Struan Gray
29-Nov-2006, 06:29
This is a spill-over from the B+W magazine thread (http://www.largeformatphotography.info/forum/showthread.php?t=6487). It is not really an LF-topic, but I hope the moderators will indulge me, as it is concerned with thinking about the medium in the conscious, contemplative way that comes as second nature (or is forced upon) LF workers.
Photographs contain a time structure: that is, they are a record of the light that hits the film during the period that the shutter is open. When the light does not move or vary, all that is recorded is the Guild Photographer's perfectly focused spot, but if the lights or the camera change during the exposure you can analyse the resulting blob on the film to learn something about the subject, the camera, or yourself. All three are valid topics for art or science.
In analogue photography this is the basis of canonical gems such as Lartigue's racing car, the motion or gesture blur that led animators to draw lines behind fast-falling coyotes, and the old schoolboy gag of turning up twice in the same group photograph. There are also technical toys like finish line timers and streak cameras for very high speed photography. In all of these you can think of the photograph not only as a spatial collection of tones, but also a temporal one: the exposure is made up of a series of time slices, which are combined to form the final image.
Digital has two advantages over analogue even for these sorts of traditional uses. First, digital is linear and - given the right camera - has better low-light performance (it is also simpler with digital to recover signal to noise by averaging). This makes it easier to tease out the separate time slices because you are less likely to lose information to reciprocity failure. Second - and perhaps more important conceptually - you can combine the time slices in more complex and subtle ways.
A classic case would be the long exposure that removes people and moving cars from street scenes. With digital you can mimic the analogue case with a single long exposure, but you can also play other games if you take a succession of short exposures and combine them with a rule that is more complex than simple addition. For example, you could photograph a town square or plaza and instead of averaging away the people, combine the individual shots so that the final photograph included everyone who visited the square that day. Or everyone who sat down for longer than two minutes. Or everyone who wore red clothing.
Another example would be the way that analogue photography privileges bright objects. Your Guild member gets understandably upset when a jet or satellite swims into his night time ken, and all those conceptualists whirling their cameras by their straps over their heads soon learn that only the brightest things leave a trail. Digital combinetrics would allow you to select for dark trails instead, simply by changing the combination rule. With digital it is as easy to photograph the swirls of a flock of rooks as a flock of seagulls, with analogue you tend to stick to white birds.
Arguably all the above could be done with analogue techniques, given enough time, internegatives and masks. I feel though that the ease of the digital way is so much greater that it amounts to an enabling technology. Milk splashes were photographed long before the electronic flash was invented, but it's Edgerton's photo everyone remembers.
And then you have the things that are simply impossible with analogue. To stop things getting too fanciful here's an example from my own experience: combining image stacks in microscopy to increase the apparent depth of field. Again, this can be done in a simple, painstaking way with analogue photography using a sheet of light and multiple exposures; but there is no way in that case to fine-tune your definition of 'sharp'. In the digital realm it is possible to tailor the combining algorithm to look for particular types of sharpness, or to concentrate on particular spectral bands, or to ignore sharp objects near to experimental artefacts, or, or, or.
Art photography hasn't really got to grips with the implications of all this, but I am sure it will, given time. There seems to be a lot of current interest in very long exposures, averaged faces from classes or groups, and things like whole journeys seen out of car windows. My point is that these sorts of concerns can be addressed with more finesse and descrimination by using more sophisticated selection rules to create the combined image. For the moment though, the scientists are some way out in front.
My own desire would be for a high-fidelity single-shot back to use for playing with motion in landscapes. Breaking waves photos that show the turbulence in sharp relief for the whole motion of the wave. Darkness trails on a windblown sea. The colour structure of a snowscape photographed over the course of an arctic summer's day. Are you listening Santa?
Photographs contain a time structure: that is, they are a record of the light that hits the film during the period that the shutter is open. When the light does not move or vary, all that is recorded is the Guild Photographer's perfectly focused spot, but if the lights or the camera change during the exposure you can analyse the resulting blob on the film to learn something about the subject, the camera, or yourself. All three are valid topics for art or science.
In analogue photography this is the basis of canonical gems such as Lartigue's racing car, the motion or gesture blur that led animators to draw lines behind fast-falling coyotes, and the old schoolboy gag of turning up twice in the same group photograph. There are also technical toys like finish line timers and streak cameras for very high speed photography. In all of these you can think of the photograph not only as a spatial collection of tones, but also a temporal one: the exposure is made up of a series of time slices, which are combined to form the final image.
Digital has two advantages over analogue even for these sorts of traditional uses. First, digital is linear and - given the right camera - has better low-light performance (it is also simpler with digital to recover signal to noise by averaging). This makes it easier to tease out the separate time slices because you are less likely to lose information to reciprocity failure. Second - and perhaps more important conceptually - you can combine the time slices in more complex and subtle ways.
A classic case would be the long exposure that removes people and moving cars from street scenes. With digital you can mimic the analogue case with a single long exposure, but you can also play other games if you take a succession of short exposures and combine them with a rule that is more complex than simple addition. For example, you could photograph a town square or plaza and instead of averaging away the people, combine the individual shots so that the final photograph included everyone who visited the square that day. Or everyone who sat down for longer than two minutes. Or everyone who wore red clothing.
Another example would be the way that analogue photography privileges bright objects. Your Guild member gets understandably upset when a jet or satellite swims into his night time ken, and all those conceptualists whirling their cameras by their straps over their heads soon learn that only the brightest things leave a trail. Digital combinetrics would allow you to select for dark trails instead, simply by changing the combination rule. With digital it is as easy to photograph the swirls of a flock of rooks as a flock of seagulls, with analogue you tend to stick to white birds.
Arguably all the above could be done with analogue techniques, given enough time, internegatives and masks. I feel though that the ease of the digital way is so much greater that it amounts to an enabling technology. Milk splashes were photographed long before the electronic flash was invented, but it's Edgerton's photo everyone remembers.
And then you have the things that are simply impossible with analogue. To stop things getting too fanciful here's an example from my own experience: combining image stacks in microscopy to increase the apparent depth of field. Again, this can be done in a simple, painstaking way with analogue photography using a sheet of light and multiple exposures; but there is no way in that case to fine-tune your definition of 'sharp'. In the digital realm it is possible to tailor the combining algorithm to look for particular types of sharpness, or to concentrate on particular spectral bands, or to ignore sharp objects near to experimental artefacts, or, or, or.
Art photography hasn't really got to grips with the implications of all this, but I am sure it will, given time. There seems to be a lot of current interest in very long exposures, averaged faces from classes or groups, and things like whole journeys seen out of car windows. My point is that these sorts of concerns can be addressed with more finesse and descrimination by using more sophisticated selection rules to create the combined image. For the moment though, the scientists are some way out in front.
My own desire would be for a high-fidelity single-shot back to use for playing with motion in landscapes. Breaking waves photos that show the turbulence in sharp relief for the whole motion of the wave. Darkness trails on a windblown sea. The colour structure of a snowscape photographed over the course of an arctic summer's day. Are you listening Santa?