Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: MarkSe on September 12, 2012, 11:48:00 am
-
I use a hot shoe bubble level for my architectural photography (5dmkII ts lenses)
Problem is that its too inaccurate so i have to spend a lot of time correcting in photoshop.
Is there a more accurate alternative to hot shoe bubble levels?
-
I "carry" the level in the hot shoe but I don't use it there. I put some protective tape on one side to protect the screen and just hold it up against the viewing screen. For twilight shots it gets illuminated by the screen too. This doesn't cover the horizontal leveling but its easy to eyeball that.
-
Good idea- I will try that!
-
The problem's likely in the shoe - not necessarily level. I use a hot shoe level but not in the shoe, either - I simply put it on my Arca plate and level that. Of course, the marks on bubble levels are not always accurate, either. If you level the plate and it's still not right, the bubble may be the problem.
-
Buy a Mk III ;D
The inbuilt level is very useful.
-
I do what Kirk does. But have you ever heard this carpentry joke: "If you have three levels and they all agree two of them are wrong?"
-
Acratech makes a large bubble level on an ArcaSwiss plate. The assumption is that your camera sits square on its own A-S plate.
-
I use a hot shoe bubble level for my architectural photography (5dmkII ts lenses)
Problem is that its too inaccurate so i have to spend a lot of time correcting in photoshop.
Is there a more accurate alternative to hot shoe bubble levels?
I place the bubble level on the lens (the T/S lenses have a nice flat area to set it on). And, because bubble levels themselves are not all that well made, you can calibrate your level by carefully framing a known straight object or structure, and then making new marks on the bubble level.
-
Is there a more accurate alternative to hot shoe bubble levels?
Hi,
The topic has been discussed before (http://www.luminous-landscape.com/forum/index.php?topic=46439.msg389056#msg389056), but what ultimately remains is the unknown amount of sensor rotation. So even if you have an accurate level, and you can find a surface that's truely level with the camera's bottom plate (the hot-shoe rarely is), there can still be a small deviation. The electronic in-camera digital levels with an 'accuracy' of 1 degree are not accurate enough for really critical leveling.
BTW you can check for sensor rotation with a TS lens, by shifting left and right and trying to align the two images where they overlap. In e.g. Photoshop you can use two layers with the top layer blending mode set to difference. You'll notice that not only is (obviously) a left/right shift required, but very likely also a small vertical shift. When the sensor is slightly rotated, then the vertical shift is required as well to get closer to perfect alignment. Actually a rotation is needed, but that will require interpolation of one of the images, which will lose some pixel accuracy due to the old fashioned bicubic resampling algorithm.
When you know the rotation, you can adjust the camera level by slightly compensating with a very accurate level (set it to the negative of sensor rotation instead of 0.00 degrees).
Cheers,
Bart
-
Some tripods have levels built into the head.
-
FWIW, you can overcomplicate this action. Buy a high quality bubble level-the super cheap ones are never properly manufactired (seconds I think of the more expensive ones). I shoot architecture for a living, day in and day out. If the procedure I described did not work would I keep doing it for 34 years (on LF film I put the bubble level on the ground glass-unfortunately there is no screen to light it up for twilight shots) making more work for myself?
-
Some tripods have levels built into the head.
True enough but how accurate they are depends on how well made the head is and then how the position how"squarely" the camera fits onto the head.
-
An architecture/interiors photographer I know recommend this: http://www.ihandysoft.com/carpenter.html I have not tried it yet. Reportedly, it can be calibrated, and so can account for, say, a hotshoe not being aligned with the camera.
-
And, because bubble levels themselves are not all that well made, you can calibrate your level by carefully framing a known straight object or structure, and then making new marks on the bubble level.
To test a level, place it on any surface (it does not have to be a perfectly level one) and note the relationship of the bubble to the marks. Rotate the level 180 degrees. If the bubble/mark relationship looks the same, the bubble is accurate.
-
I use a Gitzo leveling base which has a fairly large spirit level first and then the small level in the clamp on my RRS head. It's been right on when checked against the built in level of the 7D and a 3 way spirit level set on the top of the TS-E 17 and 24 II lenses. I have only had to correct an image when my assistant kicked the tripod and didn't tell me :o
If I'm shooting / copying artwork I use an inclination vial on the subject and then on the LCD.