Nodal points G11

[flowe]

Hi all,

Where are the Nodal points of the G11?
I have no means to find out - who does?

Thanks in advance
flowe

[Erik_L]

After googling "Nodal points", I am still wondering what you need to know....

How close can it focus? Focal length?

[forno]

After googling "Nodal points", I am still wondering what you need to know....

How close can it focus? Focal length?


IIRC it is the point around which to rotate the camera so as to get the edge overlap correct in pano's

I think it is a function of sensor plane and focal length

[flowe]

IIRC it is the point around which to rotate the camera so as to get the edge overlap correct in pano's
I think it is a function of sensor plane and focal length

forno, thanks for the clarification. Taking it one step further: the focal length is only a "handle" for the nodal point. Defined it is by lens and body design and the movement of individual lenses, shifting the optical center of the system in a way not readily known. The nodal point is given as distance along the optical axis from the tripod socket plane - with reference to the focal length, but as an entirely different number of millimeters. Consequently, the nodal information pertains to a given lens and body and would be different between say a G6 and a G10/G11.

flowe

[CyberPet]

Wouldn't the nodal point be exactly in the middle of the lens? So if the tripod mount hole is not exactly over the lens, then you have some corrections to be done, as it'll not rotate around the lens nodal point, but around the tripod mount holes nodal point.

[Erik_L]

I would assume that if the photo were taken from far enough away, this would be negligible.

[flowe]

Wouldn't the nodal point be exactly in the middle of the lens? So if the tripod mount hole is not exactly over the lens, then you have some corrections to be done, as it'll not rotate around the lens nodal point, but around the tripod mount holes nodal point.

CyberPet: "middle of the lens" is ambiguous - the nodal point lies *somewhere along the optical axis* of the lens. Then you have to take into account the height of the optical axis above camera base plane (if camera vertical), as well as the horizontal deviation of the tripod mount hole from the optical axis. Rotating the camera *truly* around the nodal point is what commercial and makeshift pano adapters are about, i.e. compensating for the deviations in all three axes.

Erik_L: Quote: "I would assume that if the photo were taken from far enough away, this would be negligible." - Right, but panos live with the depth of the subject area, and you wouldn't remain satisfied with all those flat panoramas!

flowe

[CyberPet]

Wow flowe, that totally went above my head. Thanks for the explanation. :D

[forno]

I remeber a PDF file floating around a year ago or so, that set out how to find the nodal point.

I think you need to find 2 object in line with each othe, say telephone poles, position the camera at "straight ahead". At this point you should only see 1 telephone pole with the other being hidden behind the first.

Now rotate the camera to the right, if you can still only see one telephone pole you are rotating at the nodal point.

This is from memory, dont hold me to it:lol::lol:

[DaryleH]

I remeber a PDF file floating around a year ago or so, that set out how to find the nodal point.

I think you need to find 2 object in line with each othe, say telephone poles, position the camera at "straight ahead". At this point you should only see 1 telephone pole with the other being hidden behind the first.

Now rotate the camera to the right, if you can still only see one telephone pole you are rotating at the nodal point.

This is from memory, dont hold me to it:lol::lol:

Perfect, you need to move the camera on a slider to find the perfect point, but you hit it on the nail head.

Good shooting
Daryle

[Erik_L]

I have experienced this accidentally while shooting macros with reflective surfaces. It appears that no matter how I rotate the camera, I can see "Canon" reflecting off the glass. If is shift it just slightly and then rotate it, I can get the correct angle to lose the "Canon" from showing up.

[cccc]

I have experienced this accidentally while shooting macros with reflective surfaces. It appears that no matter how I rotate the camera, I can see "Canon" reflecting off the glass. If is shift it just slightly and then rotate it, I can get the correct angle to lose the "Canon" from showing up.
I think a polarizing filter will fix your problem.

[flowe]

I tried to establish G11 nodal points with a rather elaborate setup - tripod, pano head, Soligor slide and two suspended blue and red sewing threads loaded with screws in 80 and 100cm distance. But sorry, no success. Lining up and focussing straight ahead was easy, but after panning to the edges of the monitor, I haven't been able to establish conclusive results. On the wide end, the threads became almost invisible. So for the time being, I gave up. Any better ideas?

On the other hand, I've done some panos - even with handheld camera - lining up on assembly without the slightest ghosting etc. At present, I prefer "PanoramaStudio 2.0" by http://www.tshsoft.de with very pleasing results. It doesn't look like versions other than in German would be available, but a very decent user interface might help without knowing the language - a demo version is available. PanoramaStudio reads all required info from EXIF and applies correct debarrelling automatically. (BTW: what are experiences with the Canon supplied pano software?)

I feel that the nodal question is probably not very critical with the G cameras and similar (due to the narrow optical system?) Sensibly lined up on a pano head plus appropriate software will produce most satisfactory results. One more aspect: camera in vertical position makes for a larger vertical view angle and less panning between shots and therefore a less critical nodal issue.

Good pano! :smile:
flowe

[denncald]

Re: Canon pano software - I've had fairly good results with this software and my G9. However, I more often prefer the results achieved using the free Microsoft ICE (http://research.microsoft.com/en-us/um/redmond/groups/ivm/ice/) software for panos.

Dennis

[rpolitsr]

Hi flowe

I think you will do a great job with your setup, except for one thing:

two suspended blue and red sewing threads loaded with screws in 80 and 100cm distance

You do not need thin threads to do the job, and it is better to place them near to the lens the first reference, (say, 80cm) but the second one far enough to simulate infinite. I used a 19mm aluminum tube close to the camera, I think it was at 100cm from the lens, and a tall pole (12 meters) holding a lighting fixture in the street, 50 meters from the lens. It worked, with the knowledge and guidance of Terrywoodenpic (http://photography-on-the.net/forum/member.php?u=45685). Check one of his threads: G3 G5 accurate nodal points. (http://photography-on-the.net/forum/showthread.php?t=143537)

[DaryleH]

This digital photography just gets easier every day, keeps the grey matter working though:lol::lol::lol::lol:

[flowe]

Hi rpolitsr and everybody interested

In the meantime I've been at work with the G11 and its Nodal points.
You find here (http://homepage.hispeed.ch/flowe/digifoto/nodal/9b15_nodal_points_g11.pdf) my findings with comments and more on a printable pdf.

Please comment!
Regards - flowe

[flowe]

Hi all

just curious... great effort... and no comments at all??

Some readers might have got stuck with my last paragraph re focal length with zoom lenses. I'm referring to an article on Spiegel-online of 8. November 2009: http://www.spiegel.de/netzwelt/gadgets/0,1518,658375,00.html#ref=nldt

The problem is that the well known formulae for hyperfocal distance and Depth of Field refer to prime lenses (i.e. with one single permanent focal length), moving as a solid unit along the optical axis while focussing - similar to a handheld loupe. This appears as entirely different with zoom lenses and increasingly so when focussing below infinity. The focal length from Exif most likely applies to infinity only and then changes in an unknown way. This would imply a moving nodal point depending on the current focal distance and the current true focal length - all unknown.

The figures such as given in my table may serve as an useful approximation, and the final outcome will largely depend on the capabilities of the software used - and a some luck. Let's enjoy it... :) :)

Please let me know what you think.
Regards - flowe

[cboshuizen]

Hey, I just found this forum yesterday. I checked out your results. Do you mean that the nodal points are on the lens side (front face) of the camera (except for 140mm case)? So we'd actually be rotating under the lens more or less?

I found this great T-adapter that might help with the G11 tripod mount offset, and bring the tripod under the lens for the first zoom stop.
http://store.nodalninja.com/product_p/t-20.htm

Chris

[flowe]

@Chris. So it is, as stated in the first para of my pdf paper: The Nodal points for all zoom settings lie in front of this plane (as per table) – except the one corresponding to focal length 30.5mm, which lies 7mm behind this plane.

Re focal lengths, as mentionned in my last post. The "Spiegel" reported that zoom lenses like 28-300mm for DSLRs may loose up to half the focal length when focussed below infinity. If similar with the Canon G cameras, this would render any discussion of nodal points futile. In the meantime, I investigated this with the G11 and can give a conditional "all clear". Focal distances like 500-1000mm reduce any set focal length by roughly 10%. Starting with 2000mm, the real focal lengths remain more or less close to nominal. The shortest setting 6.1mm over most of the distance range is really 6.4mm. The longest setting 30.5mm varies the most, going from 26mm at 500mm to 35mm at infinity. I do have the detailled figures, but try to summarise.

Generally speaking, the shifting of focal length over most of the distance range is unlikely to interfere noticeably with the nodal issue. This - as far as I see it - still requires to know the set focal length, either by counting the zoom steps or using the index device I suggested. And it requires ways and means to set up and rotate the camera accordingly, including thread offset. And not to forget: manually taken images may lead to pleasing results too - with some luck.

Regards - flowe