The camera is Kate’s new Samsung NV15. She wanted something pocketable with good manual controls. Out of all the shiny digicams at the big box store, the Samsungs seem to be the only ones with an interface that makes sense. Instead of the normal profusion of dials, four-way pads, and select buttons, the NV15 uses one top dial to select the mode - manual, auto, etc - and array of touch sensitive bumps along the sides of the LCD. It’s kind of like a touch screen interface, just with the buttons off the screen, if that makes sense. In any event, it’s a fast and intuitive interface because nothing gets buried under layers of menus. The image quality isn’t half bad either.

Nature’s histogram?

A couple of weeks a ago I bought a copy of ColorNeg. Colin has written extensively about using ColorNeg with VueScan’s raw scans, so I thought I’d give it a shot. The trial version of the plugin is pretty much worthless as it obscures the results with a heavy grid of noise, but I plumped down my sixty bucks anyway. I was quickly disappointed. Although ColorNeg worked wonders with some scans, with others it turned out flat, muddy junk. ColorNeg’s manual suggest that non-linear scans are to blame for poor results. I’d been scanning in 16 bit linear mode on my Scan Dual IV, but I thought it was possible that the scans weren’t truly linear. That ‘linear’ scans using the negative and positive settings didn’t look exactly the same lent some credence to this hypothesis. In hopes of getting truly linear scans, I plunked down another $80 for VueScan Pro. Ah, now that works.

Has Kate seen the light too?

Call me a convert. After years of dismissing VueScan as irrelevant at best, I’ve discovered what many already knew; VueScan’s raw scanning feature is the bomb for B&W. If you let VueScan do the inversions, it’s no better than any other scanner software, but the raw scans are something else entirely. The raw Tiffs are all bunched up in the highlights, but they expand nicely, much more so than simple positive scans. I suspect that this means I will no longer have to result to underdevelopment to protect the highlights in my negs, which should help shadow detail, something that suffered with my previous technique. A little extra PS work is required, and there is room for error here, but in the end, the raw scans make scanning much more predictable. In some ways, it’s more like working in a darkroom; I can tell by looking at the negative how I should work it in PS, something which wasn’t always possible before.

After a night of celebrating the glories of VueScan

I should note at this point that I still haven’t had much luck with ColorNeg. It does work, but not significantly better than doing the inversions and adjustments on my own. It does seem to work better with TriX and HP5 than with Delta 400, which makes me suspect that its gamma is geared towards working with films with a more traditional curve than the straight line T grain films. I haven’t played around with it enough to be sure.

Filed under Doh!

For the source of this quote, see this auspiciousdragon post about the new Zeiss Sonnar in S mount. This rumor about a new Nikon S mount digital RF is probably just that, a rumor, but it’s just outrageous enough to be believable. Have Nikon and Zeiss built a relationship around the ZF lenses (or is the F mount public domain)? Are Zeiss, Cosina and Nikon working together to bring out a full frame digital Zeiss Ikon or updated digital Nikon SP rangefinder? Or is some forum nut laughing his but off right now? Magic eight ball says “answer unclear.” Ok, actually, the first time around it didn’t understand the question. The second time around, after I capitalized and punctuated the question, it answered yes. I’m not sure I trust oracular pronouncements from something that needs punctuation in order to understand the question.

Picture number 03580035?

Marek asks “have you already made more than three and a half million pictures in your life Matt?” Er, well no, not exactly. Here’s how the serial numbers work. The first four digits are the roll number. The last four digits are the frame number. Any letters indicate that there are multiple workings of that particular frame. So picture 03520008b (Thinking Wide), is the second (b) version of frame 8 of roll 352. Why 8 digits? Room for growth. Who knows, film may one day come in 9999 frame rolls ;-) I’ll wish I had gone for 16 digits when I break 10,000 rolls of film, but until then the system works.

I started using this numbering scheme just a few years back, so I actually have shot more than 382 rolls of film (scanning 383 as we speak). Really, I have. For each frame from each of those 383 rolls, there are a number of resulting files. 16 bit 3200 DPI TIFFS of the original scans get archived off to external drives. Full resolution JPEGS of every frame go into a folder on my computer for back up and quick reference. Thumbnails, the 740px wide images that end up on this site, of the edited frames go into an Index folder divided up into 50 roll sub-folders (Roll 1-50, Roll 51-100, etc). These Index folders are easy to eyeball for just the frame I’m looking for, and, since they just contain 100kb or less jpegs, they browse quickly even on my crappy computer, much faster than browsing in Photoshop for example. For each frame, the file number stays the same through all the different file sizes and types. So for frame 03520008 there’s 25mb 16bit TIFF (03520008.tiff), the PSD file I print from (03520008.psd), the full size JPEG (03520008.jpg) in the JPEGS folder, the small web sized JPEG in the Index folder (also 03520008.jpg), and the alternate version presented in the previous post (03520008b.jpg). It is not necessarily efficient for disk space concerns, but it makes things easy to find.

If the sign says it, it must be true. This is a “lost” frame discovered while I was doing some backups; I could have sworn that I posted it before, but I couldn’t find the post.

Not through any design, other than sheer laziness, I tend to do my backups in cycles. When I start running out of room on the internal drive(which keeps happening ever more frequently with all those Battlestar Gallactica episodes clogging up my iTunes directory), I make a compressed copy (jpgs) of all the scans and archive the originals (tiffs) off to an external drive. When I get time, I go back and burn dvd’s of the tiff and offsite the jpgs to my webiste. In this process I usually find a couple of good frames that got missed during the first round of editing. In this case, so far I’ve got the frame above (wouldn’t that look great in my office?), a neat picture of some shoes dangling from a power line, and a sort of American Gothic meets the 21st century portrait of Kate and me.

There’s not really a moral to that story. Just explaining why the image numbers suddenly jumped back 30 rolls.

OK, fine, you want a moral. Keep everything. Do your backups. And revisit your old stuff occasionaly.

The Sigma 30 F1.4 is just too sharp for portraits particularly with the D80’s 10MP backing it up. That’s just not a complaint you hear very often, and it’s testament to just how far Sigma has come from it’s days as a low-budget, low-performance lens maker. Sometimes, too sharp really is too sharp. Casual portraits taken with the Sigma require a bit more work with the healing brush and clone stamp and a little lighter touch on the sharpening.


Nikon D80, Sigma 30 F1.4

Portrait photographers have long known that too sharp can be a bad thing. Occasionaly they’ve gone too far; the softfocus wedding portrait is often a horrid, misty mess that goes well behind obscuring that little blemishes we all have. But there is a place for lenses that render just a bit less clinical detail. The 50 Hexanon is a good example of such a lens. Up close and wide open, fine details just melt away. Looking at a print, you wouldn’t say that it wasn’t sharp, but you also wouldn’t be counting individual eyelashes.


Hexar RF, 50 Hexanon

Undoubtedly, some of this softness is the result of some of the limitations inherent in film and it’s resulting workflow, which brings up the question, “Is digital too sharp for portraits?” Probably not, but it might require some changes in technique. A softar filter might be worth investing in, or learning the appropriate use of photoshop’s blur tools. Or it might be as simple as letting the focus go a little off or using a longer shutter speed to induce a little camerashake. Sacrilege, I know, but not every picture requires every detail to be render with such aching fidelity.

I don’t seem to know how to resize the right way. I often end up with jaggies and other artefacts around details. Take a look at this pic for what I’m talking about. Note the artefacts around the text.

So what’s the secret here? A magic $20 plugin? Selective blurring and sharpening at various stages. Not that long ago I started downsizing in steps using a PS action. It doesn’t seem to make that much difference. Kate’s old glasses seemed to present a particular challenge in this arena nearly always rendering with some jaggies at one particularly part of the curve. See below.



I know the root of the problem has something to do with having sufficient pixels to render curves or lines of particular angles. But am I approaching that threshold that often? And if I am, what am I doing differently from all the other photographers that don’t seem to have this problem as often? This isn’t the end of the world, but it’s kind of annoying. I often find myself backing off the sharpening to reduce the apparent jaggies. I’m asking too much?

A friend pointed out to me that Canon is running its double rebate deal for the year. I’m not particularly in the market for anything, but I took a look out of curiousity. For the most part, it’s chump change; thirty bucks off a $380 flash is hardly worth the paper work. However, that rebate on the 5D is a sizeable $300. Double it by buying a lens, like a 50 f.14, and you get $600 rebate on the body and a $40 rebate on the lens. With the 5D currently selling for $2800 and the fifty for $315, the rebates can get you both for less than $2500, with shipping! Yeah, yeah, the 5D is still really just a $400-$500 camera with a $2000 digital supplement, but $2500 for the body and a lens that won’t suck is pretty spiffy in the digital realm.

If I read the right forums, I bet I could find a number of people wondering if these rebates mean that Canon is clearing out stock in anticipation of releasing a 6D this winter. Note that even though the 5D is only twice the price of a 30D , the 5D gets triple the rebate of the 30D. Perhaps a 40D is not imminent. Perhaps Canon has got a $2000 6D waiting in the wings. Perhaps.

Surely Canon must refresh the 30D this winter if they don’t want sales of the 10MP Rebel to kill off their mid-level DSLR. Or maybe they are abandoning the cropped frame mid-level DSLR market. Perhaps they think the new Rebel has got this market just about wrapped up. They may be right about that. The new Rebel looks pretty capable. It’s buffer is almost as big as the 30D’s - take into account the 400D’s larger files size, and its buffer may actually be bigger in MB than the 30D’s. The AF and metering systems look to be the same, although the 400D doesn’t get a spot meter. The 400D gets Kelvin WB settings, which the 30D doesn’t have. You do get significantly more info in the VF with the 30D.

It doesn’t seem like Canon has left a lot out of the 400D. Apart from the added VF info and a spot meter, I’m not sure what the alleged 40D would have to offer. Maybe a real mirror lock up? Canon users always seem bent up on that issue. What could they add to a 40D that would justify spending $500 more on it? Nothing.

Which brings us back around to my orignial point. The 40D is dead in the water. Canon thinks they can wrap up the midlevel with the 400D. The 6D willl target more apsirational users with a price point probably under two grand and some minimal updates to the 5D’s feature set. I’m betting it gets some sensor tweaks and a higher frame rate. If they pull it off, expect to see prices on the D200 drop to $1300.

I’ve been working on assembling my first real portfolio in a long time. I had a portfolio back in highschool when I was spending twenty or thirty hours a week doing newspaper and yearbook photography. That portfolio is long gone, which is probably for the best, so I’ve been working on assembling a new one. Originally this was going to be part of my Salt application, but I’ve put that on the back burner for a while.

I’m kind of taking a shot in the dark on how to do this, but I’m betting that showing work in a single theme is more representative in some ways than showing a whole bunch of different things, so I settled on showing class room photos from my six months in Korea. Choosing from the hundred or so available proved to be the most difficult part of the task, until I started asking for feedback. Ask ten people to pick out the weakest photo in a set and you’ll be lucky to get the same answer twice unless you’ve included some incredibly awful shots. For example the shot below has already been identifying as both the strongest and the weakest shot in the group.



Keeping that in mind, I’m soliciting comments from a number of different online communities. You can see the rough edit and make comments over on Flickr. I’ve also embedded the slideshow in the photo essays section of this site.

Embedding the slideshow was surprisingly trivial. I haven’t used Flickr much, but I’ve been toying around with it as a place to put things that might be kind of transitory. The overhead is low in comparison to creating galleries on my site, and the ability to embed slideshows gives me the best of both worlds. It took a little poking around google to figure out how to embed a slideshow based on a pre-defined set, but I eventually found the syntax for it here. To build an embedded slideshow based on tags or group/user id, check out this nifty tool.

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:

• 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?