Portfolio overhaul

With all the recent discussion on style I realised that my portfolio site was a mish-mash of seemingly random pictures: spot the amateur. Well, I've decided to overhaul the whole structure into more consistent groups and added a bunch of photos into the bargain. Now, hopefully, some of the themes that I follow can clearly be seen in context. Not quite the obsessive pursuit of an idea but much better reflects what I'm trying to get across.

I hope you like the changes.

First roll of black and white

Crane, Canary Wharf, London, July 2007

In all the years I've had cameras and taken photos I've not, until recently, actually used B&W film. Mostly, I suppose, because I was part of the happy-snapper crowd.
Well I finally got round to turning the roll I shot this summer into final images. I was using Ilford HP5 plus 400 so that I had plenty of scope for flash-free photography. I loaded it into my old SLR with a 50mm lens.

2 things surprised me about the outcome - first is that I can take decent photos with black and white. I was consciously thinking of shapes and forms. For the people shots, I'd probably have converted colour to B&W anyway. The second surprising thing was the distance to subject that worked best. normally with a 50mm I think middle distances, say 10-100m (first of the examples below). With this film, however, it was the close-up that worked best. Maybe a result of detail overcoming grain, but I can't really be certain. Anyway, here are some of the results:

Getting through the tedious tasks

Having largely finished all the outstanding image work, I finally got round to all those small things yesterday: backing-up, software updates, archiving etc. I spend far too little time keeping my system organised when I'm busy working on the pictures which is going to get me in trouble one of these days. For example, I had over 500 RAW and 1000 finished pictures to back-up. Took hours for it to get completed.

Maybe I should have less on the go at once, but that would mean taking less photos.

Next up, the scanning: I haven't even got around to scanning the dozen 4x5s and 10 rolls of 120 that I've got ready to go.

So much to do and so little inclination to do it.

Style, empathy, obsession

Barrier Lake, Alberta, September 2007

There has been discussion of late on several blogs about the nature of style, how style defines great photographers and how style is developed. Probably the best discussions have been over at the Landscapist (start here and work forward to today). Much mention has been made of the obsession of great photographers to constantly go out and create images in a particular style, often of particular subjects.

In a comment to this post, I disagreed with the obsession idea, I think it is more a matter of empathy that then feeds the development of a subject and style.

I've been experimenting with the links between style and subject matter. The best landscape work I've done (for me at least) I've produced in a relatively high-contrast and low saturation manner in cool colours. It seems to suit the sort of open, day lit subjects I shoot. It doesn't, however, seem to fit all subjects.

The image with this post is one of a number of processing experiments I've tried the last few days. The original is below. This particular frame is badly exposed: there was far to much dynamic range to get an effective capture - the sun falling behind the mountains was still to high yet for the sky to have darkened yet the mountains are fully in shadow.


The shot was never about colour but for, so a black and white treatment was appropriate. When I did the initial conversion it looked a lot like old photographs of similar subjects, hence the treatment I tried above. I've slightly overdone the effect but it served the purpose: this style seemed to suit the subject and certainly worked given the technical deficiencies of the original.

Other experiments I've tried also seem to suggest that the right technical style can make a given photo and that most subjects have a style that suits. If, however, one does not connect with either subject or style, then one is never going to create a great body of work from them. Once a connection is made, then I think the two aspects can feed off one another: the continuing work on the subject and the development of the appropriate style.

This, maybe is where most of us fall down - we're too busy trying to make each and every image a good one, fitting the styles as suit the subject rather than focusing on those for which we have the greatest affinity.

Changes to the look

If you've been here before, you'll notice some changes to the look of my blog. It's all an attempt to get more on the screen. I finally took the time to update the template to take a smaller sidebar font and tweaked the line spacing & stuff. Let me know if it's illegible but as most visitors have relatively large screens (or at least hi-res) then I'm guessing this should help.

I've also added a new website link to Weeping Ash. Got this via Doug Stockdale's blog. It's a really nice site and lots of info on stuff you don't find elsewhere, including English photographers (especially Edwin Smith).

Lightzone 3.1 - a more detailed look

WARNING: this is a looong post (and taken a week to put together).

Following my earlier comments on Lightzone 3.1, I've taken a closer look at the new version.
A bit of preamble first: these are my opinions based on using the tool and my deductions as to things going on. I've been a user of Lightzone since about v1.2, and currently use 2.4 as probably my main editor. This is really an upgraders view, rather than a ground-up analysis. I'm focusing on differences. I'm exclusively a windows user, this is based around my home XP Pro set-up (dual core 3GHZ, 2GB RAM).

I'm trying to be as objective as possible & will give reasoning as to my opinions where appropriate. I do a lot of software testing as part of my regular job and work with both development and user communities in that role. Many of my opinions come from that experience.

The good

There are things that warrant merit in the new version:

Relight tool - the big new addition. Compared to the ToneMapper, there is much improved shadow lifting together with better highlight recovery. This is a really powerful tool and can, in some instances, replace bracketed and HDR blending. It also adds something like hiraloam sharpening for contrast, too. Thus, tone compression & mid tone contrast in one.
Colour masking - a very powerful control to limit the tonal or colour range over which a tool is applied. Just one early use one of my favourite ways to use this is for colourising monochrome by letting the highlights run to white and the shadows to black. Also has a feathering function for the tonal range applied so the results blend in nicely. So much better than the ill-defined shadow/midtone blending options.
Zoom buttons - a small thing but it's nice to have the option. it seems that zoom-fit now also a applies a small buffer around the image rather than butting it to the side panels. Much nicer for viewing.
Variable panel size - enables the Zone map/histo window to be enlarged. Also means I can set the main window to zoom-fit to actual print size for regular SLR shots which is nice. It's also nice to be able to collapse the panels.
Tool displays now have some nice indicators: a marker to show if there are regions applied, little "light-ups" to show if the blending mode or colour mapper has been changed. Makes for easier tracking of what's been done without clicking through all the tools.
Return of the progress bar so I can actually tell if it's saving/exporting etc. In 2.4 it was a bit hit and hope.
Zoom drag - a little icon in the corner highlights the zoomed portion and can be quickly dragged around. Saves the interminable scroll bar work.

The bad

As I'd mentioned before, Lightzone also has a bunch of problems. here I'm going to focus on lost or missing stuff.

Rotate - from powerful, exact tool to arbitrary and imprecise. Moved from one of the best rotation implementations using a wedge to precisely align verticals/horizontals to something almost completely unusable.
Crop - incorporating rotation. Unnecessary and means I'm rotating when I want to crop. There are also all kinds of arbitrary limits on when it will and will not change the size. Seems unable to determine in which direction it's free to move.
File renaming - when exporting (batch and single) my index numbers are changed instead of appended. If I add extras, they get dropped in batch mode.
Loss of XML sidecars - I've left this to last deliberately - for me, at present, it is a deal breaker. I don't use Lightroom or other DAM software. I've got a perfectly good, cross-referenced filing system. One thing that the XML gave me the ability to do was to change file references, necessary when moving files around or renaming them. I'd be happy if such info was in the sidecar metadata but it's not. That means a lot of manual re-tagging in Lightzone, which is tedious and slow. Right now it's a big issue for me because I need to re-index a huge number of files but I can't be the only one who moves & renames batched of files.

And the ugly

There are also quite a number of areas where the UI falls down and thus spoils the overall user experience. This should not be underestimated. In my experience, software users will become non-users if they are frustrated by the interface or usability. It should also work as they expect relative to their OS and other programs.
Documentation: what documentation? It's rubbish, quite frankly. With these tools being unique to Lightzone, I want a decent technical reference to explain what they're doing so I can better use them. Whilst getting going is easy, really getting the most out does require quite a lot of trial and error. There seems to be a policy of leaving this to the online community - great if it's open source or shareware: that's part of the deal there. Plus, that sort of things happens anyway (just look at the industry surrounding Photoshop). This, however, is commercial software at $250 a pop - I expect the supplier to be properly documenting the product at this price. Lightcraft's attitude just appears lazy and inconsiderate.
Large files - Lightzone balks at anything much more than a single SLR RAW file. Not exactly putting it in the "essential for photographers" branch they claim. What about my stitched images, or scans from film? If I've got 1.5GB of free memory, a 300-400MB file should be quite easy to handle. Every other bit of imaging software I've got (and there are a few) has no problem (example - using PTGui this week, it finally balked at a 150MP, 16-bit TIFF: but that would be 7GB, no problems at 20-30MP). Telling me to add hardware is not helpful. If other programs manage, so should this one.
Auto-renaming, where Lightzone decides it knows what file names I want to use. I use an indexing system that ends -xxx.ext, where the x's are numbers. If I have a number -001, then Lightzone decides I want that to be -1 on conversion (or even _lzn-1). If I've added something (say "-218 BW" for my monochrome version), it'll happily batch convert that to -218 (or even -219 if a -218 already exists). I can control this for individual files, for batches I cannot. This then means I have to either do several batches or manually rename a whole lot of files.
The there is the batch converter itself - not much improved from before. Slightly faster but still the selector slows right down after picking about 10-12 images (because it's busy loading stuff into memory, more of later). I wish I could just pick all the lzn files for conversion but I can't. So I have to do several batches. And batching doesn't run in the background, so I'm stuck while it's converting.
Lightzone has the worst memory management of any program I have ever used. It's the only one I've used that regularly (as in every time I load it) uses all resources available. I load it up and it reads a whole lot of stuff into memory. Not that it helps in caching - switch to editor mode and you still have to wait an age for the image to load. I also tried opening a folder, then deleting all of the images (i.e. nothing left to need a cache). It was still using full allocation of memory. There are limits in how much memory LZ will access: set the limit in preferences which is split 50/50 between RAM and virtual. Single images appear to use many times more memory than required. There is no way to clear the undo history cache to free-up memory. All out, this is rubbish. I've done like-for-like memory use comparisons with Photoshop and Lightzone is using about 5 times the memory and doesn't free-up unused memory properly either.

Concluson

From all this you may think I hate Lightzone, and Lightcrafts as a company. Actually, I'm a fan of the software. It's easy to learn, powerful and for general editing gets me to results faster than anything else I've tried. I am, however, disappointed and frustrated. This is a tool with huge potential to meet the needs of serious & not-so-serious photographers alike yet the potential is being squandered with sloppy implementation and a poor user experience.
The tools are great, there is some real power there in an easy to use format. All the productivity I gain there, however, is lost in the management stuff: conversions, batch processing, large files, UI niggles. It's all about quality assurance and control. Effort spent on quality issues for software should be at least as much as on development. I'm not getting that feeling. This still feels more like a hobby development program than proper professional/commercial software. There are plenty of small UI items carrying across versions which is inexcusable - I've got shareware (heck, even freeware) that does better than this.

Of bicycles & cameras, cycling & photography

One of my fixed gear bicycles (I have 2)

My other great love is cycling, if you'd not noticed. I try and spend as many hours in the saddle as possible (although never as much as I'd like). My vacations are usually organised around doing something on 2 wheels. As i went out yesterday afternoon it struck me that cycling is a lot like photography and bicycles are a lot like cameras in a way.
Just as much gear envy and technical discussion occurs amongst cyclist about bikes as does talk of cameras amongst photographers. There is always something newer, better that is going to make one improve. I own a lot more bicycles than I do cameras (although each has an intended purpose).

I could go on for some time as to how a bicycle frame compares to a camera lens in the way it defines the characteristics of the results. I won't. there are also different types of bicycle suited to different tasks or personalities. The fixed gear I rode yesterday (pictured) is like a large format camera - the very simplest essence of the device demanding total commitment to the process (if the wheels go round so do the pedals, so you can't stop pedalling a fixie).

That all said, there is a truism in cycling: it's nothing to do with the bike, it's all about the engine. Give me a Cervelo P3 and I will not be beating Fabian Cancellara any time soon, just as a Leica will not make me Cartier-Bresson.

Change the digicam

Sometimes I wonder if the people making cameras actually observe those using them.

I've noticed 2 things about the average snapper and the way that small cameras are used:
1. Camera phone users are very comfortable taking quick snaps with one hand
2. Small camera users seem to take forever and rarely look comfortable using their cameras

This applies to the young and old alike. Cause? For me it's all in form factor (shape). Virtually all camera phone users hold the camera upright like a phone and use their thumb on the shutter release. Cameras, on the other hand, are all designed with an index finger trigger, making them awkward to use with one hand or at low angles.

So why isn't the digicam redesigned for the casual snapper? I noticed even the phone makers show their camera function product shots aligned in the traditional "landscape" format, even though no one uses them that way.

Here's a cobbled-together mock-up I've put together for such a camera. Notice also I've gone for a square aspect ratio: gets over the whole portrait/landscape decision which seems to cloud people's judgement on how to hold the thing. Simple controls, all in reach of the thumb.

Remember, you read it here first.

New pocket digicam mock-up

The attraction of a rangefinder

The photographer, Seville, November 2006

I've been thinking about the whole idea of a rangefinder camera for a while now. Various reasons for this, which I'll come onto in a moment.
If Leica were the only rangefinders around, it would be out of the question - far too expensive and not exactly perfect at that. There are, however, some real alternatives about: the Cosina made Bessa line and the new Zeiss Ikon. Whilst nobody could really call these cheap cameras at least they are affordable. Especially if you only want a lens or 2.
Colin Jago's recent words on the whole issue don't help me get over new kit syndrome here, either. Fortunately he posted a link to Big Bubbles in my comments to this post which was useful in tempering enthusiasm.

So why would I want (yet) another camera, especially an "old fashioned" thing like a rangefinder? It's for what they do best: street photography. Picture for this post is a perfect example. While I used a DSLR with a zoom lens, I manually zone focused, set to f/8 and shot with the camera on my lap. Never raised it to my eye at all. I doubt anyone knew I was there, certainly no one overtly noticed the camera. The photographer of the title spent ages tracking & zooming and fiddling. Doubt she ever got a shot in.
Street photography is, for me, all about the discrete. If I can carry a camera that doesn't get noticed and can be shot from the hip (literally), that's ideal. This is where the whole rangefinder thing comes in. Also quite handy for the casual sort of snapping. Some fast film, wideish lens (say 28 - 40mm) and you can cover a wide range of situations.

I did take a look at the new Canon G9 briefly (whole other camera but similar use for me) which looks like it could also be usable in these cases. I think I'll have to take it for a test drive. Pity I can't do the same with an RF, too.

Another couple of images from Alberta

Flowing clouds, Middle Lake, Alberta, September 2007

Driftwood and rocks, Alberta, September 2007

Too much of a rush

Last night, I was definitely rushing when i worked on the first image I posted: seeing it side-by-side with the others i realised it was not what I intended. So today I spent a bit more time and here is the result.

Middle Lake, Alberta, September 2007

I've realised that this cool, low saturation look is something I was producing early on with the digital stuff. it was a deliberate choice and reflected the way I wanted to present the world. I've been drifting further and further from it and been getting less results that I'm truly happy with. I'm going to have to overhaul my portfolio gallery (not quite yet, though) as a result.

Here's another one from the last Canada trip, with the new-old look. it also incorporates something else that I do, using selective saturation enhancement to highlight a given feature, which goes some way to mimicking the way the eye sees features in front as opposed to in the periphery.

Autumn colours of Kananaskis, Alberta, September 2007

Defining my style

I was thinking to myself on the tram home the other night: what defines my photographic style? Do I, in fact, have one? I was coming to the conclusion that no, I did not and that much of my work recently was getting a bit "average Joe" - derivative snapshot stuff. maybe not, maybe so. I'm pretty certain it hasn't been distinctive.

Then Mike Johnston posted on the self same subject, which prompted me to look afresh at some of my recent stuff. I also thought back to some of the work I'd done last year that i really liked. So here are some of the results of that thinking, several presented together so hopefully my ideas can be seen running through the set.

Middle Lake, Alberta, September 2007

Getting it in the frame

There was an interesting discussion over at the Landscapist on what & how to teach in photography. Led me to think: what is the process I use to make an image?

I see it in 3 steps:

1. Looking for the subject

2. Capturing the subject

3. Processing the image

First up, I spend quite a bit of thought time (although that may be short in clock time) looking around and at things. If I see something interesting, I figure out what interested me and then look for an angle to best get that interest to the picture. These days, I usually go with the immediate reaction to something to determine its interest. In the past I might have spent more time considering whether it would make a keeper, these days I'm happier to throw stuff out than I was then.

Second is the actual getting it to the capture medium. This involves the mechanics of the camera, setting up tripod, careful checking of angles for clashing elements and other such technical stuff. I've never really specifically regarded compositional devices such as rule of 3rds, leading lines etc. Generally, stuff just looks right through the viewfinder. Sometimes I'm looking to bring specific elements together in a specific way. Mostly, though, it is about getting the stuff I want into the frame.

Lastly there is the bit about getting the image out of the camera and to the print (I always think of print as the final destination, whether it is or not). Apart from all the colour balance, contrast, sharpening etc there is the important bit on cropping. George Barr covers this much better (e.g. here) but my approach is similar to his - check the edges the ensure the wanted stuff is in and the unwanted (e.g. distracting) is out. Once in a while I can't quite crop out all that I want so I've been known to do a little cloning at the edges to remove the last trace of intruding elements. I don't care if that's cheating.

There it is - no formal art training here, I do look at quite a lot of pictures to see what works for me or not but that's the limit. I end up happy with the results and that's what matters to me.

Not as it first appeared

Age and discovery, Alberta, September 2007

For me, an interesting image.

Normally, I'm pretty sure of what I'm photographing, by which I mean the intention I wish to convey. This was one of those that was not quite the same.

Once in a while, I'll shoot something interesting and then find something else entirely the attracts me to the result. In this case it was the interesting geometric shapes that first drew me to take the picture but then it became about something else entirely.

As ever, the real attraction of the Rockies for me is the demonstration of powerful natural forces: the upthrow of the strata and the glacial erosion.

For me, the caption says it all.

Darkness, metering & mathematics

Evening falls, Alberta, September 2007
This was a 1 minute expose & a lot of post-process. Could easily have gone to 4-5 minutes

During the last trip to Canada I was experimenting with long exposures (like the one above). here, I'm talking the stuff measured in minutes. Cliched stuff: mountains, lakes, setting sun etc. Quite a few were with the 20D but I'm not a fan of digital long exposures: too much shadow noise.

Having my (relatively new) medium format with me, I tried some film-based long exposure, too. Should be better, no noise to contend with for one. Any colour cast can be dealt with by the software. trouble is, it requires metering in tricky conditions and offsetting for middle grey and reciprocity (this last is pretty minimal with modern film). This is where my rubbish mental arithmetic fails me. I keep adjusting in the wrong direction or by too few stops. You'd never think that I deal in numbers & advanced maths all day.

I've just picked up the developed film from the shop and so the upshot is the image here (which is pretty rubbish - certainly compared to what I had in mind) was the best I got. More work needed: or a meter that does all the calcs for me!

The feedback on Lightzone 3.1

Bridge over Le Sereine, France, June 2007

Here, I was playing around with softening effects on an otherwise dull image. I'm looking for ways to add particular feelings to photos.

Colin Jago, of auspiciousdragon, posted the following comment on my thoughts on Lightzone 3.1:

"I'm still using Lightzone 2.3 (I hit a bug in 2.4 and trashed the installation). It has always frustrated me. It has loads of potential but the guiding hand behind the product seems out of touch."

I must say, I sympathise. I've found more than 1 bug in 2.4 but felt the tool improvements were worth living with them - just.

From my original observations I contacted Lightcrafts about the new lzn.jpg files. What I wanted to know was whether there was a way of editing the info in them as per the XML .lzn files of old. (When the jpg sidecar is opened in a text editor all the old XML data is visible, just not editable.)
Their response: "sorry, you can't create lzn files from 3.1" - an answer to a question never posed.
I replied as much and asked if there were plans to return the lzn files - that was a week ago, no response since.

Not only are the development team wandering in the wilderness, the support team has upped and joined them.

When I first signed up the LZ, response was quick and effective. Just like the product, I want the old stuff back, please.

Is the effort worth it all?

Beaver's domain, Alberta, September 2007
I'm sure the web version doesn't do the subtlety of such an image justice

So this is one of my favourite images from my last Canada trip. Simple yet complex. it displays a whole range of issues associated with my feelings coming from the early morning exploration and contemplation.

thing is it was a single capture - 1 shot, 1 view. No fancy preparation and no experimenting with exposure and framing. no search for the perfect image. yet no worse for all that (maybe better).

Should I just point and shoot? Or is it such that sometimes effort is required, sometimes not?

I don't have the answers, I'll just enjoy the results as they happen and hope I regret nothing.

Another way to print your books

I didn't give the credit at the time, but hendrik replied to my post on bobbooks with an interesting note on a similar service - blurb. Whilst based in the US, they send to Europe. There are 2 distinct differences, firstly that they used a network of print shops and so cannot guarantee consistency, leading to the second point that a guide to blurb colour management has been put together. Find that here.

This latter point is really useful - I might well give blurb a go with a project I've got that wants printing in this way.

Unusual inspiration

Receding sun, Alberta Rockies, September 2007

It's not often I can say that photos look better on the web.

When I posted the above picture here, I noted that I was unsure where to go with it. I certainly like it a lot - it properly captures the mood & feeling of the time I was there but the deep shadows were proving tricky. Seeing the web version, with the very "blocky" yet subtle differences in the shadows I figured what changes I wanted: very little.

The top image is the result (almost). This is going to be tricky to print so I actually have a few versions for printing to try and get the right effect in the shadows. There is very little in the way of adjustments: a couple of areas of clouds have contrast curves and there is a small amount of contrast work in the shadows to ensure that the different hills are distinct in print. That's about it - I'm not even bothering with sharpening, which ruins the overall effect in this case.

Canon EF-S 17-55 f/2.8 IS: a footnote

In my part 2 of my 2 part review (part 1 here) I made the statement regarding tilt of the lens due to the zoom mechanism. I left the statement hanging somewhat and realised that I should do a check to see if the impact is as expected.

As a result, I've run tests on a series of targets with perpendicular features (wall & floor) to test for non-parallel focal plane.

Test run by taking shots at f/2.8, manual focused on the wall at the centre of the lens. I've then checked the relative focus of the paving slabs on the floor to see where the focal point lies. If the focal plane is parallel to the film plane, the wall should be in equal focus, if not, there will be a point on the slabs that is sharper.

Overall conclusion: at a variety of shorter ranges (less than hyperfocal distance), focus accuracy and possibly depth of field (combined with sensor resolution) swamp any tilt effect. I've also discovered my 17-55/2.8 has a slight tendency to front focus (but that's another matter entirely).