Is God in the details?

Videbaek · Sep 12, 2007

I'll be referring to developments in digital photography but bear with me, it's about analog photography too... Lately it seems that HDR (high dynamic range) photography is sweeping the land. Without getting into the verboten digital details, HDR is about maximizing the detail in the low, mid and high tonal ranges by making multiple exposures and merging them. All well and good, surely, but I've been surprised to see that quite a few photographers whom I thought had good aesthetic sense have lately become enamoured of HDR with what I think are catastrophic results -- pictures so overloaded with detail that the eye simply gives up at first glance (at least mine does). I know that this is function of "if technology lets you do it, then it must be good" and the power of marketing and fad, but it's about something so central to photography and picture-making in general, namely composition and the handling of detail within a composition.

The "trouble with detail", if anything, is much more of a concern in the realm of "fine art" photography where the ability of B & W film to record fine detail as championed by Adams, Weston and a host of others is thought to be of artistic value in itself. It emphatically isn't. God is found not in the mass of details but in the well-chosen detail -- the pearl ear-ring of Vermeer's "Girl With A Pearl Ear-Ring", the breathless gap between the fingers of Michelangelo's God and Adam, the top-hat on the carriage driver of Steichen's "Flat-Iron Building".

Of course it's much easier to avoid making decisions and record everything indiscriminately. But composition is about discrimination and decisions. Great composition is about great decisions. I'll be watching with spread of HRD with trepidation.

Andy K · Sep 12, 2007

Its the latest 'new toy'. Recently 'bokeh' was the buzzword, now it's HRD. It'll pass.

juan · Sep 12, 2007

A sharp, detailed photograph of a poorly chosen subject is still a bad photograph. Unless the camera club judge gives it an award.
juan

Struan Gray · Sep 12, 2007

A single main subject or a 'punctum' that defines the centre of your particular engagement with the photograph is the norm, and sometimes hardens to a rule, but it's not the only way, especially in as broad a topic as composition.

I have been inspired by abstract paintings and by the art of China, Korea, Japan and elsewhere in the Far East, all of which often lack a single main subject or centre of interest, and which can be filled from edge to edge with fine detail. For me, the distinction is between a photograph that shows a single thing or event, and one that shows a state of being or an environment.

For a canonical photographic example see Winogrand's "World Fair" bench.

Ray Heath · Sep 12, 2007

g'day Svend

excellently put, i totally agree

i've often thought the same thing in regard to more traditional/normal techniques such as split grade printing

bjorke · Sep 12, 2007

BTW, people (even in the digirealm) have started to catch on to the idea that wide-blur, low-intensity unsharp masking (and inverse masking) can recreate many of the tone mapping effects popular in "HDR" processing.

Unsharp masking is a rather old FILM technique that has been adapted to digi. You can still do it with film, though it's easiest with big negs because the alignment and registration is a real PITA.

(how it works, very roughly: if you make a blurred copy neg from a neg (use an enlarger or stat camera out of focus and 1::1 -- or make two copies, one sharp and one not), then sandwich them together, the areas where they are teh same - no detail - will be black. The areas where the blurred and focused images are different - the details - will let light through. Based on this you can start to use the mark (or its inverse) for burning and dodging.....)

Michel Hardy-Vallée · Sep 12, 2007

Pardon my language, but I have always found the mindless rush to maximize detail (shadows, highlights, HDR, sharpness, resolution, larger format, larger enlargements, etc) somewhat of an anal fixation, a pissing contest, or a dick-size contest, depending on the mood I'm in.

It's not surprising, though. The Daguerreotype was (and still is) an incredible feat of detailed image technology. Can you believe the impact of the dag on people who seldom had any contact with images at all? Sometimes I have the impression that we want to re-create the daguerreotype's precision but we're not interested in doing it right, so we artificially squeeze all we can from other materials.

Having details is also like bureaucratic power. Think of the leader who has a humongous database of your each and every moves. What power lays hidden in such an amount of precise details! In photography, we can achieve something similar very easily. Some would call that the society of surveillance.

I think it also belies a certain fear of the world. By trying to capture so much detail at once, you are in fact refusing to make the effort of getting closer to your subject. If you can just take one big high-resolution picture of a crowd, and get away with selective blowups instead of taking portraits one by one, then you're not interested in engaging with your subject. That also reminds me of the rush from still photography to video: we would like to capture everything possible, and edit off-line instead of editing on-line.

In fact, I'm sure we would love to have a Google Earth that's updated in real time, replicating molecule for molecule what happens on the current Earth. That way, we could always rewind and freeze-frame everything that happens at any time.

Oh wait, that's what God (if she exists) is supposed to be able to do!

Akki14 · Sep 12, 2007

I had HDR explained to me as trying to make digital do what film does already. Film already has a high dynamic range - that's why you can dodge and burn and splitgrade to your heart's content. Unfortunately, bad HDR looks kind of like a poor dodge&burn job with haloing etc. I have to be honest but that's what bugs me more about HDR-type photos than the details.

On the other hand, maybe this is why I find St. Ansel photos so awful. Most of his photos have far too much detail for me to like. I always thought it was purely the subject matter not interesting me but maybe it's a combination of that and the excessive detail.

somewhat offtopic question, is bokeh really just depth of field? Or do I just know dumb digiphotogs who use it as a synonym for depth of field? I could have sworn it was more specific to the lens and aperture than just "the fuzzy out of focus bit of the picture".

Dave Miller · Sep 12, 2007

Strange really, for years most of us have been chasing detail via the purchase of ever more exotic glass, and hardware, bigger film and hang the expense, but now the digi-camp are trying to catch up it seems to be wrong.:confused: I wonder why?

catem · Sep 12, 2007

Akki14 said:
somewhat offtopic question, is bokeh really just depth of field? Or do I just know dumb digiphotogs who use it as a synonym for depth of field? I could have sworn it was more specific to the lens and aperture than just "the fuzzy out of focus bit of the picture".

This is one discussion I remember, (there are sure to be more!)
(there was a url link here which no longer exists)

Dave Miller said:
Strange really, for years most of us have been chasing detail via the purchase of ever more exotic glass, and hardware, bigger film and hang the expense, but now the digi-camp are trying to catch up it seems to be wrong. I wonder why?

Have we though? I haven't. Probably those who were always into sharpness and detail (and all that) over everything else have simply transferred to digital as it offers them what they were always looking for...

HDR makes me think 'graphic art' more than anything else, landscapes seem more often than not like something out of 'Shrek' - though I have seen a very limited amount of excellent images that are doing more emotionally and spiritually than this. I guess there are too many people 'having a go' without really getting to grips with it as artists (is that a kind way of saying what I really think, which is that there is just too much digidross around on internet forums...)

darinwc · Sep 12, 2007

Dont fault people for learning techniques to overcome equipment limitations.
Thats what LF Photographers do with a large neg and movements, right?

Some will brag about the new thing they can do but thats just people.

dslater · Sep 12, 2007

bjorke said:
(how it works, very roughly: if you make a blurred copy neg from a neg (use an enlarger or stat camera out of focus and 1::1 -- or make two copies, one sharp and one not), then sandwich them together, the areas where they are teh same - no detail - will be black. The areas where the blurred and focused images are different - the details - will let light through. Based on this you can start to use the mark (or its inverse) for burning and dodging.....)

The first time I saw unsharp masking it was a technique for reducing the contrast of astronomical photographs. What you did was to make a thin slightly blurred positive from your negative, then sandwich the negative and positive together and print. What happens is that shadows on the negative are given some extra density by the positive, while highlights are not affected. This reduced the overall contrast of the negative so you can print the highlights down more without making your shadows too dark.

Alternatively, if your negative has a normal contrast range, you can use
unsharp masking for sharpening. Because the contrast of the sandwich is reduced, you can also choose to print on higher contrast paper which increases acutance.

dslater · Sep 12, 2007

bjorke said:
BTW, people (even in the digirealm) have started to catch on to the idea that wide-blur, low-intensity unsharp masking (and inverse masking) can recreate many of the tone mapping effects popular in "HDR" processing.

This doesn't seem right to me - these techniques are only going to work if the detail is present in the original digital file. The original purpose of HDR was to work around the fact that digital sensors are only 12-16 bit devices. As such, they have a very limited dynamic range. HDR allows you to take a few frames at different exposures and combine them into a digital file that has 32-bit pixels. With 32-bits, you have much much more dynamic range. Once you have that 32-bit file, I can see how an unsharp mask techniques can then be applied to get the effects you describe.

dslater · Sep 12, 2007

Ultimately, I think HDR will be replaced by Digital cameras with 32-bit sensors. Since current 12-14 bit digital cameras seem to have a dynamic range of about 5 stops, I would estimate that a 32-bit sensor would have a dynamic range of 10-11 stops.

Roger Hicks · Sep 12, 2007

The maximum log brightness range of a print -- any print -- is at best about 2.2 - 2.3.

How you fit the subject brightness range onto the print brightness range is part of learning to make good pictures. Use the wrong technique with the wrong subject, and the Zone System and HDR are equally crap. Do it right, with the right subject, and they're equally good.

One difference is that more people put more effort into the ZS, working around its limitations and carefully considering the aesthetic result. With HDR, it's easier just to push the buttons and rely on the software. But there's still an awful lot of trash with both systems, because too many people regard photography as only a technical challenge, for sharper lenses or longer tonal ranges or both.

So God isn't in either. If She exists, She is in the photographer's eye, brain or heart.

thebanana · Sep 12, 2007

If She exists, She is in the photographer's eye, brain or heart.

Indeed.

bjorke · Sep 12, 2007

dslater said:
This doesn't seem right to me - these techniques are only going to work if the detail is present in the original digital file. The original purpose of HDR was to work around the fact that digital sensors are only 12-16 bit devices. As such, they have a very limited dynamic range. HDR allows you to take a few frames at different exposures and combine them into a digital file that has 32-bit pixels. With 32-bits, you have much much more dynamic range. Once you have that 32-bit file, I can see how an unsharp mask techniques can then be applied to get the effects you describe.

The particular "look" of HDR images is tonal compression in the extreme areas. Thiscontext-sensitive tone-mapping requires a "deep" picture, but I see a lot of "HDR" images on flickr that are made from a single RAW file, which may in many cases have less tonal range than a color neg

Nils · Sep 12, 2007

On the other hand it could be easier to choose a medium that won't let you record everything 'indiscriminately' - avoiding certain decisions because they have already been made for you...

To use an analogy similar to an earlier one, companies record huge amounts of data - are all those details useful? are they useful in isolation from other details? Only if they are, or can be, analysed and interpreted in a manner that has some meaning and value to the relevant people.

HDR is not necessarily high resolution either; nor does it need you to engage the subject any less - especially given that the current most accessible method of the technique for the enthusiast usually requires 3 exposures.

Regarding 'fear of the world', who is the more frightened - the person who ignores the details that are present? or the person who tackles them head on? (whether or not they choose to disregard them later on)

No art is particularly easy to master and, even if mastered, it still may not be appreciated - plus the 'reliance' on software is pretty much the same as the 'reliance' on film manufacturers or lens makers etc...

Anyways, is god in the details? Sometimes... Sometimes not...

dslater · Sep 12, 2007

bjorke said:
The particular "look" of HDR images is tonal compression in the extreme areas. Thiscontext-sensitive tone-mapping requires a "deep" picture, but I see a lot of "HDR" images on flickr that are made from a single RAW file, which may in many cases have less tonal range than a color neg

Yes - I think in many of these cases, the users do not really understand HDR. it is likely that they could achieve the same result without converting the 16-bit image data to 32-bits.

DrPablo · Sep 20, 2007

dslater said:
Ultimately, I think HDR will be replaced by Digital cameras with 32-bit sensors. Since current 12-14 bit digital cameras seem to have a dynamic range of about 5 stops, I would estimate that a 32-bit sensor would have a dynamic range of 10-11 stops.

Bit depth has nothing to do with dynamic range. It only has to do with the number of intermediate tones between extremes.

If a sensor can record detail over a 1000:1 brightness range (approx 10 stops), then increasing from 24 to 32 to 48 to 96 bits is not going to change the upper and lower limits of that range -- it will only change the number of intermediates.

To get a higher dynamic range, you need a physical sensor that is responsive over a greater range of light. Once you've accomplished that, then you'll need higher bit depth to accomodate all the data -- but it's the sensor and not the bit encoding that makes this possible.

The 32-bit files in HDR are necessary because each individual pixel holds the R/G/B information from 5, 8, 10, etc individual captures. So you need extra bit depth just to hold all that data. But it's not the bit depth per se that creates the higher dynamic range -- it's the content of that pixel information that does.

dslater · Sep 21, 2007

DrPablo said:
Bit depth has nothing to do with dynamic range. It only has to do with the number of intermediate tones between extremes.

If a sensor can record detail over a 1000:1 brightness range (approx 10 stops), then increasing from 24 to 32 to 48 to 96 bits is not going to change the upper and lower limits of that range -- it will only change the number of intermediates.

To get a higher dynamic range, you need a physical sensor that is responsive over a greater range of light. Once you've accomplished that, then you'll need higher bit depth to accomodate all the data -- but it's the sensor and not the bit encoding that makes this possible.

The 32-bit files in HDR are necessary because each individual pixel holds the R/G/B information from 5, 8, 10, etc individual captures. So you need extra bit depth just to hold all that data. But it's not the bit depth per se that creates the higher dynamic range -- it's the content of that pixel information that does.

I am well aware of this - However, I think it is pretty obvious that digital sensor manufacturers wouldn't create a 32-bit chip unless the sensors had the dynamic range to take advantage of it.

DrPablo · Sep 21, 2007

True. Though I think the dynamic range of current digital cameras is pretty much fine. All they need to do is make the shadows far less noisy and make the highlights end much less abruptly. DSLRs already have a much greater latitude than slide film, and they're comparable to a negative with standard development.

dslater · Sep 21, 2007

DrPablo said:
True. Though I think the dynamic range of current digital cameras is pretty much fine. All they need to do is make the shadows far less noisy and make the highlights end much less abruptly. DSLRs already have a much greater latitude than slide film, and they're comparable to a negative with standard development.

Hmm - that's not what I understood. everything I've read indicates digital sensors have about the same dynamic range as slide film. In any case B&W film processed properly has a much wider dynamic range than digital sensors or color film. I'd like to see a digital sensor with at least as much range as B&W and preferably more. There are always difficult lighting situations that need more range to capture.

DrPablo · Sep 22, 2007

Some guy on FredMiranda did a huge study in which people put their cameras on manual, and shot a series of pictures of a wall, adjusting exposure by 1/3 of a stop, and the dynamic range was determined to be the number of stops between the lowest and highest exposure setting where there was recognizable detail.

The average was something like 7 2/3 stops, and some people (mainly with full frame cameras) reported over 9 stops. I'm sure some of this DR is not necessarily useful DR, like the bottom 1 stop and the top ~ 1/2 stop. But still, that's a pretty good DR. Part of the problem is that there is no shoulder, though, so highlights end up being all weighted equally until they cut out. So the DR is qualitatively different than on film.

I haven't studied this formally, but I'd give Velvia maybe 4 useful stops in my experience, and Astia maybe 6 or 7. I definitely have to be more careful these days shooting slide film than I ever did when I predominantly shot digital. Then again I could shoot 800 pictures to make sure I hit the exposure on one of them, and I'm not doing that with 8x10 slide film.

bjorke · Sep 22, 2007

I think "Photo Techniques" did a similar study (was it "Digital Photo Pro"?) that showed varying dyanamic range between a # of cameras. The best cameras did very well. What was interesting to me was that the camers with the widest DR were not known as the best cameras -- because, apparently, they also had a lot of variance (noise) in the extremes of their ranges. Cameras with low variance AND wide DR were, as you might expect, top-dollar items.

(At this point the little clown on a tricycle comes through reminding us all that film frames do not display these artifacts, and behave with great consistency across all acceptable ranges -- that is, they are still a match for the best commercial/consumer digi sensors)

Is God in the details?

Membership Council

Privacy & Transparency

Privacy & Transparency