• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Helpful Hints/D.I.Y.: Leveling wide angle Fields of View

Hi folks.

I recently acquired a 24mm tilt and shift lens and experienced how difficult it is to get the horizon level (and the lens is almost free of distortion, so that's not the reason). Since the lens is capable of producing an even wider FOV by exploiting its shift capability, it became even more important to get things leveled properly.

I know that the stitching software allows to correct for horizon and verticals, but sometimes it only adds time in postprocessing, and it may cause the need for some cropping. Also, there are subjects that don't give good clues for control point placement (e.g. rough landscapes), or require an accurate display of the angles that are designed by e.g. an architect.

Having come to the conclusion that the various bubble level fitted accessories only have a limited degree of accuracy (and can also get misaligned a bit in use), I spent some time in finding a better tool for when eyeball-ing isn't accurate enough.

For indoor architectural use I already used an automatic leveling laser guide which is quite accurate (+/- 0.5 degree), but a bit delicate and bulky in transportation. It does work fine in combination with Live View or tethered shooting, provided that there is not too much (sun)light falling on the subject (up to 6 metres distance before it gets too dim anyway), and that it can be positioned to align with the edge of the frame or one of the lines that can be superimposed over the Live View display. Aligning with the camera's viewfinder is error bound, due to slight rotation misalignments between the groundglass and the actual sensor.

I now purchased a relatively compact electronic level (via eBay), after comparing several similar devices. The one I chose is more accurate than some of the others, yet relatively affordable. It is the Digi-Pas DWL80Pro . The Pro version offers 0.05 degree accuracy versus 0.1 degree for its cheaper non-Pro sibling. It therefore also allows to calibrate more accurately, and verify the need for recalibration.

While more expensive than a bubble level this device allowed me to see that the built in levels in my Pano gear are not accurate enough, which explains why I had to correct in post processing which also costs, time=money.

This is a very good companion to my RRS BH-55 with rotating PCL-1 clamp. It allows to set the rotation plane level, and now I nolonger have to rely on inaccurate bubbel levels or plumb lines.

I'll probably now will find out how far off the sensor is mounted in my camera ;-) , but at least I will have a good reference to use when squaring things up before the shoot. I can now at least use a known rotation offset to compensate for any sensor misalignment should it manifest itself.

One more variable quantified/eliminated.

Cheers,
Bart
 
Last edited:

Michael Fontana

pro member
Bart

I had the same problem here and use a Stabila precison level, Typ 80 E with a precision of 0.029 degs, since about 3 years. Even when buying that one, I used t more precise one, to grab the fine one out of the bunch.
One of my best investements ( 80 bucks) ever - better than any Photohop-tool.

They had some digital levels too, but I decided to go for a conventionel one, as I feared the battery to stop working just in the moment I need it.
 

nicolas claris

OPF Co-founder/Administrator
Amazing !
I had this morning exactly the same discussion whith the local guy or does clean my sensor and do some adjustment on the gear…
I told him that despite the viewfinder with a special grid, I have always (wether with horizon shot handheld) or with the tripod a tendancy to shoot always 0,5 to 1 degree clockwise.
He said that some good tools exist but couldn't remember what that were…

And Bart do post that now… "close minded" isn't it ? ;-)

Micahel, any link to Stabila precison level, Typ 80 E? Googles gets me to "nowhere" precise…
 
Bart

I had the same problem here and use a Stabila precison level, Typ 80 E with a precision of 0.029 degs, since about 3 years. Even when buying that one, I used t more precise one, to grab the fine one out of the bunch.
One of my best investements ( 80 bucks) ever - better than any Photohop-tool.

They had some digital levels too, but I decided to go for a conventionel one, as I feared the battery to stop working just in the moment I need it.

Hi Michael,

Yes battery dependence is a potential issue, although it's supposed to warn when the power is getting low. I've also looked at the accurate (0.5mm / metre) bubble vial levels like the one you mention, and accurate bullseye types, but they tend to be a bit bulky. That's what I like about this electronic one, it fits easily in the bag, although there are smaller ones (but less accurate).

This electronic one also has traditional bubble levels built in which adds a bit of length, but it's nice to get things approximately adjusted, before the final touches are applied. What's nice is that the calibration adds to the confidence one can have about ones tools.

Bart
 
Amazing !
I had this morning exactly the same discussion whith the local guy or does clean my sensor and do some adjustment on the gear…
I told him that despite the viewfinder with a special grid, I have always (wether with horizon shot handheld) or with the tripod a tendancy to shoot always 0,5 to 1 degree clockwise.
He said that some good tools exist but couldn't remember what that were…

And Bart do post that now… "close minded" isn't it ? ;-)

It must be telepathy!

Micahel, any link to Stabila precison level, Typ 80 E? Googles gets me to "nowhere" precise…

I found a link here, but it seems a lot cheaper now than what Michael payed. But then Switzerland seems to be on the pricy side for many things.

Basically the accurate ones are labeled "0.5mm / m", and there are other common brands as well.

Bart
 
Sensor rotation?

Just a quick one.

To test for sensor rotation one can place the camera on a flat surface, in front of a mirror. Shoot the image of the camera itself while centering the reflection of your lens in the viewfinder. The image taken should show the rotation of the camera versus the sensor and vice versa.

Bart
 

Ken Tanaka

pro member
The new Olympus PEN E-P1 actually has an electronic leveling display option. When activated an indicator along the right border shows the camera's pitch while an indicator on the bottom shows roll. It's really quite nifty, although I've not found it useful for the type of photography I'm doing with this particular camera. I wonder why Canon and Nikon don't incorporate such a feature in their big bodies?
 
The new Olympus PEN E-P1 actually has an electronic leveling display option. When activated an indicator along the right border shows the camera's pitch while an indicator on the bottom shows roll. It's really quite nifty, although I've not found it useful for the type of photography I'm doing with this particular camera. I wonder why Canon and Nikon don't incorporate such a feature in their big bodies?

Hi Ken,

I'm not sure, but I thought Nikon also had something like that (in Live View, in the viewfinder?)? Don't know how accurate it is though.

I'm testing the digital leveler on my RRS pano setup, and it's interesting what can be learned. It looks like it that adding a heavy camera and ditto lens will flex the aluminium bars enough to cause a rotation of 0.3 degrees. So, I'll start my cameraless leveling to show a 0.3 degree angle in the other direction to get the desired neutral working position. To measure is to know.

I love it when a plan comes together.

Bart
 

Asher Kelman

OPF Owner/Editor-in-Chief
Hi folks.


I now purchased a relatively compact electronic level (via eBay), after comparing several similar devices. The one I chose is more accurate than some of the others, yet relatively affordable. It is the Digi-Pas DWL80Pro . The Pro version offers 0.05 degree accuracy versus 0.1 degree for its cheaper non-Pro sibling. It therefore also allows to calibrate more accurately, and verify the need for recalibration.

While more expensive than a bubble level this device allowed me to see that the built in levels in my Pano gear are not accurate enough, which explains why I had to correct in post processing which also costs, time=money.

This is a very good companion to my RRS BH-55 with rotating PCL-1 clamp. It allows to set the rotation plane level, and now I nolonger have to rely on inaccurate bubbel levels or plumb lines.

Bart

Hi Bart,

Thanks for sharing this valuable information. How does one use a level. Until now, I put the level on my LF camera and that works fine. With a DSLR, are you putting the DWL80 on the PCL-1 clamp?

I have the manfrotto stage and a series of RRS clamps to go to the L bracket on my camera. There's a little play that worries me.

Is it that the PCL-1 clamp is true with the L bracket?

I'm concerned about play in the set up.

Thanks,

Asher

Asher
 

Michael Fontana

pro member
Nicolas

I' ve two with 30 cm lenghts, each, not the cheap yellow ones:

typ-80-e-30-cm-2473-0.jpg



I didn't went to the cheap homebuilders store - if you compare them with themselfs, you realise pretty fast that these cheap are crappy - but to the more expensiv pro stores. Maybe that's the reason - beside the swiss beeing milk cows - they were more expensiv as well, 80 sFR. each - but precision is worth it.

Precision depends on the lenghts, too: the longer the better, so I took about 10 of the 30 cm long ones, and verified their precision with the help of a longer and even more expensiv and precise level and finally bought the two best.
 

Michael Fontana

pro member
It looks like it that adding a heavy camera and ditto lens will flex the aluminium bars enough to cause a rotation of 0.3 degrees.
Bart

Yep, that's true - with a lens of 1 kg, it'll be even more.
I had enforced my Seitz pano head by the company, as heavy cam and heavy lens resulted in quite noticable differences. So there's a tradeoff between weight and stability/precision.
 
Hi Bart,

Thanks for sharing this valuable information. How does one use a level. Until now, I put the level on my LF camera and that works fine. With a DSLR, are you putting the DWL80 on the PCL-1 clamp?

Hi Asher,

Yes, that's what I first tried. That is relatively easy. One just releases the PCL-1 rotation lock, and slightly releases the ballhead itself, and puts the level on the PCL-1 base that's fixed to the ballhead. The weight of the PCL-1 is very low, so by carefully tapping on the side that needs to move up the level can be set to zero for that axis (the level indicates which end should go up or down). Then one rotates the PCL-1 by 90 degrees and repeats the exercise for that axis. One can double check wether the other axis is still level. Then the camera is mounted.

That will work fine with a camera only, or with a single multipurpose rail for flat stitching.


I have the manfrotto stage and a series of RRS clamps to go to the L bracket on my camera. There's a little play that worries me.

Is it that the PCL-1 clamp is true with the L bracket?

I'm concerned about play in the set up.

Indeed, so was I. With the other "Omnipivot" setup for 3D rotations, I have found that the weight of a very heavy camera/lens combination will flex the rails a bit. Now that I can kind of quantify the effects of gravity, I'm arriving at the conclusion that a useful workflow is to determine the amount of de-leveling caused by gravity, and set up the Omnipivot pano package with a bias so that it levels after mounting the camera. It's still a bit of work, but now there is no guesswork anymore. The effects can be quantified, and be compensated for.

Bart
 

Asher Kelman

OPF Owner/Editor-in-Chief
Hi Asher,

Yes, that's what I first tried. That is relatively easy. One just releases the PCL-1 rotation lock, and slightly releases the ballhead itself, and puts the level on the PCL-1 base that's fixed to the ballhead. The weight of the PCL-1 is very low, so by carefully tapping on the side that needs to move up the level can be set to zero for that axis (the level indicates which end should go up or down). Then one rotates the PCL-1 by 90 degrees and repeats the exercise for that axis. One can double check wether the other axis is still level. Then the camera is mounted.

That will work fine with a camera only, or with a single multipurpose rail for flat stitching.
That's great, I'll review whether I might buy the RRS PCL-1 to go all RRS. Do you use RRS stage too so that you get the "NP" in the right place.

With the large format camera there seems to be more accuracy in the reference to the back. With the DSLR, one doesn't know because the lines in the viewer may not be correct. However, one could, I guess find out how good that is by checking the how a horizontal laser line lines up in the viewfinder to ground glass lines and verticals too. I'm not sure if one can fine adjust the ground glass in the viewfinder if it's off.

The Ricoh digicam GX100 has a built in electronic level which is nice but I don't know how accurate it is.

Indeed, so was I. With the other "Omnipivot" setup for 3D rotations, I have found that the weight of a very heavy camera/lens combination will flex the rails a bit. ...The effects can be quantified, and be compensated for.
I would think with the electronic level, one could tilt the RRS stage exactly the right amount for that camera. (Maybe one could even fit the electronic level on the RRS L bracket of the DSLR? Then one could correct for sag routinely). A handy option is to have a stage that can be reproducibly adjusted in it's correction to flat.

This is where the Arca Cube would be handy! I only know one person who owns one.


629979.jpg


Photo BHPhotovideo​

This is a well engineered device for $1924, it had better be! Even though it's very expensive, I can see having a chart of adjustments for sag according to the lens and shift used. Obessional, yes, but if one can make it routine and one get's paid for photography then it might be worth it. The other thing is a more beefed up rig.

I personally resent having to straighten things in Photoshop as digital photography is a larger black hole for time than the chemical darkroom. It seems so much faster but then we take more pictures and get sucked into the black hole of PS corrections of images.

It might be that once I have the perfect view I like, I can do it in one go with my LF camera with the Digital Fusion stitching back, where everything is in the same plane.

Asher
 
Do you use RRS stage too so that you get the "NP" in the right place.

Hi Asher,

Yes, I use the full "Ultimate-Pro Omni-Pivot Package" on a BH-55 ballhead.

With the large format camera there seems to be more accuracy in the reference to the back. With the DSLR, one doesn't know because the lines in the viewer may not be correct. However, one could, I guess find out how good that is by checking the how a horizontal laser line lines up in the viewfinder to ground glass lines and verticals too. I'm not sure if one can fine adjust the ground glass in the viewfinder if it's off.

It can probably be adjusted by a qualified service center, but it's not going to be exact because even a minute displacement results in a larger angle when magnified. The cameras that allow the user to replace the groundglass will still have a little tolerance in the final position.

I would think with the electronic level, one could tilt the RRS stage exactly the right amount for that camera. (Maybe one could even fit the electronic level on the RRS L bracket of the DSLR? Then one could correct for sag routinely).

There are a few flat areas that can be used to put the level on, the L-plate itself, when mounted to the camera, doesn't offer enough room:

7551_Web_S.jpg


As you can see, the lower rail of the pano package can be used to level that. That is the rail that will be used for flat-stiching as well. It will allow to level the plane of rotation. When a specific camera's sensor rotation is known, it can also be used to level with a slight angle to compensate for the sensor rotation. Having the tool to set the bias accurately is therefore a great help.

When using the other parts of the pano package as well, one could compensate for mechanical error of the construction, as well as for the torque that will be excerted on the construction (in addition to the sensor rotation offset). It will alter the plane of horizontal rotation a bit though.

The following images show that there are some spots where the level could be placed to get the other axes level:

7552_Web_S.jpg
7553_Web_S.jpg


The first image showns the difference between the lower rail and the upper one (0.30 degrees). In fact, as it turns out, despite the computer machining of the Alumini(m)um, 0.20 degrees are due to tolerances in mounting, and 0.10 degrees as a result of torque. In the second image I rotated the level a bit to show its display, but normally one would align the level with the rail.

Measuring the leveling with and without the camera mounted, will allow to quantify the torque effects for a given camera/lens combination (depending on the entance pupil and camera orientation, the force arms can/will differ per configuration).

A handy option is to have a stage that can be reproducibly adjusted in it's correction to flat.

This is where the Arca Cube would be handy! I only know one person who owns one.

Yes, Jack is fortunate to have such a fine piece of enigineering. I wouldn't mind having it, if only for the construction (I'm sensitive to the beauty of design). However, the expensive/heavy contraption will almost certainly not have its center of rotation at the entrance pupil or even on the optical axis, thus making it less useful for precise pano stitching.

I personally resent having to straighten things in Photoshop as digital photography is a larger black hole for time than the chemical darkroom. It seems so much faster but then we take more pictures and get sucked into the black hole of PS corrections of images.

It's always a trade-off. Sometimes it's faster or better to do it when shooting, sometimes it can be part of the post-processing workflow. When stitching e.g. the absolute vertical alignment is part of the stitching routines, but it usually helps quality if a proper effort was also taken when shooting. And not all shots are stitched, in which case leveling the camera (+sensor) makes sense.

Bart
 
Top