Part One: Overview and Setup

LensAlign ruler revealing a good balance of depth of field on either side of the "0" mark | Canon 5D Mark II and 100mm f/2.8 macro lens

LensAlign Pro

LensAlign Pro

Earlier this month, Michael Tapes, president of RawWorkflow, e-mailed me because he saw my mention of  Datacolor’s product, aimed to complete with his LensAlign, the Spyder LensCal. He also referenced my account of working briefly with one of his company’s LensAlign Pro units this summer.  Michael then offered  to have a review copy of the LensAlign Pro sent to me for a closer look at this useful, if somewhat expensive, tool for properly calibrating the autofocus performance of a camera and lens.

While many manufacturers, including Canon, Nikon, Sony, and Olympus, now provide a means for end-users to calibrate the autofocus performance their prosumer and professional cameras via “autofocus microadjustment” (AFMA), none of them offered a tool for making these changes in a meaningful way.

LensAlign, then, is a platform for measuring the focus accuracy of a camera/lens combination.  One takes this information and then, if necessary, change the AFMA to shift the depth-of-field (DOF) so that it straddles the “0″ mark on the LensAlign ruler.  I will explain the overall functionality (although I do not intend to re-create the wheel–the LensAlign manual is on the company’s resources page) and consider why one would purchase such a product.

Debunking the 1:2 Depth of Field Myth

One of my initial concerns about this process is that I’ve heard for years–from anecdotes in photography books to some of my professors in graduate school–that depth of field is split in a 1:2 ratio around the point of focus.  That is, 1/3 of the DOF resides in front of where one focuses and the remaining 2/3′s lies behind it.  But LensAlign is designed to make the DOF 1:1 in front and back of the focus mark, which seemed to go against the grain.  Someone has to be wrong!

In researching this, I found a page of “Common photography myths,” and sure enough, number 14 was that DOF was 1/3 in front and 2/3 in back.  Essentially, DOF is mostly 1:1 front to back. However, as magnification is decreased (distance from subject becomes greater, i.e. farther back in the frame), DOF slowly grows to become closer to the 1:2 ratio of lore.  Then, as one focuses more towards infinity, the ratio becomes 1:∞, and hyperfocal distance is achieved.

The DOF ratio is not always 1:1, hence my saying “mostly.”  For example, with wide-angle lenses, particularly at close distances, the split becomes 40% in front and 60% behind the focus mark.  If you want to know more about how depth of field is calculated, you can find the formula here, and a list of known values for the circle of confusion of several camera models can be found here.

All this aside, I suppose if  one desires a 1:2 ratio for depth of field, one can use LensAlign to shift it to one’s liking, but it is a choice to alter the way that things are designed to work for one’s creative needs, it is not “the way it should be.”

Inside the Box & Initial Setup

Upon opening up the white clam-shell box, one finds the main unit, and a postcard detailing how to set up the following components:  its metal ruler, the “sight gate,” a magnetic on an adhesive strip for the “sight gate,” and four small, cylindrical magnets for use with the LensAlign’s “enumerator.”

The body of the LensAlign Pro is made of four pieces of extruded sheets of PVC that have been cut with a router.  Two vertical pieces are parallel to one another–the front containing a the focusing target that also holds the “0″ mark of the LensAlign ruler.  The rear vertical piece contains the bulls-eyes for making the unit square to the camera position.  A third piece is perpendicular to the other two, and keeps them straight.  All three are screwed into a square of PVC with a 1/4″ tripod thread in the center.  To keep the two focusing pieces parallel to one another, and to prevent them from bending in transit, a length of poster tube is packed between them for shipment.  Keep this piece in the event you ever want to travel with LensAlign Pro!

The ruler is a 9.5″ strip of metal, and it is how one gauges the focus calibration of their camera/lens.  Every other part of LensAlign is designed to work to support this ruler.  So, while in the initial setup the user concentrates on everything except the ruler, during the actual measurement stage, the ruler will be all that matters.

The sight gate is a plastic card that will slide up and down to open and close the bulls-eye’s in the front of the LensAlign.  However, to make it slide, one has to adhere a magnetic strip on the back of it (it’s a little odd that the user has to do this, and I assume it has something to do with labor costs to produce LensAlign, but it is a trivial matter to do it one’s self).

Once set up, the main unit could be used on a tabletop, but a 1/4 inch thread allows it to be mounted onto a tripod.  One could mount the unit onto a light stand, but for reasons explained later, it is better to have the unit on a tripod head that allows movements, such as a ball head.

The enumerator and the ruler

The ruler has various positions that it can be set to (and held in place by magnets). They range from virtually parallel to the imaging plane to almost perpendicular.  However, the LensAlign documentation recommends beginning with the ruler at the third position, and my experience seems to confirm that this is usually a good starting point.

That said, I tend to use the “flatter” ruler positions, 4 and 5, just as frequently as I do position 3.  The greater the distance between the camera and the LensAlign, the flatter the ruler should be, as it makes it easier to distinguish the boundaries of the DOF.

LensAlign Enumerator: Fresh from the factory (left), set up with a neutral AF adjustment at a distance of 5 feet from target (middle), and set with a negative 3 AF adjustment at a distance of 25 feet from target (right)

The enumerator is not essential for the operation of the LensAlign, but it is a handy feature if one makes a series of images at different AFMA’s and reviews them later on a computer screen, as it provides a visual record of the AFMA setting.

Working with Test Photos

That is to say that there are at least three ways to work with the test photos of the LensAlign: review them immediately on the rear LCD screen and tweak the AFMA on the spot, shoot tethered to a computer and review the pictures in the field with the camera, or make the initial (neutral AFMA) test picture and review it in-camera, and then make a series of incremental AFMA changes and download all of the images to a computer for review.

The advantage of the first option is that it might be a little faster, but the latter two choices tend to be more accurate as they involve larger computer monitors for viewing the test pictures.  Hence, if one chooses to review the images on the computer, the enumerator provides an easy reference for the settings used.

Personally, I have tried all three options, and find that they all have their time and place.  I don’t have a setup like Joe McNally, but if I did I would shoot tethered for everything.  (To make his setup, one needs a tripod, a Manfrotto dual-head “bar” to mount on top of it, a laptop table to mount on one side of the bar, and a tripod head of some kind for the other.)  Since I don’t have those parts, I chose to tether only when I was indoors.  Outdoors, I find that I can roughly eyeball whether I’m front or back-focused on the rear LCD of the camera.  Once I believe that I have found the proper AFMA for the lens I’m using, I’ll bring it inside and check it on my computer monitor.

A Word About Placement

Orienting the LensAlign towards the camera and lens. Note that I have taken the unnecessary, but helpful step of placing a bubble level on the LensAlign platform.

Before one can make a test image to determine any front or back-focus, one must first place the camera and LensAlign on separate platforms, space them, and make them square to one another.

It really helps to have the LensAlign either indoors with indirect lighting, but I suppose if you’re feeling motivated, you could set up flashes to illuminate it.  When using it outdoors, be sure to place the unit in shade, or use it on an overcast day.  The ruler is extremely reflective, and direct sunlight turns it into a blinding mirror.  You want the light falling on the ruler to be even for easy-reading!

The online distance tool helps to determine how far apart the camera should be from the LensAlign, as well as how “deep” the DOF should be with a given focal length and sensor size.  This distance can also be calculated manually by multiplying the focal length by 25.  So, for a 400mm lens, one would need to be 10,000mm from the target, or 32.8 feet.  TIP: A 50-foot reel measuring tape will be very helpful for this step if you own telephoto lenses!

I have also tried calibrating lenses at their minimum focusing distance (MFD) instead of the recommended distance.  In a way, it’s easier because one just moves back as far as one needs in order to get LensAlign in focus.  Additionally, the target will be larger in the frame the closer one is to it, making the markings on the ruler easier to read.  In fact, in my first draft of this review, I suggested that MFD might be all you need.  But then I was proven wrong!

My advice for telephoto lenses 400mm and longer would be to first make an adjustment for the MFD, and then back up to the suggested distance to confirm the adjustment.  I recently calibrated a 500mm f/4L with my 1D Mark III and discovered that the adjustment made to counteract back-focus at MFD was too extreme, and actually caused front-focus when set at 41 feet from the target.

Aligning LensAlign

Making the lens and target may involve tilting the camera downwards or upwards at the target, depending on height differences. In any event, it is not necesary to make them level to the world, but to make the two devices square to each other.

The first step you should take is use your tripod head (you have to be tripod-mounted) to position the camera at the LensAlign.  Aim the center focusing point (the most accurate of the focusing points, irrespective of camera brand or model) at the center bulls-eye of LensAlign.  Then, run behind the LensAlign, and looking through the “peep hole” through the center bulls-eye, lock it in position so that you are looking through the center of the lens.  Often, you can see light coming in through the viewfinder of the camera through the middle of the lens, which can help to make it just right.Return to the camera, and switch on Live View (or take a picture and review it on the LCD) and zoom in to see if the center bulls-eye is truly centered in its viewing port.

LensAlign's bulls-eyes - note the (close to) bilateral symmetry. At shorter focal lengths, not every bulls-eye will be centered, and the target of greatest importance is always the bulls-eye in the center

An aside: To my knowledge, every camera that has the ability to make AFMA changes also has live view.  However, the LensAlign is helpful even with cameras that do not have this feature, as it can confirm whether or not there is a focus problem with a given lens.  If there is a problem, the user can take the data and decide whether the front or back-focus exhibited warrants sending the camera and lens to the manufacturer to be professionally calibrated.

While the LensAlign manual does explain how to make the unit square to a camera, I have found that it helps to pay attention to more than the center bulls-eye, but to consider all of them.  Especially at shorter distances and wider focal lengths, the other four bulls-eyes will not be centered in their viewing ports, but making sure that they are symmetrical on opposite sides of the center bulls-eye ensures that the plane of the LensAlign is parallel to the camera’s sensor plane.

Corrections and A note about Keeping Things Level

In the original version of this review, incorrectly stated that the body of the LensAlign is made from black masonite.  While it has that apperance, it is actually made from PVC.

Additionally, in the original posting I emphasized that keeping both the camera and LensAlign level, by means of a hot shoe bubble level or otherwise, would be a good idea.

After speaking to Michael Tapes directly, I stand by thinking that it’s helpful to keep things level, but agree that it is not essential for the operation of LensAlign. Hence, I have removed it from the section about aligning the focus target to the camera lens.  In fact, he told me that the absence of a level “wasn’t a casual omission.”  He believes that adding a level the product would make it appear to the user that being level was part of making the camera and focus target square, when it is not, in fact, a necessary step.

A level camera and a level focusing target will not change the DOF–it’s a constant.  Camera lenses do not draw rectangular images, but circular ones, so how the camera is rotated on the lens doesn’t change the DOF.  However, I find that leveling both helps in reviewing the images from LensAlign, because my beleaguered brain doesn’t like to look at picture of a ruler that is skewed at a 5° angle.

The LensAlign functions the same whether level or not, but you may find you prefer to read a straightened ruler as opposed to a tilting one!

For me, I find the lack of a level to be my lone frustration with the product.  I say this because it is downright annoying to constantly walk back and forth between the camera and the LensAlign while trying to remember to take the level with me each time.  When I walked 41 feet out to the 500mm lens that I was trying to make square to the focus target, and realized that the level was still sitting on the back of the LensAlign, I really started grousing!

But to reiterate: it is not necessary to do this extra step of leveling the two devices, I just prefer to work this way, and I think many photographers who will take the time to change the AFMA in their cameras will feel the same as I do that looking at a level measurement tool is easier than looking at one that is tilted, even if the data is the same.

Read part two of the review now!

Tags: equipment, LensAlign, Reviews
Posted in Reviews | 5 Comments »

Washed ashore, North Beach, Racine, Wis. | Canon 1D Mark III and 70-200mm f/4L IS lens @ 70mm | Exposed 1/250 sec. @ f/8, ISO 400.

I am presently writing a full review of RawWorkflow’s LensAlign Pro–a product I briefly worked with in June–so I’ve been busy adjusting the autofocus performance of my cameras with several of my lenses.  While it’s nice to look at a test photo of a focus alignment ruler and see your depth-of-field falling exactly where it should, there’s no substitute for confirming that the changes are for the better than to go out and make some images!

Note that both images in this post were created at f/8, but for different reasons.  In the image of the feather and debris on the beach, I wanted to make sure that the feather, which was almost flat against the beach, as well as the taller matter, would all be rendered in sharp focus.  However, for the photograph of the Ring-billed Gull below, I was using a Canon 1.4x II teleconverter, and I always try to stop down either 2/3 or a full stop from wide open in order to eliminate the vignetting introduced by the teleconverter, and to counteract any loss of sharpness.

Ring-billed Gull stretching, North Beach, Racine, Wis. | Canon 1D Mark III and 400mm DO IS lens with 1.4x II TC | Exposed 1/1600 sec. @ f/8, ISO 400.

Tags: birds, Canon 400mm DO IS, Canon 70-200mm f/4L IS, EOS 1D Mark III, nature
Posted in Updates | No Comments »

Datacolor SpyderChekr

Earlier in the month, Datacolor announced the SpyderLensCal, a product that was a clear imitation of the LensAlign autofocus microadjustment calibration tool from RawWorkflow.  Although it would be unfair to dismiss it outright as I’ve not actually worked with one, SpyderLensCal has several shortcomings compared to the LensAlign, including the inability to confirm that it is parallel to the camera.

Now, Datacolor has announced another product that seems to be an homage to a competitor: the SpyderChekr is aimed squarely at X-Rite’s ColorChecker Passport that I reviewed early this year.  The gestalt of the two products is the same: a plate of pigmented squares with known color values is embedded into a hinged case to be carried on location and photographed under the lighting conditions for the photo shoot.  Then, in RAW conversion either in Adobe Camera RAW or Adobe Lightroom, the photo of the unit is exported to the manufacturer’s software to be profiled.  That profile is then applied, at the user’s discretion, to the photos from the shoot.

However, there appears to be a fundamental difference in the way that Datacolor approached the SpyderCheckr compared to the SpyderLensCal: price point.  The SpyderLensCal is priced to undercut what has been the only game in town, the LensAlign, and to do so by $20 at an MSRP of $59.00.  To get down to that price, Datacolor clearly cut corners.  The SpyderCheckr, on the other hand, has been announced at $129.00, a full $30.00 more than the X-Rite product, and actually builds upon the idea of camera calibration by adding some interesting features.

The Datacolor product can be mounted on a tripod, whereas X-Rite’s hinged case was designed to support the color charts, like the fold-out arm in a picture frame.  A tripod thread adds versatility.  The color charts in Datacolor’s SpyderCheckr can be flipped over to reveal gray-balance targets.  That means that one could photograph the necessary color chart for camera calibration and a greay-balance target in the same frame.  X-Rite’s passport does include a gray-balance target, but it is mounted opposite the color charts, so they could never be together in the same photograph.

SpyderChekr's reversible (and replaceable) charts | rendering from Datacolor's SpyderCheckr user manual

Furthermore, that Datacolor’s color charts are reversible means that they are also user-replaceable should they begin to fade or become scratched.  The Datacolor Web site for the SpyderCheckr suggests that these spare color charts will be available for sale “early in 2011.”

There are some differences between the SpyderChekr and the ColorChecker Passport that do not make one superior to the other, but are of interest just the same.  First of all, the Datacolor product is physically larger.  In fact, I believe it is close to the size of a traditional Gretag-Macbeth (now X-Rite) color chart.  This means that it won’t fit in a camera bag quite as easily.  Even when I’m not working on a studio shoot, I always carry the X-Rite product in my bag, because it is small and easy to hand-hold.  However, I can’t mount it to a tripod (or light stand), so that may be irrelevant for some.   Another difference is that X-Rite’s ColorChecker Passport creates DNG profiles for its camera calibration, whereas Datacolor’s software manipulates the Hue/Saturation/Lightness sliders in Adobe Camera RAW / Adobe Lightroom.  I’m inclined to believe that DNG profiles may be a better route, but I’m not a technical expert on this particular part of Camera RAW / Lightroom, so I will just offer that they are different routes to achieve the same objective: more faithful color.

Many of the design and operational differences between the two products are actually outlined in a chart on page six of Datacolor’s SpyderCheckr manual, although you’ll notice a natural bias for their own product!

Finally, if you’d like to watch a video explaining the features and the supporting software for Datacolor’s SpyderChekr, you can watch this PhotokinaTV’s video featuring David Tobie, product technology manager for Datacolor:

Tags: Datacolor SpyderChekr, equipment, Previews, X-Rite Color-checker Passport
Posted in Announcements | No Comments »

Meadow Lake Wind Farm, Wolcott, Ind. | Canon 5D Mark II and 70-200mm f/4L IS lens with circular polarizer | Exposed 1/13 sec. @ f/20, ISO 50 (+2/3 EV)

This past Sunday, as I was driving up I-65 North through Indiana on my way up to Racine, Wisconsin, I came upon the Meadow Lake Wind Farm.  I exited off the interstate and pulled into the parking lot of a gas station and Dairy Queen in one (where else can you can get dip cones with a faint aroma of gasoline?) and was impressed by the imposing structure of the wind turbine directly in front of me.  It was about 4:00 in the afternoon (Indiana is on Eastern time) so to help the color I screwed on my circular polarizer.  The sun was coming from about 90 degrees, so I knew I could get optimum polarization if I wanted it, but I really only wanted a touch.  It’s easy to make the sky turn blue-black if you’re not careful!  The tricky part was conveying a sense of movement.

My tripod was packed away in the trunk of my car, so I braced myself against a car door and switched to shutter speed priority.  I knew this would mean that a slow shutter speed, in broad daylight, could only be achieved with a low ISO and a tiny aperture.  The polarizing filter was already helping by knocking the light back two stops, so a small aperture would finish the job.

Thing is, while the 5D Mark II has built-in sensor cleaning, it’s also the camera that I use the most, and I knew that it had some dust spots that needed to be cleaned off.  So, to get a slightly less dust-revealing aperture (originally f/32), I knocked the ISO down to 50.  Now, if only the wind turbines in the background had been slowed down enough to show that they were moving, too!

Tags: EOS 5D Mark II, landscapes, pictorial, wind energy
Posted in Travel | No Comments »

Dropbox sits in the Windows 7 task bar

There’s one thing that I use with all of my computers (laptop, desktop, netbook, and even my iPod Touch) that makes the task of moving files around so painless that I’m not sure how I ever got by without it.  I’ve been using it for almost two years, and I continued to be baffled by how uncommon it is to find other users.  I’m talking, of course, about USB flash drives Dropbox.  It’s a way of having your files on your computer and in the Cloud all at the same time.

Read More after the jump!

Read the rest of this entry »

Tags: Reviews, software
Posted in Reviews | 4 Comments »

A Review of the Canon 135mm f/2L Lens

Weighing the options, Crown Honda, Durham, N.C. | Canon 5D Mark II and 135mm f/2L | Exposed 1/400 sec. @ f/2, ISO 400.

Preamble

The Canon 135mm f/2L lens is one of the more highly regarded lenses.  Together with the 35mm f/1.4L and the 85mm f/1.2L II lenses, it is popular among photographers for the special “look” that it gives images, and in a way, deservedly so.  The subject is easily isolated from the rest of the frame, making the challenge more of creating a pleasing composition than of worrying if the image will be too “cluttered.”  However, my experience with the lens for a week, courtesy of Canon Professional Services, has left me wondering if the reputation that the lens built for itself is as deserving in the age of 21+ megapixel camera bodies and circular aperture blades.  While it surely is effective wide open, the lens appears not to be as sharp as it could be in all circumstances, and stopping the lens down reveals the potential for distracting backgrounds, as my comparison between this lens and the 70-200mm f/4L IS zoom lens will reveal later in the review.

Looking its best

Without question, wide open the lens creates a very compelling and desirable effect for its focal length.  The advantage of telephoto compression means that the foreground can be easily distinguished from its background, and the lens’ wide maximum aperture (f/2) helps to obliterate what’s left of said background.  With such a fast telephoto lens, foliage only a few feet behind a subject can be turned into an almost seamless background of solid color.  Furthermore, the glass is cut in such a way that some subjects seem to literally pop out of their two-dimensional planes, giving them very “3d,” almost Zeiss-like appearances.  Indeed, the color and the contrast that come from this lens is impressive.

Below are a few examples that illustrate some of the very best traits of this lens:

Duke Gardens, Durham, N.C. | Canon 5D Mark II and 135mm f/2L lens | Exposed 1/6400 sec. @ f/2, ISO 400 (0 EV)

In this example, the background flowers are no more than inches behind the central subject (with the yellow flowers about two feet behind).  And at 100%, the sharpness is quite stunning despite being made wide open.  I must note, however, that images made of subjects physically closer to the lens (particularly in the 3-5 foot range) appear sharper than do those made in the 10 foot to infinity range.  However, as I do not have any lens AF Micro-Adjustment equipment on hand, I can say only that this is a perception, but I cannot say an incurable one–AF Micro-Adjustment very likely would address this issue.

More after the jump!

Read the rest of this entry »

Tags: Canon 135mm, equipment, Reviews
Posted in Reviews | 1 Comment »

Datacolor SpyderLensCal - image from datacolor.com

Have I seen this before?

Yesterday, Datacolor announced a new addition to its “Spyder” line of calibration tools: the Spyder LensCal.  In many ways, it’s a spitting image of the LensAlign Pro I briefly reviewed here earlier this summer, and significantly less expensive at a suggested price of $59.00 (the LensAlign Pro sells for $180, while lensAlign Lite goes for $80).

While the design of the two systems is uncannily similar–an autofocus target with a ruler to the right–there is a significant difference between them: I see no way to confirm that the camera and the focus target are perfectly square to one another with the SpyderLensCal.  This is a significant advantage for the LensAlign, and in fact begs the question of just how accurate the Datacolor product could be if it is not feasible to make the target and camera square.  If they are skewed, so too will the out-of-focus areas, and the reliability of the product comes into question.

Furthermore, the SpyderLensCal offers no advice about the distance that should separate the camera from the target (only that it be “a fixed distance”), nor do they offer any clues about how much depth of field should be expected to be in front or fall behind the “0″ point of the ruler.  (While the depth of field for many lenses is pretty much 49% in front of where you focus, and 51% behind it, with extremely wide focal lengths, such as 16mm, the depth-of-field shifts to become 40% in front and 60% behind the point of focus.)  I can already envision many of Datacolor’s customers having to rely upon LensAlign’s Web site for their online “Distance Tool.”

Of course, it will be interesting to see how these models really stack up once they can be compared side-by-side.  The SpyderLensCal will begin shipping in about three weeks, during Photokina.

Tags: Datacolor LensCal, equipment, LensAlign, Previews
Posted in Announcements | 3 Comments »

"The impeded stream is the one that sings" (Wendell Berry), Eno River State Park, Durham, N.C. | Canon 5D Mark II and 135mm f/2L lens | Exposed 1.3 seconds @ f/32, ISO 100 (-2/3 EV)

For the past week I’ve been working with a Canon 135mm f/2L lens from Canon Professional Services.  I’ll be publishing my thoughts on this lens soon, but until then, a bit of a “teaser” from last evening.

Tags: Canon 135mm, EOS 5D Mark II, equipment, landscapes, Reviews
Posted in Reviews | 1 Comment »