution, dilution, agitation) can take a lifetime to master (although passable results are as simple as following the directions on the box). Small changes in any of the many variables can cause wildly diverging results. And that’s before you get to printing or scanning (and then printing digitally, which is its own alchemy). For a certain kind of mind - inquisitive, stuborn, tedious, etc - it’s all damn good fun.
One would think that it might be easier to just take digital files and convert them to B&W. In fact, many cameras include modes that do this in camera including the D80. When you know what B&W can look like, these modes just don’t satisfy. Neither do many of the more ‘advanced’ methods of doing these conversions in post. There’s a least three ways to do it in photoshop (channel mixer, plain old desaturate, switching to lab color and splitting the channels). There are probably other ways as well. And then there’s the various stand alone and plugin apps that promise either better or easier to achieve results. Lightzone seems to be one of the newest and the most promising.
Getting the most out of these tools require a lot of not easily acquired information. If you don’t know the spectral response and density curves of the B&W film you are emulating or how to put that information to use, the results are rather hit or miss. Along with all that take into account that digital sensors respond to light completely differently than B&W; how does one map a linear response onto a curve with big shoulder and toe regions? How do you get shadow densities comparable to B&W without pushing your highlights all out of wack. This is hard enough on B&W film, which already has an acommodating curve. On digital it seems near impossible to do with any regularity.
For all that, there are some people out there doing incredible B&W work on digital. It’s apparently not impossible, but I sure don’t understand how to do it, yet. See the pic below. Bringing up shadows moved the highlights up to 255. Even ignoring the highlights, the noise in the shadows had already taken on offensive look. Blah.
See the photo blog for a color version.
One would think that it might be easier to just take digital files and convert them to B&W. In fact, many cameras include modes that do this in camera including the D80. When you know what B&W can look like, these modes just don’t satisfy. Neither do many of the more ‘advanced’ methods of doing these conversions in post. There’s a least three ways to do it in photoshop (channel mixer, plain old desaturate, switching to lab color and splitting the channels). There are probably other ways as well. And then there’s the various stand alone and plugin apps that promise either better or easier to achieve results. Lightzone seems to be one of the newest and the most promising.
Getting the most out of these tools require a lot of not easily acquired information. If you don’t know the spectral response and density curves of the B&W film you are emulating or how to put that information to use, the results are rather hit or miss. Along with all that take into account that digital sensors respond to light completely differently than B&W; how does one map a linear response onto a curve with big shoulder and toe regions? How do you get shadow densities comparable to B&W without pushing your highlights all out of wack. This is hard enough on B&W film, which already has an acommodating curve. On digital it seems near impossible to do with any regularity.
For all that, there are some people out there doing incredible B&W work on digital. It’s apparently not impossible, but I sure don’t understand how to do it, yet. See the pic below. Bringing up shadows moved the highlights up to 255. Even ignoring the highlights, the noise in the shadows had already taken on offensive look. Blah.
See the photo blog for a color version.
PTLens seems to work just dandy. It has a predefined setting for the Sigma 30 f1.4 - and a bunch of other lenses as well - which seems to correct the barell distortion quite well. And it slipped right into the batch actions I normally use. See the corrected photo below.
Anybody know a good tool for correcting barell distortion? One that doesn’t cost an arm and a leg? I dig this Sigma 30 F1.4, but it kind of gives things the bends.
Take it as hyperbole if you like, but I see life through a camera’s viewfinder, and I’m sure I’m not the only one. We all have ways of structuring our reality. Some folks have religion. Some have Nascar. Others barbeque on their back deck every weekend. The viewfinder is my way, so it’s really dissapointing that they keep building these digital cameras with such crummy viewfinders.
The D80 is supposed to have a pretty good VF for a cropped frame DSLR, and from having used the competition, I’d say the D80 is pretty much the state of the art. But that’s not saying very much. Even the full frame 5D has a VF that would have never made it into a basic manual focus SLR from the late 70’s or early 80’s. Those were pretty good cameras in a lot of ways. Nice big viewfinders. Although not as big as something like an F2, they were pretty good. They still are pretty good. If you haven’t ever picked up a Canon AE-1, a Minolta X-something, a Nikon FE or best of all an Olympus OM-1, don’t do it. You’ll hate your DSLR viewfinder. Pick up a good rangefinder, and you’ll be seriously depressed.
How much difference does this make? It’s hard to say. On a psychological level, this is pretty big hurdle to get over. I’m used to the disconnect between vision and print coming at a much later stage of the photographic process. With film, if the end result doesn’t match my vision, it’s usually due to bad processing. It’s not something I’m aware of at the moment of taking the photo. With digital, the bad viewfinder puts the disconect right at the moment of photographing. Making the most of that disconnect is going to be the most challenging aspect of digital for me. The rest of the chain is cake by comparison.
The D80 is supposed to have a pretty good VF for a cropped frame DSLR, and from having used the competition, I’d say the D80 is pretty much the state of the art. But that’s not saying very much. Even the full frame 5D has a VF that would have never made it into a basic manual focus SLR from the late 70’s or early 80’s. Those were pretty good cameras in a lot of ways. Nice big viewfinders. Although not as big as something like an F2, they were pretty good. They still are pretty good. If you haven’t ever picked up a Canon AE-1, a Minolta X-something, a Nikon FE or best of all an Olympus OM-1, don’t do it. You’ll hate your DSLR viewfinder. Pick up a good rangefinder, and you’ll be seriously depressed.
How much difference does this make? It’s hard to say. On a psychological level, this is pretty big hurdle to get over. I’m used to the disconnect between vision and print coming at a much later stage of the photographic process. With film, if the end result doesn’t match my vision, it’s usually due to bad processing. It’s not something I’m aware of at the moment of taking the photo. With digital, the bad viewfinder puts the disconect right at the moment of photographing. Making the most of that disconnect is going to be the most challenging aspect of digital for me. The rest of the chain is cake by comparison.
I’m terrible at color correction. I just don’t have an eye for it. I’m not color blind, but I just really can’t eyeball a photo and fix the color with any degree of accuracy. Recently I stumbled across Chris Nicholson’s “Even a color-blind person can color-correct a photo.” Sounds perfect, and actually it is perfectly correct color, although the method is a little slow. Chris outlines a method for using PS’s eyedropper tool, info palette, and curves do color correction by balancing the numerical inputs and outputs for Red, Green and Blue. It sounds more complicated than it is, and with a little practice it gets to be pretty intuitive. If you’ve struggled with color correction, give the article a look:
Uncorrected image, as shot
Image “corrected” via PS auto color
Corrected via Crhis’s method. See a larger version of this image on my photo blog
If you follow Leica stuff at all, you know by now that Leica’s first digital RF, the M8, is now real as in you can look at pictures and video of it on the internet. I’d love an M8, two actually. While the price tag ($5000) is pretty reasonable compared to other pro digicams, it’s way out of my league given other goals I have at the moment - like getting into and paying for Salt next spring.
So what’s the point of this unique wedding of old fashioned RF technology and brand spanking new digital hotness? Having used the D80 for a couple of days now, I think I can enumerate a few points where an RF is still superior:
So what’s the point of this unique wedding of old fashioned RF technology and brand spanking new digital hotness? Having used the D80 for a couple of days now, I think I can enumerate a few points where an RF is still superior:
- Size: while the D80 is pretty small for a DSLR, it’s nothing like an RF body. When you get to lenses it’s game over. The RF allows for considerably smaller lenses.
- Speed of operation: I’ve now used DSLR’s from all the major manufacturers, so I think I can make a pretty broad statement here. In comparison to an all manual camera, DSLR ergonomics suck. Aperture rings on the lens, shutterspeeds on a thumb actuated dial ala Hexar RF, and a highly refined manual focussing system whoop the pants off Multi-badass 10000 AF modules, command and sub-command dials, reprogrammable buttons etc etc etc.
- Lenses: you can do cool things when you don’t have a design around a big flapping mirror.
- No VF blackout: being able to see your subject at the actual moment of exposure gives the RF user a better sense of timing and the ability to handhold at slower shutter speeds.
I’d also take one of Michael Johnson’s DMDs. Think Hexar AF with a digital sensor, built in A&S and a few other goodies. Heck, I’d even pay $1000 for one. I will not however, pay any amount for a Richoh GRD, no matter how many posts Mitch makes about it on photo.net. Mitch seems like a pretty decent guy, and the GRD a pretty decent camera, but 28mm lenses just don’t float my boat, and I know that I require something a little faster to keep my frustration level at the right level for photographic production.
My expectations for backup solutions are shaped by spending five years doing IT work. Consequently, a lot the solutions out there for home use seem either inadequate or just plain clunky. I know a lot of photogs use disk arrays of one sort or another, either RAID solutions or the poor man’s solution of just a bunch of external drives. Some folks incorporate DVD backups and even off site storage, but this usually amounts to giving a stack of dvds to the guy across the street or a friend in another state. Although this protects you from a house fire or similar disaster, it’s kind of clumsy and slows access to your backed up data. And lets face it; digital has already turned photographers into lab technicians. Do you really want to become a system admin too?
In the enterprise world, data storage and back up now often involves network attached storage - multiple big drive arrays often geographically dispersed and hooked up via high speed network - with a secondary backup of tapes off site. The combination of the two provides you with quick access to backups that are online but offsite as well as distant offsite storage on a completely seperate media. Redundancy and quick access. Network attached storage, or NAS for short, is what I’d like to see come to the home user particuarly internet attached storage. There’s lot of ways to hook all that RAID or even all those external drives up to the home network, which is all well and good for quick access, but it doesn’t protect you from a local disaster. That’s where the internetworked attached storage comes into play.
There’s a lot of companies offering online storage, but the cost per gig isn’t reasonable for photographic uses. See this NYT article for providers and prices: http://www.nytimes.com/2006/09/07/technology/07basics.html. Apple for example charge $100 a year for a gig of storage. A much cheaper solution is usually available from website hosting companies. I pay godaddy.com $75 a year to host this site and a few others. That gets me 100 gb of disk space, which is way more than I need for hosting this site, and enough transfer to handle all the site requests and access to my backup files. 100 gb isn’t enough for you to backup all your raw files, but it is enough to archives full res low compression jpegs of the selects.
That 100gb of storage is always available no matter where I go in the world. In the last week I’ve accessed it from my apartment in Korea, over a wireless network in the Osaka airport and from the pseudo-inlaws house in Iowa. And the hosting company takes care of migrating all that data to new storage media when the old stuff goes belly up. I don’t have to worry about lubricants drying up in a seldom used external drive or about improperly stored DVD’s having data faults. I also don’t need to worry about finding a computer with the right interface to access some obsolete storage media. Ever wondered how you are going to get at your old ATA drives when they stop building ATA controllers? SATA and USB will probably be gone in 10 years. But the network won’t.
The only problem? Getting the data there. If you haven’t ever noticed, your dsl or cable connection caps your upload speed at a much lower rate than you get for download. That means uploading 10gb of image files can take days instead of hours. So don’t fall behind. Luckily download speeds are much better.