The Terra Galleria Blog – QT Luong

The Treasured Lands 2019 Calendar was released just a month ago. For this edition of the namesake calendar, we chose to include in the calendar only images that have appeared in the book. Since I strive to visit the national parks in various seasons to capture their changing mood, it was easy to find seasonal images that match precisely each particular month. Like he did for the book, the designer avoided the monotony typical of calendars with a few surprises.

The calendar has a respectable size, actually slightly larger than the book. The printing is better than some of the calendars that have been produced with my images in the past. However, this time we have another printed reference, the Treasured Lands book, and it sets the bar high. The book’s printing is undoubtedly richer with more accurate and vibrant color, and what looks like deeper blacks and better highlight detail. However, the differences are clear only if you were to look at the book and calendar side by side – always an instructive comparison since the only controls are quantities of cyan, magenta, yellow, and black inks. The quality of the paper contributes significantly to the difference. The edge contrast is also better in the book (this is most noticeable in the Zion photograph) and this could be because I personally sharpened all the files – using Photokit Sharpener – whereas for the calendar this was left to the publisher and printer. Applying sharpening for print is not intuitive: with the proper amount for the printed page, the image on screen looks over-sharpened.

Regardless of its printing limitations, the calendar is an economical way to bring the national parks to the walls of your home or office, and they make for affordable gifts at a street price of about $10. Like in previous years, I do not sell the calendars, but copies may be ordered on Amazon.

The dual side of Hanford Reach, and reason for the accidental existence of Hanford Reach National Monument, is the Hanford Site (Hanford Nuclear Reservation) the former top secret Manhattan Project plutonium production facility established in 1943, which is almost entirely surrounded by the national monument.

Visiting the mysterious site used to require a security clearance, but since it became part of the Manhattan Project National Historic Site in 2015, any U.S. citizen can sign up for a free tour – and proof of citizenship was not even required.

After the government chose the Hanford Site for its remoteness and supply of water/electricity via the Columbia River, two communities were condemned and evacuated. One of the tours visits the remains of those communities. The other, probably the most interesting one, takes you inside the historic B reactor.

The B Reactor was the first large-scale nuclear reactor ever built. It supplied the plutonium used in the first atomic explosion (at the test Trinity site in New Mexico, July 1945) and in the Fat Man bomb dropped over Nagasaki on Aug 9, 1945. Remarkably, although the Manhattan Project started only in August 1942 and construction at Hanford started in March 1943, by November 1944, B Reactor produced its first batch of plutonium. In February 1945, as Franklin Matthias, who had supervised Hanford’s construction, hand-delivered in Los Angeles the first 100-gram batch of plutonium to a courier heading Los Alamos, he told him the 3-inch object had cost $300 to make. During the Cold War, the site expanded to nine nuclear reactors that produced most of the plutonium for more than 60,000 nuclear weapons. After the end of the Cold War caused the Hanford Site to be decommissioned, those reactors have been entombed (“cocooned”) in concrete, except for reactor B which has been designated a National Historic Landmark by the National Park Service (NPS) and partly converted into a museum.

Except for the tours, the Hanford Site is not open to the public. Research facilities and commercial nuclear production are still active. More notably, the largest environmental Superfund cleanup effort in the country takes place there. The site is considered to be the most toxic place in America and has been dubbed “America’s Chernobyl”, with estimates by the Government Accountability Office for the clean up reaching more than $100 billion. The reactor tours are conducted by the Department of Energy in partnership with the NPS (information and advance reservations here), daily or twice a day except on Sundays from the end of May to mid-November. In addition, there is an annual tour of the cleanup process ran by the Department of Energy.

You meet at a temporary visitor center located in an anonymous business park in the city of Richland. There, you confirm your online sign-up and watch an introductory video before boarding a comfortable bus for a 45 minutes ride straight to the B reactor, during which a volunteer details more facts about the project. The reactor visit starts with another presentation in front of the impressive reactor’s core, after which visitors split into guided groups to view different parts of the facility such as the cooling system and control room before being left on their own. SCRAM is a jargon term (of uncertain etymology) meaning emergency shutdown of a nuclear reactor effected by immediately terminating the fission reaction.

I found the tour to be professionally conducted, explanations to be interesting and in-depth, and appreciated that there are no restrictions on photography, video, or use of tripods. Overall, the tour lasts 4 hours, is well worth the time, and a great way to spend those mid-day hours indoors. Given how historically significant the once secret site was, I think it is a unique opportunity to access something you’ll certainly not see anywhere else and gain a direct understanding of a technology that has changed the course of our world.

Hanford Reach could the national monument with the most unusual history of all. Its lands, located in Eastern Washington, were initially set apart from development not for conservation, but as a security buffer zone for the top-secret Hanford Nuclear Reservation where the plutonium for the Nagasaki bomb – and many others during the Cold War – was manufactured, and which became the most polluted site in the Western Hemisphere.

After the end of the Cold War, the decommissioning of the Hanford Site’s nuclear reactors was followed by a cleanup that continues to this day, and to the establishment, in the year 2000, of a vast nature preserve in the buffer zone that surrounds it. The Hanford Site once provided most of the jobs in the area, but agriculture has quickly taken over. Without the nuclear activity, the former buffer zone, which now stands between industrial agriculture and a Superfund site, would have certainly been developed, especially since the unassuming lands are not home to any significant landform. The mighty Columbia River is heavily dammed, and its last remaining free-flowing section, which stretches 51 miles, is called the Hanford Reach, hence the national monument’s name.

The river in the Reach is lined up with islets, ponds and sloughs that have all but disappeared in the other sections of the Columbia, supporting one of the most productive salmon spawning areas in the Northwest. By contrast, most of the terrain of the monument is desert covered by the largest remnant of the shrub-steppe ecosystem that once blanketed the Columbia River Basin. The area receives between 5 and 10 inches of rain per year.

Approaching the monument, I was struck by the vastness of the area, but also by how dry and flat it appeared at first. More power lines than I’d seen elsewhere criss-crossed the land. It was not helping that I visited in mid-summer – as a family man, I was trying to put to good use a week that was freed because my children were at a camp. Under the scorching sun, with temperatures reaching 100 degrees, the air was constantly filled with haze. Along Hwy 24, the main highway inside the monument, a few signs point to the history of the area, but as far as facilities are concerned, that’s pretty much it. The monument has no visitor center, no campground nor commercial facilities, no established trails, and within its 304 square miles, I saw a single portable toilet.

As often, areas that from a distance and a zipping car look barren reveal their beauty if you know where to go and take a closer look. Although the monument appears large, the areas with public access are relatively limited, so it was not as difficult to find places to hike and photograph as it appeared at first.

Wahluke Unit

The most remarkable part of the monument is located in the eastern part of the Wahluke Unit, which is accessed via a well-graded dirt road starting south of Milepost 63.2 on Hwy 24. After a few miles, you arrive at an intersection. Heading east, you reach a riparian area called the Wahluke Ponds, which are dominated by tall verdant reeds contrasting contrast with the area’s arid grasslands. Continuing south leads to the White Bluffs Overlook, which offers an excellent roadside view of the Columbia River from the top of the bluffs. The west branch of the road ends at the White Bluffs Boat Launch. The river-level view there isn’t great, but approximately a half-mile mile before the end of the road, there is a large pullout that serves as an informal trailhead. From there, a user trail leads in half a mile to impressive close views of the White Bluffs that tower 400-feet above the river. The White Bluffs face West, so the hike is best undertaken in the late afternoon. The trail then climbs and follows the edge of the cliffs, and about three miles later, ends at extensive sand dunes that you will not find criss-crossed with footprints.

Ringold Unit

The Ringold Unit offers the opportunity to see the Columbia River and the surrounding habitats at river level along its well-graded, unpaved 8 miles, with bluffs and golden hills on the other side. The Ringold Unit is adjacent to the Wahluke Unit, but the 2-mile section of road linking to the White Bluffs Overlook, although open to foot traffic, is closed to vehicles. Driving from one to the other requires quite a big detour.

Saddle Mountain Unit

The area is part of the Columbia River Plateau, formed by basalt lava flows and water erosion. That geography is best observed from Saddle Moutain. A road, starting north of milepost 60.1 on Hwy 24 mostly, paved although indicated “high clearance vehicles”, gains about 1,500 feet elevation for excellent views of the Reach. At an intersection, make a right following a sign, the left branch continues for a long distance and eventually descends into the opposite valley. At first, it looks like all rolling hills, but a hike down the saddle from the end of the road reveals large basalt cliffs. I didn’t see a single person on the road or on the off-trail hike to the Saddle Mountain summit, and in the quiet of the place where only the sound of the wind was heard, the nuclear reactors and patches of agricultural lands looked far and insignificant.

Unfortunately, my selection of images is quite limited due to a big data loss, but I still hope that they have offered a glimpse of the beauty found there. I am looking forward to return in the springtime.

Summary: One of the photographer’s greatest fears is to lose a significant chunk of images from a big trip or event. In this long-form article, find out how a memory card failure caused a week of photographs to disappear, what I did to try to recover them via software, then physical data services, and the valuable lessons, counter to common knowledge, to be learned about memory cards, dual card slots, and backups to prevent such a nightmare scenario from happening to you.

Landscape expeditions can be taxing in the long days of summer, even more so if you are also doing night photography. After flying to Seattle, I arrived at the coast of Olympic National Park around 11 PM – many view Treasured Lands as a culmination of my work in the national parks, but I am far from being done with them! Seeking stars, I woke up before 2 AM for the short window between moonset and astronomical twilight. However, the marine layer had rolled in while I was hiking to the beach, and I shivered until past sunrise time without even seeing a sliver of sky. The next day, since I had to drive from Heart of the Hills Campground and hike 45 minutes to Hurricane Hill, I rose before 1 AM.

On the last day, temperatures in the inland plains of Hanford Reach rose above 100F. When I came home from the week-long trip, I went straight to bed. The next morning, I reached for my cameras, took the memory card out, and inserted into the card reader. This resulted in the dreaded:

Attempting to read the card with software

I reacted with mere annoyance at the computer, and tried to read the card from the built-in memory card slot of a laptop. The same message appeared. Growing worried, I inserted the card into a standalone backup device, and it said “Memory card not found”. Surely I would be able to see the files in the camera, since the last time it had been turned on, less than a day before, everything was normal and I was able to scroll through some images? Nope, like the computer, the camera said “Unable to use memory card. Format?”

The PhotoRescue software installed on my computer had been successful at rescuing files from a corrupted card in the past, however, it did not uncover a single file this time. Hoping for better luck with Lexar’s own Image Rescue software that came with the card, without the activation code handy, I contacted Lexar’s customer support. They were quick in providing me a download, but it did not help:

Note that the computer sees 8.2 GB, but it was a 64 GB card. Lexar support suggested several other apps, including Ontrack® EasyRecovery which has a “technician” version with a $500/year licensing cost. As those apps include a trial mode that allow you to attempt to discover files, and require you to pay the fee only to actually recover them, I took each of them for a spin without success. There are a lot of recovery programs around, but if the error is hardware rather than software, you can try all the programs in the world, and they won’t do you any good.

Dealing with physical data recoveries services

It was time to contact a physical data recovery company. All those companies have a similar mode of operation. You send them your media with pre-paid overnight shipping at their cost, they diagnose it and provide an estimate. If you approve the estimate, they attempt to recover files and charge you if the recovery is successful. Seems fair, right? The problem was that the quote from the Lexar-recommended company was quite a bit higher than I expected:

After a bit of shopping, I found another company which provided me a lower quote. I sent them the card via Fedex overnight and got the following diagnosis:

Despite numerous attempts to identify the source of the problem, we were unsuccessful in reaching a solution. At this stage it is clear that this is an extremely complex case that will require extensive research, time & use of in-house donor parts. Your media will be passed to our Research and Development department; They will be able to research & execute techniques such as safe removal, reballing and replacement of the core processor and other components containing device specific information and repairs to fractures in the tracks of the logic board.

The work we would attempt is a highly skilled process that is not infallible. The chance of a successful recovery once these processes are completed is approximately 74%. At this point most Data Recovery services would send the media back to you as they don’t have the ability to do such complex work, let alone any sort of R&D department. Our R&D department is one of the best in the world, and they are responsible for creating new techniques and ways of recovering data that allow us to sustain an overall 92% recovery rate. Currently, we have two options for you:

Option 1. You allow our R&D department to use the lab time, parts & cleanroom facilities needed to gain access to your data. This is a standard process that will be completed within 15-20 business days. To go ahead with Option 1 an upfront payment of $800.00 is required, and this covers everything the R&D department will need to provide you with feedback: all lab time, parts required from our library, and cleanroom usage. If we are unable to gain access to any of your data that initial fee is not refundable.

Option 2. We return your media to you via either our courier or standard option, and you incur no recovery costs as we were unable to recover your data.

Although the fee wasn’t too bad, I wasn’t going to pay upfront without a recovery guarantee, especially after my friend Tommy, a technology entrepreneur and all-around geek opined “These are extremely difficult and risky techniques. I’m skeptical that they have such capability and even if they do, I doubt that the success rate is 74%. I guess more like 25% or less”. I declined, received my media back a few weeks later via regular mail, and sent it to yet another data recovery company:

Would you have proceeded with the recovery at this price?

Although you don’t often read that in reviews, my main complaint with the Sony A7R2 cameras is sensor dust resulting from the mirrorless design and an ineffective sensor dust cleaning system – it mechanically shakes the sensor using the image stabilization actuators. To cope, I work with two camera bodies in order to minimize lens changes. The excellent 24-105 FE (review) stays on the primary camera which is used for most of the photographs, while I reach for the secondary camera when I need more specialized lenses. On that trip, I had failed to do drive backups and the damaged memory card was in the primary camera, which meant that it contained the majority of a week’s worth of work, and probably the best photographs. However, I retained usable images from the second camera, a few of which illustrate this post.

I decided against proceeding with the attempted recovery, saving me the potential disappointment of failure or costs. It is not that the pictures aren’t worth the amount asked. Rather, including all expenses, the trip cost me only a fraction of that amount. This math didn’t account for my time, but no matter how tiring the effort felt, that time was spent on a process that I largely enjoy. Losing the pictures did not rob me of the experiences I had nor of the scouting I did, and rather than looking back by investing in the recovery, I chose to look forward by saving the money for a repeat trip – and a new camera.

The larger conclusion here is that attempts to read a card with recovery software may not always work, physical recovery services are expensive, and also not guaranteed to work. Even after three decades in photography, the incident reinforced several lessons for me.

Memory cards can fail

The disaster drew home the point that memory cards do fail catastrophically. Both data companies found serious physical damage, but that was a card that had been moderately used for a year without any single glitch, so neither “dead on arrival” nor past any reasonable life expectancy – which by the way nowadays is longer than technological obsolescence. It had never been subjected to any form of abuse before as it spent most of its life in a single camera, was formatted after each download, and prior to failure went straight from the camera to a card reader. There were certainly no warnings nor reasonable explanations.

I have been using digital cameras since the first days of full-frame in the early 2000s (remember the $8,000 Canon 1Ds series?) without any card failure, while during that time, I have had to replace a half-dozen failed hard drives. This made me overconfident in flash technology.

If you browse the internet, you will see that I was far from being alone. Quite a few other professional photographers (some with scores of workshop clients) state that they have never experienced any card failure and that when it happened, they were always able to rescue images with recovery software. Clearly, my experience has been different. The fact that you’ve been lucky doesn’t mean that your luck won’t run out at some point, as it did for me – and others. As we will see next, a quick perusal of customer reviews shows that memory card failure is not that rare.

Cards are not equally reliable

The card that failed is a Lexar Professional 1000x 64GB SDXC UHS-II/U3, which is amongst Lexar’s top line of cards, and deemed “professional” by the manufacturer. I used to believe that any memory card from a reputable brand would be reliable. If in addition, you bought it from a reputable vendor, chances that you’d get a counterfeit of questionable reliability would also be low. The preferred vendors are specialist stores such as B&H, but Amazon is fine, as long as you don’t buy from their third-party merchants. Sandisk and Lexar are two of the most well-known brands, and I’ve used exclusively their cards, depending on the best deal I could find at the moment.

Because of that belief, I didn’t pay much attention to customer reviews, adopting the attitude that nothing is 100% foolproof and unlucky folks can have a bad experience with any product. Besides, a quick glance at the ratings show that almost all cards are rated between 4-stars and 4.5-stars, so they must be good products, right?

It would have done me more good to read the Amazon customer reviews before buying the card, but after the card failure, I looked them up. One of the first 1-star reviews I read described the exact same experience I had:

I literally had just reviewed the pics on my Nikon D610 camera and inserted the card into my card reader and got a message that it was not formatted (which it was – I format every card when it’s new). I put the card back into my camera – and same Format error.

Although I didn’t read all the 300 1-star reviews, the ones that I sampled overwhelmingly bemoaned card failure. Since this was becoming quite relevant, I looked at the 1-star review tally: 15%. That’s almost 1 out of 6 reviews, odds similar to the Russian Roulette. If someone killed themselves playing the game, I don’t think you’d attribute his death to “just bad luck”. Of course, this is not a scientific observation because several factors affect review-writing, but you get the idea. On the other hand, it is instructive to compare the percentage of 1-star reviews for a few other UHS-II cards:

• Sony M 3%
• Sony G 6%
• SanDisk 7%
• Lexar 1000x 15%
• Lexar 2000x 17%

If we assume that 1-star reviews are exactly the type you’d leave if the card totally failed, from that small sample, we can see that some cards are four times more likely to fail than others. This data also sugggests that there is a problem with those UHS-II Lexar cards. On the other hand, the Lexar UHS-I card that I have used for several years gets a convincingly low 3% of 1-star reviews. The lesson here is that not all cards are equal, even amongst those from a top brand. And if they can have such a high failure rate, think about cards from less reputable or conterfeit brands! By the way, looking at those numbers also indicate that failure rate with SD cards is far higher than CF cards.

Pay attention to negative customer reviews

Some negative customer reviews are frivolous because they are rooted in user error, or because they concern themselves with delivery rather than the product’s quality or performance. However, negative reviews are generally more significant than positive reviews.

If you think that one shouldn’t focus on the negative while the vast majority of reviews are positive, consider that on Amazon, the average rating for a product is 4.4 (out of 5) as found here by analyzing 7 million reviews. Even a product with an average 4.0 rating (4-star) is below average. The large majority of products are rated above 4.0, so the difference between a great product and a subpar product is less than 1 (star) on average. On the other hand, we’ve just seen that the number of 1-star reviews for different cards varies by a factor of four.

Consider dual card slot for backup

If your camera has dual memory slots, the most obvious and foolproof way to prevent data loss from memory card failure is to set the camera to write to two cards simultaneously so that it creates a back up in real time. Now that memory card have become very affordable, you can buy two sets of cards with enough capacity to last you for your whole trip so you don’t have to reuse any card, and you always keep two datasets.

Dual memory card slots are standard in high-end DSLRs cameras, and after omitting them in their first two generations of mirrorless cameras, Sony has started providing them in the A9 and A7 mk3 series (a good example of listening to customers, since there were complaints about the single-slots in previous cameras, many of them from Canon and Nikon users), with the caveat that the second slot is UHS-I, so using simultaneous writing will negate the benefits of the faster UHS-II main slot. I was thinking of skipping the A7R3 generation and wait for the inevitable A7R4, but the incident prompted me to upgrade.

Recently announced full-frame mirrorless cameras from Canon and Nikon have been greeted with an inordinate number of Internet comments about their single memory card slot. We saw many claims that no professional would use cameras with a single memory slot.

But the fact is that some professional photographers refuse to use the second slot for back up even when their camera has one, and for specific reasons. To start with, when I was shooting the Canon 1Ds series, I did not set up the cameras to simultaneous write, possibly because the size of the memory cards available back then made it mandatory to perform daily backups. More recently, Lloyd Chambers uses his dual-slot Nikon as single slot SD cameras because he is annoyed by the camera defaulting to the wrong card – this reminds me of Ted Orland’s aphorism “Owning more than one lens assures that you will always have the wrong lens on the camera for any given picture” – while Thom Hogan uses his dual slot Nikon as a single slot XQD camera because the SD slot slows down the camera. Colby Brown thinks that “there is no point in making two copies of your SD cards” and accordingly sets his Sony A7R3 to auto switch as he estimates he has a higher chance of missing a shot because of a full card than a SD card failure – what I used to do with the 1Ds.

I wonder if those statements about the rarity of card failures do a disservice to less experienced folks, because what isn’t clearly disclosed is that, although those photographers apparently don’t fear card failure, they also have extensive backup strategies using hard drives.

Have a solid backup strategy

If there is one thing that I wish others learn from my misfortune, it is that a solid backup plan is necessary. You want multiple copies of your data in multiple places. There are quite a few ways to go about it.

Several brands now offer ruggedized portable drives. SSDs, which have fallen in price, are much less prone to damage than HDDs. Using drives for backups, you can do a daily (or even more frequent) backup, and have more than two copies of your data.

On the other hand, compared to the in-camera dual slots backups, drive backups are not in real-time, so you could possibly lose a day of data. More importantly, you need to remember and take the time to perform the backups – when maybe you’d just rather go to sleep. During that ill-fated trip, I carried a portable drive, yet due to a combination of fatigue, loaded schedule and complacency, I did not use it. Another reason was that my portable drive had experienced a glitch during the previous trip, forcing me to skip backups, which in turn broke my habit of making them regularly. This brings up the point that with drive-based backups, you have to carry more gear, which could also fail. Even if you carry several USB external drives, you still depend on your laptop for your ability to make drive backups.

With in-camera dual-slot backups, given the availability of huge capacity cards, you could shoot most trips on a single pair of cards, but if something catastrophic happened to your camera, you’d lose everything. Alternatively, you could use smaller cards, and once a pair of card is filled-up (or another threshold in capacity or time is reached), each of them can be stored at an independent location to minimize the risk of loss due to theft. While the second approach increases the chances of a problem because you have more cards to manage, it minimizes the adverse effects of problems.

For now, I have settled on an approach which I think provides me the most redundancy with the least effort: use the second slot of the A7R3 for real-time backup with a medium-sized card, plus do a daily backup on a single self-contained portable hard drive. My choice is the HyperDrive ColorSpace UDMA3 that I fit with a SSD drive. I generally prefer such a device to a laptop because they are considerably smaller and much faster to deploy for backup.

Eventually, my data loss was caused by my own neglect. I let my guard down and did only a few trips without backing up, and see what happened. This is just my experience and one data point, but I hope it’s been useful to you to read about it. What is your backup strategy? Did you experience a catastrophic memory card failure that you’d like to share?

Observing natural changes is a rewarding aspect of revisiting a place, as I did for the Tinkers Creek Gorge of Cuyahoga National Park – obscure and far from California. The water flow fluctuated, and you can see in this post which conditions worked best for each scene and how I adjusted my composition in response to the changes.

Bridal Veil Falls

Cuyahoga Valley National Park is unusual in that the National Park Service allows local parks within its boundaries to operate independently. Located in the northeast corner of the park, the Bedford Reservation is one such local park, managed as a city park by Cleveland Metroparks. Its highlight is Tinker’s Creek, the largest tributary of the Cuyahoga River, which has carved a deep gorge declared a National Natural Landmark.

Bridal Veil Falls is reached via a short walk (0.3 miles) from a trailhead along Gorge Parkway in the Bedford Reservation. Since it shares its name with the prominent 620-feet waterfall in Yosemite Valley, I was quite underwhelmed by the sight on my first visit. Although the narrow stream of Deerlick Creek tumbles down maybe 30 feet, it does so along a gentle slope rather than a steep drop, and the water flow was low enough that it would have taken some imagination to see a bridal veil there – the photo processing I did was unusually heavy. However, twelve days later, robust rains had changed the conditions enough that most of the shale rock was covered by whitewater. With that infusion of water, Bridal Veil Falls began to live up more to its name.

Sony A7R2, TS-E 24mm, 0.6 s. at f/16, ISO 50

Sony A7R2, 16-35 @ 19mm, polarizer, 3.2 s. at f/13, ISO 100

Sony A7R2, 70-300 @ 80mm, 0.8 s. at f/16, ISO 100

Viaduct Park

Viaduct Park, also part of Tinkers Creek, is under the radar, maybe because it is located just outside the boundaries of the national park. The steep forested gorge, which was not logged due to its rugged terrain, makes it easy to forget that Viaduct Park is surrounded on three sides by urban development – which is never too far in Cuyahoga Valley National Park. Tinkers Creek being a much larger stream than Deerlick Creek, the increased water flow and water opacity resulted in a reduced contrast between the whitewater and the smoother parts of the current. More water doesn’t necessarily make for a better stream photograph.

Sony A7R2, 24-105 @ 24mm, polarizer, 0.8 s. at f/22, ISO 50

Sony A7R2, 16-35 @ 21mm, polarizer, 0.8 s. at f/11, ISO 50

In addition to the creek, Viaduct park includes ruins of an old mill, an old viaduct, and a man-made tunnel through which the creek flows. If I had used the same composition for the high-flow image as for the low-flow image, a mostly cream-colored creek would not have balanced well against the dark forest as in the initial composition where the water was darker. However, I noticed that the higher flow created eddies not present with lower flow. A long exposure captured a circular shape which together with a dark rock formed a counterpoint to the tunnel opening.

Sony A7R2, 24-105 @ 40mm, polarizer, 0.6 s. at f/22, ISO 50

Sony A7R2, 16-35 @ 19mm, polarizer, 4 s. at f/22, ISO 50

Great Falls is not a tall waterfall, but its width and straight drop can make it more impressive than its a 15-foot height would suggest. On my first visit, no water was flowing over most of the drop. I used a tight composition to emphasize the section where most of the flow was occurring, also minimizing the visual extent of the bare rocks on the river bed, and using fallen branches to direct the eye towards the most appealing part of the waterfall, where low flow created texture. On my second visit, the waterfall was now spanning the entire 80 feet width of the outcrop. I framed the photograph with a wider angle of view again, this time to depict the almost the entire wall of cascading water, and now that the riverbed with filled with water that continued the flow of the waterfall, I also made it a prominent part of the composition.

Sony A7R2, 24-105 @ 60mm, polarizer, 0.6 s. at f/22, ISO 50

Sony A7R2, 16-35 @ 30mm, polarizer, 1.3 s. at f/16, ISO 50

Mammoth Cave National Park is generally a dry place. I explore a few out-of-the-way watery places within the park, while discussing some of the compositional strategies I used in those relatively plain scenes.

Sloan’s Crossing Pond

Water has made the longest cave in the world. The karst landscape of Mammoth Cave National Park is characterized by water being quickly channeled underground. Streams are few and little surface water is available. Sloan’s Crossing Pond provides important wetland habitat. Its textures are not found elsewhere else in the park, and it is also a rare spot with an opening from the forest.

How did I photograph Sloan’s Crossing Pond? I timed my arrival for the late afternoon. Reflections in water surfaces are an important part of their photographic appeal, and they are more clear when the water is in the shade. I walked the half-mile trail around the pond to look for different viewpoints.

My normal operating mode is upon arriving at a scene to first seek a wide-angle image. Usually, a quick glance is enough to tell if there is a compelling composition, and then it is a matter of framing and balancing elements, which I do instinctively. There is much more to landscape photography than the wide-angle image. The second phase involves more discovery and observation, as I identify the details of the grand scene that I find the most interesting. It takes time and a more deliberate mind to find those images, but the images reveal the natural features of a place in a deeper way and are often more personal. An easy way to start finding those images is simply to switch to a longer focal length and scan the scene. The following images are examples of such details found from the previous image.

Green River Ferry

The Green River crosses the entire length of the park, and from what I read makes for a wonderful float trip, but there are relatively few access points to see the river from land. The most well-known, because it is close to the cave entrances, is the Green River Ferry. While it is fun to make the crossing or observe the unusual ferry operate, the surroundings aren’t too natural.

The Green River Ferry is the only ferry which is currently active in the park, but there used to be more. Although they are now abandoned, their access roads lead to river viewpoints where you’ll find few other visitors.

Houchin Ferry

The river access point with the most possibilities is Houchin Ferry, located at the western edge of the park. The Houchin Ferry Campground is a delightful place to spend the night, as you can set your tent on a grassy area next to the river. It is much smaller and intimate than the campground located near the visitor center, and because it is out of the way, even arriving on a Friday evening without a reservation, several campsites were available. Instead of securing a site upon arrival, I proceeded to catch the sunset colors.

Mammoth Cave National Park consists of heavily forested gently rolling terrain, so most of river views are limited by the forest, but they could be remarkable with the right conditions. As Kentucky has high humidity, and the many cave openings provide quite a bit of cool air, there was a good possibility that that fog would roll in at night to burn in just for sunrise, but I did not get lucky that morning. Right at the landing, a power line intrudes into compositions. Looking for alternatives, from the campground, I followed a narrow user trail leading to the riverbank, which turned out quite muddy and slippery. Besides ruining shoes, the mud doesn’t make for a great foreground, especially with the footprints. My first attempt to mitigate the issue was to hide it with river driftwood.

I then tried a second solution, which was to make the foreground less prominent. I placed its closest elements further away by placing the horizon higher in the picture. The sky, now occupying a larger part of the image, needed something more, and I partly filled it with an overhanging branch. A frequent problem when “framing” pictures with trees is that they bear little relationship to the rest of the image. I avoided that by finding a branch in the water to echo it, and I made sure it was clearly defined by placing part of it against the reflection of the sky.

Dennison Ferry

Dennison Ferry is at the end of a gravel road, and lightly visited. Unlike Houchin Ferry, it is not a boat launch, but rather a canoe launch. Steep stairs lead to the river, whose bank at this point is steep, precluding wandering around, unlike at Houchin Ferry. The only place to stand at is at the bottom of the stairs. During the first exposure I made, light rain created texture on the water’s surface, and I made it the main subject of the picture.

After the rain stopped, the surface of the water changed to a mirror. I noticed a bit of fog rising in the distance, and liking the atmosphere it created, I framed a different image from the same viewpoint.

Of course, visitors come to Mammoth Cave National Park primarily to check out the cave. Besides, what interested me the most in the 80 square miles of rolling hilly country are the sinkholes and limestone cliffs, which offer subtle clues that a vast cave system lies beneath. However, this time I checked out those wet places in a quest to explore each corners of our national parks. Even if you don’t find them remarkable, I hope you’ve enjoyed the peek at my approach to composition.

The polarizing filter is such an essential tool because it allows you to control reflections, darken skies, cutting haze, and increase color saturation. Although using 24mm had become second nature for me to use, with a polarizer on the Sony FE 24-105 f/4, I found myself zooming in a bit until the dark corners (as illustrated in the previous post disappear. After a few months of that, I tried to look for a better solution, and it meant trying new filters.

The problem of vignetting by polarizer has long been recognized by the manufacturers. Several offer “wide-angle polarizers” which are designed to mitigate the problem by reducing the filter’s thickness. The thinnest of those filters have no front filter threads. This prevents the use of lens caps for protection, something that I was not fond of in my past use of similar filters.

My Tiffen polarizing filter is reasonably thin and inexpensive. In an extensive comparison, Roger Cicala found that all the tested polarizing filters worked equally well. In order to solve the vignetting problem, I tried several thinner polarizers. One of the difficulties is that the filter’s thickness is not an information is not readily provided by manufacturers, at least in a consistent way. Some provide overall thickness, while others do not count the lens threads. I was able to get this information only after buying the filters, when I measured the thickness with a digital caliper.

The filters

Here are the circular polarizing filters considered in the comparison, all in 77mm:

• Tiffen ($34) 7.7mm
• Firecrest Super Slim ($79) 7.4mm
• Hoya Pro1 Digital ($110) 7.4mm
• Firecrest Ultra Slim ($122) 4.6mm, no front threads
• Heliopan Slim ($146) 7.4mm, no front threads

A few things are worth noting about some of those filters. The Firecrests come in a useful box (similar to the B+W but larger) with hand-written “quality control” inscriptions. They have a distinctively warmer tone than others, which you may or may not like. In addition, they turn out to transmit about 2/3 f-stop more light than the other filters tested, which generally is a desirable feature. The Heliopan Circular Polarizer Slim Filter is made of brass unlike the others which are aluminium. Despite it having no front threads, it is slim only compared to the regular Heliopan, which is one of the thickest filters I own at 12.7mm – not included in the comparison because it is not a 77mm.

In addition, I also tested two “protective” filters:

• B+W UV 5.8mm
• Hoya UV 6.3mm

The comparison

I photographed a scene from the Evergreen Valley where I live at the 24mm setting on the FE 24-105mm with each filter. First, for reference, here are uncorrected images with the bare lens, showing vignetting respectively at f/4, f/8, f/16:

Uncorrected f/8 images, ordered from the least vignetting to most vignetting:

B+W UV (5.8mm)

Hoya UV (6.3mm)

Firecrest Ultra Slim (4.6mm)

Firecrest Super Slim (7.4mm)

Hoya Pro 1 (7.4mm)

Heliopan Slim (7.4mm)

Tiffen (7.7mm)

Uncorrected f/4, f/8, f/16 with the least vignetting polarizer (Firecrest Ultra):

Uncorrected f/4, f/8, f/16 with the most vignetting polarizer (Tiffen):

Lightroom-corrected f/4, f/8, f/16 with the least vignetting polarizer (Firecrest Ultra)

Lightroom-corrected f/4, f/8, f/16 with the most vignetting polarizer (Tiffen)

Results and conclusion

Here are the findings from the filter tests:

• The protective filters did not result in more vignetting than the bare lens.
• For the polarizing filters, there is a strict correlation between the filter thickness and the amount of vignetting. From less vignetting (better) to more vignetting (worse): Firecrest Ultra, Firecrest Super, Hoya Pro1, Heliopan Slim, Tiffen. In view of that, it is regrettable that the thickness information is not provided by manufacturers, but given the choice between a regular model and a slim model of the same brand, the additional expense of the slim model is justified. I am not sure why the Firecrest Ultra vignettes slightly more than any of the protective filters, despite being slimmer.
• The worst vignetting occurs wide-open. It improves by stopping down, but does not go away.
• Vignetting from the polarizer is not entirely corrected by applying a profile in Lightroom, but in the best case (Firecrest at f/16) it is barely noticeable.

I chose to keep the Firecrest Ultra. Leaving the lens hood permanently provides enough protection for me. If a filter without front threads is not acceptable, the Firecrest Super is a good choice.

Update June 2024: The Firecrest Ultra has turned out to be a very fragile filter. I have cracked two by leaving them in a pant’s pocket. Not only that, but they are currently quite difficult to replace. The only source I have found on the Internet is the UK site of the manufacturer. They ship worldwide fast, but at a cost!

Modern zoom lenses are quite amazing compared with what was available a few decades ago. A lens such that the Sony FE 24-105mm f/4 G OSS (reviewed here) offers a great combination of sharpness and versatility in trans-standard focal lengths, all in a relatively compact package – consistent with the appeal of their A7 series cameras, Sony emphasizes that this lens is the smallest and lightest of its class.

However, compromises still have to be made for such lenses. In typical fashion, Sony designed the lens with a tight image circle, meaning that the area illuminated by the lens barely covers the full-frame sensor. This results in strong vignetting, the darkening of corners and edges compared to the image center. To keep the lens compact, a 77mm filter thread was chosen, instead of 82mm. As a consequence, corners are further darkened by a thick filter as the filter cuts into the field of view of the lens.

The problem is particularly noticeable with a polarizing filter, because even a thin polarizing filter is still quite thick compared with other filters. Here is a typical example at 24mm, f/11, from the Cadiz Dunes of Mojave Trails National Monument:

The image was photographed on the Sony FE 24-105 f/4 lens on the Sony A7R2 with a Tiffen filter which, at 7.7mm, is relatively thin – by comparison, my B+W and Heliopan polarizing filters are respectively 11mm and 12.7mm thick. This seems to contradict Ken Rockwell’s statement about that particular lens:

There’s no need for thin filters. Go ahead and use your standard rotating polarizer and grad filters. I can use fat 77mm filter and get no vignetting at 24mm on full-frame

It’s not that Rockwell’s website is “a work of fiction, entirely the product of [] imagination and personal opinion”. Rather, he shoots exclusively in JPG, in which case by default the camera automatically corrects vignetting by applying a profile. For example, on the Sony there is a setting “Lens Comp.” and within, the default for “Shading Comp.” is “Auto”. With that default, the image that you see in the viewfinder or LCD has the correction applied, so you don’t even see the full extent of the vignetting. Personally, I prefer to turn “Shading Comp.” to off to see what I get in the RAW file, because such a correction is not applied to a RAW file. With “Shading Comp.” off, the vignetting within the bare lens (example here) is barely noticeable, however, the vignetting caused by a polarizer is evident.

Provided that Lightroom has a profile for your lens and camera combination, it does the correction automatically on the RAW file much the same way as the camera corrects JPGs. “Enable Profile corrections” in the Lens Corrections panel is checked by default. By the way, acquiring the FE 24-105mm forced me to upgrade my operating system, which broke many applications! For some obscure reason, the version of Lightroom I was using was not supporting that particular lens profile, so I needed to upgrade to Lightroom Classic CC. In turn, Lightroom Classic CC was not supported by the version of Mac OSX I was running.

By default, the profiles corrections consist of both distortion and vignetting (both sliders are at 100%). We are concerned here with the vignetting, so let see what happens if we move the “Distortion” slider to 0% and keep “Vignetting” at 100%:

This was not too helpful!

Getting “Distortion” back to its default 100% definitively improves things. This tells us that the improvement is due to the corners being cropped by the distortion correction rather than being brightened by the vignetting correction. If you like to compose tightly to the edges and not crop like I do, this can alter your compositions. Parts of the image’s exposure are also affected, with fewer highlights being recoverable, as can be seen in the upper corner.

For all those reasons, I leave “Enable Profile corrections” off in my own default, and instead apply it on individual images on a case-by-case basis. Most of the time, those corrections are not necessary. A slight amount of vignetting can be desirable, as it helps keep the eye in the image center, and distortion is only noticeable in landscapes with a straight horizon line. Another problem with applying those corrections by default is that they slow down Lightroom quite a bit.

By the way, you can notice in the Lightroom screenshot that “Remove Chromatic Aberrations” is checked. By Lightroom defaults, it is not checked, although of the two boxes that is the one that I enable in my own defaults because there is no downsides to it. In my opinion, Adobe got it wrong with their two defaults here.

Even with the distortion correction, dark corners are still somehow present and distracting. To finish the image, I needed to crop manually to get rid of them.

The other problem with using a polarizer on a wide-angle lens is that this can result in uneven polarization of the sky, since the polarization depends on the angle relative to the sun. The polarizer’s effect is strongest a 90 degrees from the sun. In the image, the polarization is most intense at one-third of the frame which was at 90 degrees, and as you get over or under that angle, the effect is reduced. The overall effect is a darker blob, which is quite noticeable in a cloudless sky. It is quite easy to miss that effect in the viewfinder. In this case, I previously had the zoom set at a longer focal length, where the polarizer was just fine, and didn’t pay enough attention as I set the zoom to 24mm. Be careful with polarizers on wide-angle lenses!

Totally fixing the uneven sky in processing is very time-consuming, since it involves rebuilding the sky in Photoshop – this discussion gives an idea of the techniques involved. I quickly mitigated the problem in Lightroom by reducing the contrast and brightening shadows on a gradient selection of the sky, resulting in the finished image above.

I have found the adage “Bad weather makes for good photographs” generally true. Cloudless days are not what I look forward to for daytime photography. There are two approaches to finding favorable weather conditions for landscape photography. You can go to a location based specifically on the weather forecast – the best is little-known Meteoblue because it includes cloud cover at different altitudes. Instead of looking for weather, you can wait for it. I somehow used both approaches on the same day in Shenandoah National Park.

I’ve heard some photographers and workshop leaders drive hundreds of miles in a day to place themselves at a scene with “epic” light for a sunset. I am not a big fan of that approach. When I am out in nature, the experience of enjoying the environment is paramount. Spending hours rushing in a car is not a way to do it. Although I admire those epic pictures, I feel that there is more to nature than that, and I’d be missing a chance to make those discoveries if I was focussed only on getting the epic, pre-conceived, picture. Rather, my approach in looking for weather is to make smaller adjustments to my itinerary based on weather conditions, as well as looking for subjects appropriate for the weather.

The day had started sunny, but as I drove down the Skyline Drive in Shenandoah National Park and the road climbed to higher elevations, clouds started to roll in, obscuring views. I abandoned my plans for a long hike to a mountaintop, switching my interest to the forest. I love the depth and atmosphere that fog brings in any forest, and the thicker, the better.

Any place in the park is a good starting point to explore the forest. Whenever I encountered a section of dense fog along the drive, I would look for the nearest pullout or trailhead. Opposite the Little Hogback Overlook, beautiful ferns cover the ground, and so do they along the section of the Appalachian Trail accessible from the Hogback Overlook. When the fog started to dissipate, I’d go back to the car, and then keep driving until another area of dense fog.

By the time I arrived at Big Meadows, the rain was quite heavy. Rainy weather is excellent for photographing in the forest, but having photographed in rain with the benefit of fog earlier in the day, I didn’t feel like venturing in the downpour. I went to the Big Meadows Lodge for a picture of the inside. There, I found a cellular signal, which allowed me to figure out that the rain might stop. If the weather was to break, Big Meadows was a good place to be as, like the name implies, it is only one of the large open areas in the park. It is not a natural meadow but rather was cleared by the early settlers who were evicted to make room for the national park. While most of the park has been allowed to revert to a wild state, the NPS has chosen to maintain Big Meadows in its historic condition. Hawksbill, highest peak in the park, wasn’t too far, but the sound of thunder made it clear hiking high would be too risky of a proposition. I hung around and got out when the rain started to subside, and almost like on cue, the rainbow appeared just before sunset time.

Shenandoah National Park is a long and skinny park. The Skyline Drive runs its entire length of 105 miles. For most of it, the road is never more than a few miles from the park’s boundary on both sides. For this reason, besides the Appalachian Trail that closely parallels the Skyline Drive, there are few hikes in the park that aren’t short. The two best ones are the Old Rag Loop (8.8 miles, 2,400 ft elevation gain) which offers great scrambling and views, and the Whiteoak Canyon/Cedar Run loop (8 miles, 2,750 ft elevation gain). The later combines two hikes: the Whiteoak Canyon, a steep gorge with six waterfalls, has been called the “scenic gem” of Shenandoah, and along Cedar Run, two more waterfalls can be seen (NPS map).

Sony A7R2, polarizer, 24-105@57mm, 1/6s at f/8.0, ISO 200

During my autumn trip to Shenandoah, water levels were quite low for a waterfall hike, but last spring, conditions looked wet enough. Stopping at the visitor center for unrelated information, I thought I might as well ask for trail conditions. I was glad I did, when I was told that there was so much water that the connector between Whiteoak Canyon and Cedar Run was entirely flooded. Starting from Cedar Run, I might have never made it to Whiteoak Canyon. Instead, I changed my plans to hiking Whiteoak Canyon as a out-and-back hike. Always a good idea to do a last minute check with the rangers!

Sony A7R2, polarizer, 24-105@47mm, 1/3s at f/13, ISO 100

The Whiteoak Canyon Trail was quite wet from the start. Hiking became easier after I gave up on trying to keep my feet dry by avoiding the puddles – I had packed two pairs of shoes for my trip. The trail at first doesn’t follow the river, and the sun was still quite bright, so I didn’t photograph much until the bridge over the Robinson River, about two miles from the trailhead. Those bridges always provide interesting vantage points. I waited for a bit of cloud cover to soften the light and reduce the bright spots. Stream and waterfall images are all about the flow, and there is nothing like bright spots and shadows to interupt it. Compared with the photograph I made on the way back at dusk, the afternoon image with some hints of sunlight was more lively. Always a good idea to photograph in different light conditions!

Sony A7R2, polarizer, 24-105@39mm, 1/5s at f/16, ISO 100

After the bridge, the trail mostly followed the Robinson River, providing numerous compositions. Despite the length of the hike, I made sure to carry my tripod to be able to use a range of shutter speeds because water would be a primary subject. Photographing backlit made the leaves glow a fluorescent green. Since I wanted a slow shutter speed to blur the water, I had to wait for a lull in the breeze so that the leaves would not be blurred.

Sony A7R2, polarizer, 24-105@55mm, 1.3s at f/11, ISO 100

Although the cascades and waterfalls over Robinson River are the main attraction in Whiteoak Canyon, because of the wet conditions, water was flowing everywhere (including on the trail), creating ephemeral waterfalls of more delicate beauty than the main stream. Finding it more exciting to try and capture the transient, I made sure to look around.

Sony A7R2, polarizer, 16-35@23mm, 1/13s at f/11, ISO 100

Although a (variable) neutral density filter is ideal for photographing streams, it is not necessary. When the water is not in direct sunlight (a situation to be generally avoided), using a polarizing filter and stopping down gets me shutter speeds which are slow enough to blur the water. By varying aperture and ISO, it is possible to get a range of shutter speeds. Digital photography is particularly great for photographing water because you can experiment with shutter speeds and get immediate feedback. Some scenes look better with the water entirely smoothed, for some it is preferable to freeze the motion a bit, and often a combination of both in a single frame gives the most dynamic result.

Sony A7R2, polarizer, 16-35@16mm, 0.6s at f/8, ISO 250

As with all river hikes in Shenandoah National Park, you start from the ridge (elevation 3,385 ft), and the sound of the waterfalls entice you to hike further and further downhill. Remember you’ll have to work your way up. My favorite part of Whiteoak Canyon was an off-trail point at about three miles, elevation 1,750 ft, where several branches of the stream converge, forming a striking wall of waterfalls. In this photograph, the role of the polarizer was to increase contrast between the whitewater and the darker, more smooth sections of water, as well as increase the saturation of the leaves, rather than slowing down the shutter. Instead, at base ISO it resulted in a shutter speed which was a bit too low for the effect I preferred, so I raised the ISO.

Sony A7R2, polarizer, 16-35@16mm, 1.3s at f/8, ISO 100

On my way back, I detoured again to a rock ledge a short distance downstream from the bridge that provides a fine and vertiginous view of the upper falls, the only such large scene along the trail. In the afternoon, the waterfall was in the shade, but the surrounding trees and cliff were sunlit. Now the scene was softly lit, with a bit of fog providing some atmosphere. Even though a loop can be a more satisfying experience, revisiting a scene on an out and back trail even within a few hours pays.

Sony A7R2, 16-35@22mm, 25s at f/4, ISO 800

Further back, I had spotted a smaller river scene along the trail, but the afternoon light wasn’t favorable. Since it was now dark, I was able to illuminate the scene with a flashlight (described here) creating even light that appears at first like sunshine, but without the distracting shadows that would have resulted from sunshine. Darkness is not a reason to stop photographing.