While on a bike ride the other day, we stopped to watch a bit of hubbub as 15+ people doing a photo shoot for a catalog spent the better part of half an hour for just one shot.

Of course, Anthony doesn't care that they had a Nikon D3 or how they did the lighting, etc., so he was soon off to investigate more important things...

He eventually noticed the ones I was photographing, and he came over and got it (that's the first photo above). He also had one that hadn't yet turned fluffy, and was surprised and interested that they could both be the same plant...

But, he's five years old, so “piqued” goes only so far. His interest was soon captured by some birds in apparent need of being chased...

Yesterday's picture of a water lily was taken during a dark, rainy day at the City of Kusatsu's (“Water Forest”) park. The name in Japanese is mizu no mori (水の森). They have outside gardens with all kinds of water plants, a large greenhouse (“Lotus Hall”), a movie theater, and restaurant. It costs $3 for adults.

It was raining while we where there, which is perhaps appropriate for a “water forest”.

One thing I'll never understand is the apparent need in Japanese culture to make allowance for the fact that a smoker might not want to go for more than three minutes without a cigarette. Smoking was not allowed in the gardens, except for pretty much every place there was a bench (the little covered area in the photo above being one of the many smoking areas). We were the only ones there at the time – a rainy day just before closing time – so we were lucky, but one smoker would ruin the whole gardens for everyone else. Sigh.

The rain picked up toward the end, so we took the opportunity to have some lotus ice cream (which was quite tasty)...

Near the entrance was a large display of water lilies of many colors. Here's another look at the one from yesterday...

Once you get close to the place, the place is easy to find because there's 300-foot tall wind turbine right at the end of the parking lot...

It's 200 feet up to the center of the prop; the blades have a 230-foot circumference. It generates electricity in winds from 3mph to 56mph (generating 1500kW in winds from 29mph on). It doesn't spin very fast... the speed is limited to about three seconds per rotation, which is what it was doing while we were there.

Without limiters, it could, of course, spin much faster, but that would be bad. If winds go over 56mph (25 m/s), the whole thing shuts down so that it doesn't self destruct like this one did (slow motion view) when its limiter broke.

The caption is just a silly reference to the “art” vs. “photography” discussion in my recent post about HDR. When I first looked this picture, I wondered why it was a blurry – it seemed mosaiced or pixelated, as if it hadn't been fully loaded from disk before display – and I waited for it to finish loading and to “snap” into focus. I eventually realized that it wasn't pixelated, but rather, it just had a lot of water drops that you (or, at least, I) don't notice at first.

It's a lotus flower. [UPDATE: Peter Barnes of Barnes Botany left a comment on a followup post indicating that this flower is a waterlily, not a lotus. Oops!]

I took the shot handheld, on a cloudy, gloomy day, leaning precariously over a water-lily pond, so the full-size version is not as crisply focused as it could be, but it's good enough.

In any case, even though I know what it is, it still has a strong “processed” feeling to it, as if it been run through a few Photoshop filters. Hence, the caption.

I did crop it a bit... here's the full frame:

Last month, we visited a lotus garden in Shiga (about 45 minutes away by car) that I'll write more about another time. Fumie learned about the garden on the blog of a fellow fan of Kousuke Atari, a pop singer we've seen in concert many times. My first time was in Miyajima, and my seventh (I think... I've lost count) was this evening, here in Kyoto, where he was backed by the Kyoto Symphony Orchestra.

During the intermission, a lady came up to us and introduced herself. She's the lady with the blog, and had been exchanging mail with Fumie, but they'd never met. So, now we've met the lotus lady. This reminded me that I'd not posted anything from that trip, so having arrived home from the concert, here we are with a post.

It was interesting how she could find us among the thousand or so people at the concert (it was a small venue). In the seven Kousuke Atari concerts I've been to, big and small, I once met a guy from Sweden, whose wife had dragged him along. Other than that, I've never noticed another “doesn't look Japanese” person, so I'm probably pretty easy to pick out of the crowd. 🙂

In my running set of posts about my visit to the Fushimi Inari Shrine in south-east Kyoto, I showed the paths lined with thousands of gates that the shrine is famous for, but I ended the most recent installment – Fushimi Inari Shrine: Foxes, Treasure, and More with the teaser that there was so much more than just the gates.

We didn't have a map, so we took what turned out to be a side path that later looped back to the main path....

It was great luck to take this side path, because in doing so we discovered a wonderful visual treasure: an area jam packed with a mishmash of a bazillion “sub shrines”.

As I speculated in my first post, it seems that each of these sub-shrines represents an Inari Shrine elsewhere in Japan, providing a physical presence for the real shrine here at the head Inari shrine. (“Inari” shrines being dedicated to business success, material wealth, and foxes, among other things.)

This small area was just packed with history. You could see generation after generation after generation of additions as it was expanded over time. It's all nestled in the middle of the mountains, so expansion often involved constructing stairs and carving out a flat terrace.

Some of it looked as if it had been completed yesterday – very modern, with proper drainage and smooth, polished cut stone – but more often than not it looked to be hundreds of years old. The photo above shows some very weathered stones.

It's an incredibly visually-rich area. Anthony ran around and explored as I tried to see what I could capture photographically, but his attention span matches his age (five years old), so I didn't have much time.

In looking over the photos when I got home, I was struck by how often the sense of scale was totally lost in the photo. For example, the gate in the photo above is about a foot or so tall, with the center plaque thing being about the size of a deck of cards. (You can also see it in the center of the “Relatively Modern Area” photo above.)

On the other hand, the gate in the center of this next photo is huge, with life-sized horses flanking it...

At least I think the horses were life sized... they really look like toys in the photo, so I'm not 100% sure.

Near the entrance to the area was a large map showing the location of sub shrines by number...

The map shows no indication of the visual chaos that envelops the site. The map is boring. The site is amazing.

Of course, this is Japan, so any area of beauty or interest must be marred with lots of wires, utility poles, etc. It's the law.

Continued here...

High-dynamic-range – HDR – is an image-processing technique that's been gaining popularity over the last few years. HDR can be used to create some amazing, impactful, stunning images. For some eye-popping examples, see this page, which is just one page of many that are linked from this HDR roundup.

I haven't created anything amazing with HDR, but I utilized HDR in whipping this image together, just for this post...

HDR attempts to overcome a limitation of current camera technology... a limitation that disallows a camera from picking up fine detail in the dark shadows and bright highlights of a scene at the same time. This limitation can be easily seen in the main source photo used in creating the image above, a two-second exposure that I took during the Kyoto Higashiyama “Hanatoro” Lightup Event in March:

As you can see here, the lanterns appear to be just big blobs of white. They weren't big blobs of white, of course (they were a delicate pattern of fine colors, and looked like this), but with the limitations of my camera's image sensor, if I left the shutter open long enough for a reasonable capture of the dark street, the relative firehose of photons from the much brighter lanterns simply overwhelm the sensor in that part of the photo. With my camera's sensor, as with virtually all digital cameras today, “overwhelm” means “registers as pure white.”

On the other hand, if I use a shorter exposure – one that slams the door shut on the photon firehose before the image of the lanterns is totally blown out – the detail in the lanterns is retained, but now the unfortunate side effect is that the dark areas of the scene don't have a chance to send enough photons, and they end up black. Here's the same scene, but with a much shorter exposure that allowed only 1.7% as much light in as above...

In this particular situation, my eyes didn't have a problem capturing detail in all areas of the scene at once, but like our cameras, our eyes do suffer from this issue to some extent (try reading the label on the end of a 100w bulb while it's lit and you'll have a difficult time of it). Still, camera technology is far behind God in this respect, and that's where HDR comes in.

To create the shot seen at the top of this post, I brought in lantern detail from the “shorter exposure” image, as well as from another intermediate image, and blended it all in where the original was blown out. For having spent just a few minutes on it (and without much HDR experience to begin with), the technical result in this case is surprisingly good – the result looks completely natural – although the effort seems to be a waste of energy on such a boring scene.

Here again is the HDR version, and the three photos used to make it; mouseover the buttons below the image to see each version...

On rare occasions I do this kind of specific “filling in the details” processing, pulling in sections from other exposures of the same scene. I did it when I created the large image of one lantern (here), and perhaps I've done it one or two other times.

Somewhat similarly, at times I've “painted” in color that was lost due to the aforementioned “blown out” problem. I did it with the sky of this picture, and the blue lights in the side areas of this photo. The center section actually had white lights mixed in as you see there, but the sides were pure, retina-burning blue. Unfortunately, in the original, the brightest parts of the blue lights came out pure white, just like in this unfixed photo, so I painted back some blue.

All this so far shares a lot of DNA with HDR, but it's probably not what most people would consider true HDR. To look at that, we first need to understand...

The Problem with HDR

Here's a “real” HDR image I whipped up just for this post, using two photos I took of our room at a seaside ryokan in Ise, Japan, a year and a half ago, showing detail in both the brilliantly bright outside, and the much-darker inside...

I've got to admit up front that there are a few problems with my selection of source photos, and how I processed them. I didn't use a tripod when I took them, so the photos don't match up perfectly, and I used a different aperture (and likely a different focus) with each, which can cause further image-mismatch distortion. As for processing, I did it by hand in Photoshop, quickly and roughly, so some of the transitions are certainly harsher than they need to be.

Still, those are only small details compared the biggest problem, and that's our brain: in the HDR Combination above, the inside wall is actually brighter than the outside sea – something our brain just knows makes no sense.

The root problem with HDR is that even though, theoretically, it is able to bring to bear a wider range of detail, that detail still ends up being shoehorned back into the same “low dynamic range” of our current image/display technology.

In the “blown out” versions used in my two HDRs above, both the lanterns (of the first) and the outside sky (of the second) are encoded as “pure white”, which means that both appear – as far as our image/display technology is concerned – to be of equal brightness. Of course, that's silly because they're not of equal brightness: the sky is certainly thousands of times brighter than the lanterns, and were they to be placed in a scene together, the lanterns would be so relatively dark as to appear to be unlit.

So, if I wanted to fix the disconnect, I'd have to encode the photo so that the sky was brighter or the lantern was darker. Well, the sky is already encoded as bright as possible (“pure white”), so that means that I would have no choice but to dim the appearance of the lanterns considerably. That squeezes out detail as the whole lantern is moved toward blackness. It's closer to reality, but it's just the opposite of HDR.

Another approach would be to make the lantern dimmer, but less dim than it really should be, in the hopes of preserving some of its detail. We then run into the problem of how bright it becomes compared to other things in the scene, in the same way as the wall compared oddly to the outside sea. Considering that scene, if we added the lantern but dimmed it only half as much as reality would dictate, it may well appear reasonably natural when compared to the sky, but it would end up appearing brighter than other objects in the room that we know are brighter than it. So, then, begins the cat-herding exercise of trying to adjust relative brightnesses in such a way that we preserve local detail while at the same time avoiding the global disconnect for our brain.

“Preserve local detail” means that any one object or area takes up a good, meaty chunk of contrast, that is, of the total black-to-white range of brightnesses an image can contain. This is fine when you have everything in a similar brightness range, but when you have wildly differing brightnesses, their expanded ranges can't fit into the whole without overlapping. “Overlapping” is just the opposite of “wildly differing”, and hence our problem. In the end, you just can't fit 10 gallons of photons into a half-gallon bucket.

Sometimes the overlapping can be managed carefully, such as I did in the lantern photo, but it's questionable whether that kind of minimal treatment is really “HDR”. This road sunset is a better example of careful, prudent HDR. The room at the seaside ryokan above, and this lake scene are a bit more iffy, but most people's idea of HDR are wildly unrealistic, “gimmicky” scenes like this beach and this car.

The Fundamental Problem

The fundamental problem is that display technology has an extremely limited dynamic range. Your computer monitor can not show anything blacker than what you see when you turn it off, and it can't show anything brighter than what you see in this box: . Whether it's an overexposed firefly or the molten face of the sun, it comes out the same: . There's something wrong with that.

Frankly, until looking at a photo of the sun is as dangerous as actually looking at the sun, HDR will remain hobbled in the “reality” department.

Ouch.

Those are harsh words, yet, at the start of this post I used complimentary phrases like amazing, stunning, impactful, and eye-popping to describe HDR images. This apparent discrepancy in my thinking can be resolved by understanding the difference between “image” and “photograph”.

I sense that many who employ HDR do so feeling that it's a way to advance deeper in photography, to make an end-run around technological limitations and unleash the true potential of their camera. There are occasional exceptions, but generally, this turns out to be pure folly.

HDR as an Artistic Technique

I don't think HDR is a photographic technique, but on the other hand, I do think that it can be a wonderful artistic technique. The difference may just be semantics to some, but I think it's important to keep the proper frame of mind.

In the “artistic technique” frame of mind, HDR is in the same camp as numerous other image-processing techniques that use a photograph as their source. As one example of a “cousin of HDR”, here's a “painting” I made from a photo with Corel's Painter Essentials 3:

A closer cousin might be the “Dave Hill look” I applied to a photo in a recent post:

I developed the technique to create the following image myself, using Photoshop's “Darken” blend mode...

Here's the result of applying way too much sharpening with Adobe Lightroom...

Here's another one done by twiddling the knobs in Lightroom...

Of course, there are untold other techniques, including numerous Photoshop filters that can also do amazing things.

The point is that none of these are “photos” anymore, just as most HDR images are generally not photos anymore. HDR can be great, but the result is almost always an unnatural, unearthly look. That look, like the results of any other artistic treatment, can be impactful and powerful and compelling, but it's a slippery slope to cliché and gimmicky. Be careful.

I'd been thinking of writing a post like this for quite some time, and finally decided to do it when GMSV (a witty and informative technology blog I read daily) referred to HDR as High Dynamic Range photography. I'd rather it be referred to as “imagery”, not “photography.”

Why I Don't Do HDR

After all is said and done, why don't I do HDR? Well, I'd love to if I had the time, but there are so many things I still want to explore within the realm of photography that I just don't have time for that non-photographic tangent. If I did have that kind of time, I think I'd want to spend it exploring more what I might be able to accomplish with Corel Painter.

What do you think?