Photos not level...advice?

[Ari]

Hi,
I have a problem with about 20% of the building photos I take.
I realize I may be nitpicking here, so bear with me.
I notice that in the photos shown below, one side of the photo is perfectly vertical, while the side closest to the camera is not; it is usually keystoning, indicating that the camera is not level.
Camera is a solid monorail, tripod is solid too.
I line up the camera's rail to the leg of the tripod that faces out, then level the tripod head (3-way), then level the camera.
I use a good bubble level, made in Japan, for these kinds of shots, and I follow what it says because the bubble levels on the camera don't read the same all the time, and I don't use the GG grid for aligning, as it may be installed at a slight angle.
What am I missing here, and what am I doing wrong? Why is only one side askew?
Thanks in advance.

http://i53.tinypic.com/oppro3.jpg

http://i56.tinypic.com/jaid8w.jpg

[Michael Clark]

I run into this once in a while to,keeping the back level (horizontal and vertical) I thought was essential and maybe it could be some distortion from the lens as I try to use the front movements the most. I hope some one will help you as I'm also interested.I have to say you do a better job than me.
Mike

[Doremus Scudder]

I do a lot of architectural shots with a field camera and, as much as I try, some shots are just not "square." It just happens. They need to be corrected when printing.

But before I go into the printing corrections, let's talk about getting the image as close to correct possible in-camera.

First, I think you are making a big mistake by not using the grid on your ground glass. Levels are notoriously inaccurate; the only true reference you have are the right angles and straight lines on the gridded ground glass. Do whatever inspection and adjustments are necessary to get your ground glass mounted properly, i.e., parallel to the film edges (image opening edges of your camera back). Even if the grid is a little skewed, the only thing that happens is that the image will not be exactly aligned with the film edges (this can happen due to the film not sitting really square in the holder as well...). The image, however, will be properly squared up as long as you've used the grid to align it.

When composing the image, try to use lines as close to the edge of the image as possible. A vertical line at the center of the image will be vertical as long as you have the camera leveled side-to-side, regardless of how much keystoning the outer lines have. The same applies to horizontal lines you want square. By the way, using the grid is the ONLY way to ensure that horizontal lines are parallel. No amount of leveling will help; gravity only works up-down :-)

My sequence is as follows: First level the camera roughly with the levels.

Next, go to work aligning the image with the grid (you don't even really need the levels from here on). Adjust level side to side by choosing a vertical line at the center of the image and adjusting the side-to-side level on the tripod head to get the line to line up with the grid. If you align with a vertical on the edge, you may be lining up with a line that is keystoned! (Your problem above, I believe.) If there is no vertical line at the center, use your shift movement to place one as close as possible to the center (don't pan, since the tripod may not be exactly leveled and you will just introduce error). I also try to check a horizontal line at the center of the image as a control. Like the vertical at the center, a horizontal line running through the center of the image will be correct as long as the camera is correctly leveled side-to-side.

Next, align verticals at the edge of the image by tilting the tripod head (front-to-back) to get lines at the outside of the image to line up with the grid. If you don't have a vertical in the center, you can use the alternate technique of lining up two parallel lines at the edges of the image. This requires some see-sawing and back-and-forth, so try to use a center line when possible.

Check everything once you think you have the lines right. If things don't seem to click, make sure the vertical lines you are using are actually vertical... Keep in mind that the real world is often not as square as we would like. Old buildings are often really skewed due to shifting, settling, and sag. I often have to compromise to find the most pleasing composition.

If you need horizontals level and parallel, pan the tripod head till things line up with the grid. Again, choose lines close to the edge of the image. Recheck everything again.

If you are shooting transparencies and they are the finished product, then there is nothing else you can do to correct things, so spend time aligning the ground glass and really lining things up when composing the image. If the levels and the grid do not agree, always go with the grid.

However, the best laid plans of mice and men do oftimes go astray... and if you are printing, you can correct misalignments, even fairly severe ones, at the printing stage. Sometimes I am amazed at how far from square/parallel/perpendicular some of my lines can be, even after all the time I spend aligning... So, if you print digitally, just fix the image in PhotoShop... (I'm shuddering as I recommend that, however...).

For me, an analog printer, I prop up the easel to achieve square. My Beseler enlargers have a tilting lens stage that allows me to really adjust focus with the good old Scheimpflug principle. However, I am currently using Omega D enlargers that do not allow this, and find that the depth of field at f/16 or higher is adequate for very sharp prints with even fairly large propping up of the easel (even an inch or so).

If you are good with camera movements, propping up the easel to achieve proper alignment will be second nature. I keep a bunch of misc. pieces of cardboard, foam-core board, etc. around to use as props, which go under the appropriate easel corners. I've even had to adjust on two axes at once, getting one corner higher than all the rest (equal to using swings and tilts together) to get an image or two squared up. No problems with sharpness in the final print (I print up to 16x20 from 4x5 negs).

Hope this helps a bit,

Doremus Scudder

[Kirk Gittings]

I have struggled with this for my entire career. I haven't ever found on camera levels to be dependable or bubble levels in a hot shoe type of arrangement. The only thing I have found dependable is holding a really good level against the ground glass (with a properly mounted grid screen for determining side to side level), which I have done successfully for a couple of decades on numerous cameras including DSLRs (holding a bubble level against the screen.

[Leigh]

If you want some GOOD bullseye or bubble levels, take a look at the offerings from McMaster-Carr at www.mcmaster.com

They have a wide range, in various accuracies and sizes, for a few dollars each. Some are far more accurate than you could ever use with a view camera.

They also have an assortment of other level types, some of which are appropriate for this use. I got a four-quadrant level (called an 'angle indicator') that sits on the camera bed for front-to-back or horizontal, or can be pushed against the ground glass for vertical.

- Leigh

[Sirius Glass]

I use a six inch torpedo level across the top of my camera. Then I line up the verticals with the scribed lines on the ground glass.

[Robert A. Zeichner]

If you have an i-pod or an i-phone, you can down load an inclinometer for .99 that will give you extremely accurate readings. Just hold it against the standards and adjust for 0 degrees. Works very well for me.

[rdenney]

No small bubble level has the precision to support the accuracy you seek.

Consider this: You can notice a problem of just a few inches on a building column that lines up with the edge of the frame. 4 inches in 48 feet, for example, is an error of 0.7%. A bubble level that puts a 1/4" bubble in a 1" sight glass isn't going to do it. That 0.7% error is just 0.007" on that 1" sight-glass.

On surveying instruments old enough to be leveled using a bubble, the glass itself will be mounted in a frame about 6" long for a transit and maybe up to twice that for a Dumpy-style sight level. The tube in which the bubble floats is bigger and more precise, and most important of all, the surveyor will turn the level around for every confirmation of level to zero the error in both directions.

I have several carpenter's levels that are anywhere from 2 to 8 feet long, and they are all wrong. I know this because I turn the level around--any error should be identical in both directions to know if the calibration is correct. How is that 15mm-long spirit level on the side of the camera going to be better?

Measuring on the glass with a torpedo level that uses a tubular sight glass is much better, to be sure, but the only way to know if you have it right is by looking at the ground glass. That 4" in 48" error that is invisible on a tiny bubble levels on tripods or the equally tiny tubular level on view cameras will be quite visible in the ground glass using a loupe. An accurately gridded ground glass is essential for this sort of work.

Rick "who uses the bubble levels to get in the ballpark" Denney

[Leigh]

No small bubble level has the precision to support the accuracy you seek.
Rick,

You really need to do some research before you open your mouth.

McMaster-Carr (linked above) has bullseye levels with accuracies to 0.035"/foot, which equates to 1.68" in 48'.

They also have bubble levels with accuracies to 0.0003"/foot, or 0.014" in 48'.

There are numerous offerings with lesser accuracies (and lower prices).

- Leigh

[John Koehrer]

Rick,

You really need to do some research before you open your mouth.

McMaster-Carr (linked above) has bullseye levels with accuracies to 0.035"/foot, which equates to 1.68" in 48'.

They also have bubble levels with accuracies to 0.0003"/foot, or 0.014" in 48'.

There are numerous offerings with lesser accuracies (and lower prices).

- Leigh

My, how politely stated.
Ah, civility shot in the foot again.

[Leigh]

Ah, civility shot in the foot again.
The word 'civility' does not appear in my contract. :p

The problem is that people make blanket statements that are not correct.

Visitors retrieving threads from the archive frequently skim them, rather than reading every post. They see these generalities and think they're gospel.

The only way to counter that is with a comment that's likely to draw a visitor's attention.

- Leigh

[Merg Ross]

I have struggled with this for my entire career. I haven't ever found on camera levels to be dependable or bubble levels in a hot shoe type of arrangement. The only thing I have found dependable is holding a really good level against the ground glass (with a properly mounted grid screen for determining side to side level), which I have done successfully for a couple of decades on numerous cameras including DSLRs (holding a bubble level against the screen.

Kirk, excellent comments. Forget the on-camera level, bullseye or otherwise; they will only get you close. As suggested here, it is paramount that you have a properly mounted grid screen.

[Kirk Gittings]

Thanks Merg. In no other genre of large format photography is leveling more critical (well maybe copying paintings). When in doubt about leveling a VC?.......ask an architectural photographer.

[Ari]

Well, it's refreshing to hear that others, who have had years more experience than I, suffer the same difficulties.
I do have a very good level, but maybe I need to go digital; that 0.5˚ inaccuracy is likely the reason I have inconsistent angles.
And I'll double check the grid on my screen; I just received a Yanke screen from China today, very bright and super-thin, but one side was not true, and needed a little sandpaper.
Thanks for all the advice and replies.

[Kirk Gittings]

In 30 years of shooting architecture, I never saw the need for a digital level-one more thing that eats batteries.

[Ari]

In 30 years of shooting architecture, I never saw the need for a digital level-one more thing that eats batteries.

Kirk, can you suggest a good, accurate level, appropriate for this use?
I have four levels, not counting the camera-mounted levels, none of which are accurate enough (obviously).
Thanks

[Kirk Gittings]

I just use a high quality carpenters torpedo level.

[Leigh]

Levels are self-calibrating. Good ones are adjustable.

All you need is a smooth, flat, relatively level surface.

Make a 90° corner and tape it in place, to use as a positional reference.

Put the level in the corner and note the reading. Then turn it 180° and note the reading.

The two should be the same. If they are, then the level is accurate within its resolution limits.

If not, just make a mark at 1/2 of the error, and that is your new reference.

- Leigh

[Ari]

Thanks Kirk, thanks Leigh.

[rdenney]

Rick,

You really need to do some research before you open your mouth.

McMaster-Carr (linked above) has bullseye levels with accuracies to 0.035"/foot, which equates to 1.68" in 48'.

They also have bubble levels with accuracies to 0.0003"/foot, or 0.014" in 48'.

There are numerous offerings with lesser accuracies (and lower prices).

- Leigh

You know, Leigh, you could argue in a much more friendly way and make your point just as well if not better. You do not need to be uncivil just to rise above the noise--this is not that kind of place. Your first sentence served no purpose and was disproportionate to anything I said. I don't really care that you don't care about civility.

You mean 2308A4 in the McMaster catalog? Do you own one? Have you tested your ability to read it to its stated level of sensitivity? Can you eyeball the gap around the bubble to the extent needed? Can you look straight on the level from the camera position to avoid parallax? Is there a flat surface on the camera (not plumb--bullseye levels only work on a horizontal plane) that is reliable?

At 0.3% (0.035"/12"), it has twice the sensitivity as my example and is nearly twice the size (hmmmm). At 1-3/4" in diameter, it's not small at all compared to the levels on cameras and tripods. Or did you miss my use of the word "small"? That one, plus the two larger ones, are the only bullseye levels in their catalog that are more sensitive than the 0.7% value I quoted. All the bullseye levels 1" or smaller are worse. Their 1" level (the size used in my example) has a sensitivity of 0.157" per foot, or 1.3%--twice the error I used as an example.

And all the tubular levels that are more sensitive than my example are more than two inches long, except one (2209A71), and that one's fifty bucks. Even it is twice as long as the (adjustable) bubble levels on my Sinar. Which, by the way, I have calibrated, and are still not accurate enough for the purpose at hand.

My statement was that no small bullseye or tubular level will have the precision needed to avoid visible error conditions in architectural photography where a column is close to the edge of the photo. Having used levels of all types professionally, I stand by that statement. But those errors are much easier to see on the ground glass, which was my conclusion and recommendation.

By the way, levels are not self-calibrating. If you turn it around and find an error, you have to make a new mark or adjust the level (if it's adjustable). That sounds like manual calibration to me. Maybe you meant that they can be checked without a calibration fixture. If you are going to pick at people's words, you have to expect the same in return.

Rick "noting the difference between sensitivity, precision, and accuracy" Denney

[Ivan J. Eberle]

I made my own GG and drew grids with a Sharpie. As Doremus explained, grid lines do help with straightening out horizons and converging verticals

[E. von Hoegh]

Another vote for the grid lines. I have a small clinometer that I can hold to the GG and read through a built in magnifier, but this usually gets me in the ballpark, no farther.

My final check is on the GG and I nearly always have to tweak it.

As for the levels, I have a Starret machinist's level that is VERY sensitive but no more useful than the clinometer ( I've tried it) - plus it weighs about 3 pounds.

[ic-racer]

Technically your problem is not about the back being "LEVEL" it is a problem with the back "TILT" not being parallel to the buildings.

[Ari]

Technically your problem is not about the back being "LEVEL" it is a problem with the back "TILT" not being parallel to the buildings.

Of course; "level", to keep things simple, refers to both axes.
And according to my instruments, the back IS level and plumb; hence the mystery.
Next time out, I'll rely on the grid.

[Leonard Evens]

Technically your problem is not about the back being "LEVEL" it is a problem with the back "TILT" not being parallel to the buildings.

I agree.

One problem is that often you can't just check the detents for the standards. There is usually a slight amount of play which can induce a slight tilt or horizontal turn in the position of the standard.

I've done lots of architectural photography, and I no longer have such problems. If the building is level, then all you have to do is make sure the rear standard is level both in the vertical and horizontal directions. I do this by putting a cube level directly on the ground glass holding it against the top of the frame. This allows you to check both the vertical and horizontal leveling.

I also, check the image carefully on the ground glass using a small ruler acting as a sort of 't-square'.

It is also important for depth of field to make sure the front standard is parallel to the rear standard. I do this by turning the camera on the tripod so it is facing down, again checking that by putting a level cube on the rear gg. I then hold another level across the lens barrel, to check that it is also level. If under these conditions both standards are level, they must be parallel.

[rdenney]

It is also important for depth of field to make sure the front standard is parallel to the rear standard. I do this by turning the camera on the tripod so it is facing down, again checking that by putting a level cube on the rear gg. I then hold another level across the lens barrel, to check that it is also level. If under these conditions both standards are level, they must be parallel.

Unless they've sagged, even slightly, under their own weight.

Rick "all structures deflect under load" Denney

[ic-racer]

Of course; "level", to keep things simple, refers to both axes.
And according to my instruments, the back IS level and plumb; hence the mystery.
Next time out, I'll rely on the grid.

My experience with using the grid is that if it is level then you can get it plumb with the grid and if it is plumb, you can get it level with the grid. But, if it is neither plumb or level to start with, it is very difficult to get it right with just the grid.

[JeffKohn]

I wonder if in cases where the image seems level in one part of the frame but not the opposite, what you're actually seeing is geometric distortion from the lens. After all such distortion will be asymmetrical if using movements.

[E. von Hoegh]

Of course; "level", to keep things simple, refers to both axes.
And according to my instruments, the back IS level and plumb; hence the mystery.
Next time out, I'll rely on the grid.

Your instruments aren't sensitive enough. An accurate grid will never fail.

[E. von Hoegh]

I wonder if in cases where the image seems level in one part of the frame but not the opposite, what you're actually seeing is geometric distortion from the lens. After all such distortion will be asymmetrical if using movements.

Forgot about that. It would be worse the shorter the focal length (for a given format).

[Ari]

I wonder if in cases where the image seems level in one part of the frame but not the opposite, what you're actually seeing is geometric distortion from the lens. After all such distortion will be asymmetrical if using movements.

Hey, that's what I'd like to know, too. In fact, my impetus for asking in the first place.
Anyone know?

[E. von Hoegh]

Hey, that's what I'd like to know, too. In fact, my impetus for asking in the first place.
Anyone know?

What lenses are you using? I can assure you that some lenses will do it, but the result would probably be curving verticals and horizontals ( pincushion or barrel), ie nothing would be straight. The examples you posted are from a not quite vertical film plane.

[rdenney]

What lenses are you using? I can assure you that some lenses will do it, but the result would probably be curving verticals and horizontals ( pincushion or barrel), ie nothing would be straight. The examples you posted are from a not quite vertical film plane.

I agree. Even so, modern view camera lenses are really excellent in terms of geometric distortion, especially in shorter focal lengths. These lenses are made for these duties, after all. They don't have to zoom, or fit in front of a mirror box, or even be particularly fast.

Rick "who has never seen noticeable distortion in a modern view camera lens intended for architectural application" Denney

[John NYC]

Hey, that's what I'd like to know, too. In fact, my impetus for asking in the first place.
Anyone know?

One will definitely get "distortion" especially near the edges of the frame on rectangles or cylinders (like water towers -- something I am really familiar with). They will appear stretched. I have found that different lenses do perform better or worse at this type of thing, somewhat regardless of focal length. I've had some short lenses (like on 8x10 the 150mm SS XL) that are excellent in comparison to slightly longer lenses (like a 90mm on 4x5 that I had).

See this web page and go to the first paragraph under the section "Rectilinear vs Curvilinear Construction"

http://www.photographymad.com/pages/view/wide-angle-lenses

While I am seeing this stretching in your photos (naturally as you are using a lens that is corrected so that lines will stay straight), it also appears to me you don't have the camera level front-to-back and side-to-side and/or your camera is not perfectly squared-up and/or something you are doing with movements is doing this, and that is what I think is the cause of what is bothering you.

[Ari]

John, thanks for the post.
I don't mind a little stretching with rectilinear lenses, as long as that stretching doesn't veer off at an (unintended) angle, like the bricks in the second photo.
The photos I showed use almost no movements, perhaps some front rise.
I take the time to square up everything, plumb and level, but I may have overlooked something, as the photos attest.

[Leigh]

By the way, levels are not self-calibrating. If you turn it around and find an error, you have to make a new mark or adjust the level (if it's adjustable). That sounds like manual calibration to me.
Rick,

The term 'self-calibrating', as used in metrology, means that an instrument requires no external standards to be certified to its stated accuracy. It has absolutely nothing to do with the methods involved, be they automatic or manual.

In the case of a level, the setting is the zero point, and that can be set using only a flat surface as I described previously. All high-precision levels have an adjustment for setting the zero.

As to the accuracy of various levels, there's no way you could set a camera anywhere near level according to any of the better ones.

My two Starret 199Z Master Precision Levels (the standard of the industry for such instruments, at $733.00 each) have a sensitivity of 0.0005"/ft per division. That's the thickness of a sheet of paper at a distance of 10 feet. The 199Z is a bubble level.

As others have suggested, I only use the levels to get the camera close. After that it's align the camera to the subject using the GG grid.

- Leigh

[John NYC]

Rick,

As others have suggested, I only use the levels to get the camera close. After that it's align camera to the the subject using the GG grid.

- Leigh

Yep, that's what I do as well... but I have inferior levels compared to yours!

Here is what I use: http://www.leevalley.com/us/wood/page.aspx?c=&p=60783&cat=1,43513

I think I read about it here on the forum.

[Leigh]

Here is what I use: http://www.leevalley.com/us/wood/page.aspx?c=&p=60783&cat=1,43513
That's a nice-looking level.

The accuracy seems appropriate for this application (roughly ±1 arcminute).

- Leigh

[Tim k]

Holy Smoley, this thread makes me want to go find something round on a sand dune on a foggy day.

[Ari]

I'm just following up.
I've taken all the suggestions from this post and applied them these last few days, with a noticeable improvement; thank you to all.
But, for the life of me, I can't explain why something like this happens: the front right side corner of this building bends away to the right at a slight angle, while the other two (visible) sides of the building are apparently straight.
Can anyone explain this to me? Is it impossible to get the three sides plumb on a piece of film?
Thanks
http://i52.tinypic.com/2v7szsh.jpg

[Leigh]

Hi Ari,

The left front edge of the building is farther from the camera than the right front, so it will look smaller.

The same is true of the right rear edge, although it's less obvious.

It's one of those "optical vanishing point" problems.

You can minimize it by getting farther away from the building, thus reducing the difference between the two distances, but there's no way to eliminate it.

- Leigh

[John NYC]

Kirk will tell you that a proper architectural photo is done just like you have it there. However, it doesn't look natural to a lot of people, so what some people do is not fully correct the converging verticals with front rise. That is, they might point the camera up a little and use front rise. This would give just a little of that "building falling backwards" thing and somewhat alleviates the "building is expanding as it goes up" feeling, but not too much for most people (other than architects and architecture photographers who would shoot me for saying what I am saying).

[Ari]

Thank you Leigh, thank you John.
Leigh, I still don't see why a vertical line does not reproduce as vertical, given the chance. Are you in effect saying it's my lens that warps vertical lines the closer I get?

John, you are correct. There is even a formula somewhere for allowing a wee bit of convergence so as not to look unnatural.
But that's assuming all your lines are parallel to begin with.

I know it seems I'm nit-picking, but I will be showing some work (nothing that I've posted on the forum) in early September, and I don't feel I could show a photo that has this type of problem.
Thanks again!

[rdenney]

I'm just following up.
I've taken all the suggestions from this post and applied them these last few days, with a noticeable improvement; thank you to all.
But, for the life of me, I can't explain why something like this happens: the front right side corner of this building bends away to the right at a slight angle, while the other two (visible) sides of the building are apparently straight.
Can anyone explain this to me? Is it impossible to get the three sides plumb on a piece of film?
Thanks
http://i52.tinypic.com/2v7szsh.jpg

I just put a ruler to my screen, and the front right corner of the building measures vertical for the lines that I can see. Only the very top corner doesn't seem right. When I place the ruler along the front corner, the top swings in a bit, but only the very top 6 or 8 feet or so of the building. A ruler also shows that the line of the parapet in that corner section facing the front is not parallel to what should be the same line in the upper left corner of the building. Finally, the upper right corner seems just a touch unsharp. The misalignments are neither pincushion nor barrel distortion--the whole corner swings left and up. All those symptoms suggest to me that the film was curled up in that spot. This should not happen in a standard film holder. Did you custom-cut the sheet and miss a bit? The problem might be in the negative stage of an enlarger if you enlarged it, or in the scanner film holder if you scanned it.

I disagree with others that the building should converge a bit in this case. When the edges have such high contrast, and those edges of the building are so close to the edges of the frame, even a slight lack of vertical will become very apparent to the point of being distracting. If the building was photographed from further away so that its edges weren't right at the frame's edges, then a bit of vertical convergence prevents the illusion that it is diverging at the top. That's not what you are complaining about in any case, of course.

Rick "who has to use a straightedge to eliminate the possibility of optical illusions" Denney

[Ari]

Rick,
Thanks for going to the trouble.
I use Grafmatics exclusively, one of the reasons being the supposed film flatness inherent in the holder. And it's a scan from an Epson 4990 with the supplied 4x5 film holder.
Not all my shots turn out this way, some are rendered perfectly straight, so I might, as you imply, have to find out if I have a rogue holder.
I also try to keep in mind what a friend told me, that not all buildings are built straight and true, some human error must factor in every now and then. But I don't think this is the case here.
And I agree with you, this building is not tall enough to warrant a slight convergence.

[Kirk Gittings]

Two things:

1) Rigidly aligning the frontal perspective of a building in a photo really makes the building pop-gives it a strong "presence".

2) And if the image is really powerful small problems in the alignment, rectilinearity of the lens or problems in the building will pale and go unnoticed.

Which leads me to my real question, why photograph this building? Is it just an exercise? It is neither interesting historically or as contemporary design. it is dead as a doornail in terms of architecture-corporate pap. Photograph buildings that excite you. Photograph in dynamic light. Pay attention to details but learn to feel good architecture and try to bring that to life in your images. The rest will come.

[rdenney]

Which leads me to my real question, why photograph this building? Is it just an exercise? It is neither interesting historically or as contemporary design. it is dead as a doornail in terms of architecture-corporate pap. Photograph buildings that excite you. Photograph in dynamic light. Pay attention to details but learn to feel good architecture and try to bring that to life in your images. The rest will come.

Agreed on all points (though I did like the reflected cloud in the windows), but it would be a shame to mess up a gig photographing an important building because one of the septums in his Grafmatic was bent, causing the film to curl (or whatever the problem is).

Did you see the issue I saw in the upper right corner? I for one would be interest in your expert analysis of what you see.

Rick "who has done a lot of boring things for money" Denney

[Ari]

Excellent advice, Kirk.
I just want to find out if I'm playing with faulty equipment, i.e. my gear or my eyes :)

This was purely an exercise: I was trying out a new GG, and applying the advice given to me in previous posts concerning alignment procedure.
The building is flat-out awful (federal gov't), but it's very close-by.

I scanned the same image directly on the scanner glass to eliminate the possibility of a mis-aligned scannner holder, but the results were identical.

I was using indirect swing here, pointing the camera up, then setting both standards to plumb.

Here is another shot from the same spot, but with the standards swung to face the building; indirect swing again, and the same corner making trouble for me.

http://i52.tinypic.com/w4s3k.jpg

[Ari]

Rick "who has done a lot of boring things for money" Denney

:D :D :D

Amen to that!

[rdenney]

When I put a ruler to this one, the verticals on the side are vertical and straight, at least within the resolution of my monitor.

But the top of the parapet wall still has an angle to it. But when I put a ruler across the whole building, it seems as those the parapet wall on the left upper corner has a similar angle. Are you sure the wall sections are plano-parallel? I can't confirm it with the cloud reflections.

Rick "thinking this angle will reduce the effect somewhat" Denney

[Ari]

That may be true, Rick; my obsessive nature will get me back to the building for a closer look.
But first I'll go through the Grafmatics and see if any septa are bent; I think you may be onto something.
Thanks!