Peter Patricelli
Posts: 154
Location: Eugene, OR, United States
Registered: 20 Dec 2006
|
Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 14:19 GMT
|
|
Maybe I am the only one but the process of setting up and shooting a 360 VR panorama involves so many steps in terms of the tripod, head, camera settings, etc. which are NOT part of things I commonly think of and adjust when shooting, that I am constantly forgetting something like White Balance or turning the auto-focus off. So after shooting one or two sequences I remember something, have to re-shoot, etc.
So, to help my clunky brain, I created a 360 VR Pan checklist which I have laminated and will attach to my pano head. I offer this as proof that I really need all the help I can get AND in case anyone spots something I have missed. And if anyone else wants to copy this, that is fine.
Panorama Sequence
Tripod
1. Level tripod
2. Level head
Panorama head
3. Set positions for Camera and Lens used
Camera settings
4. Jpg versus Raw
5. White balance
6. Set to single exposure
Exposure
7. Light meter
8. Test shots (20mm lens) in A mode
9. Set ISO setting
Final Camera Settings
10. Set to M mode
11. Set f stop and shutter speed
12. Set Bracketing
Focus camera
13. Set Hyperfocal point
14. Set pitch
Shoot sequence
ANything missing?
Peter Patricelli
www.flyfishingfotography.com
|
|
alert moderator
|
|
DorinDXN
Posts: 2802
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 17:32 GMT
updated: 18 Jan 2007 at 17:33 GMT
|
|
Good idea, and useful list deppending of tripod maybe level operation to have greater order
15. mirror lookup
and deppending how big one's memory card is,
16. format or empty card
especially at multi row, you know, one is at photo number 27x3 + 2 then, card full, then must touch the camera for delete.
17. cover the viewfinder
you know your light metter shows that is good 3 sec exposure but you have lens covered, where the light came from?
18. look after your shadow, your reflexion, your bag could save time of post process
19. find a good start for anchor/ think about good start for pano also (symetry, the rule of third..)
20. extrashots if people moving (sugested by Andrey)
21. or water moving (suggested by John Houghton)
22. take handheld nadir first (suggested by Bill)
23. rotate (counter)clockwise depends of how the moving features/people are (suggested by irieman)
24. look how the ceiling is to low to high try imagine how your pano will look like
25. make some points if you don't have(with laser my suggestion, or with sticky tape suggested by John Houghton)
maybe also an mp3 player, you know those on usb .. you label the next track buton with "Checked" or so
track 1
Level tripod
than press button Checked
play track 2
Level head
one have much to look at at pano session and to hear some useful advices is good.
I know what you think of right now.. 
555-about-to-pano phone number
I'm gonna rest a while now and I may came up with some extra
Dorin
|
|
alert moderator
|
|
David Marine
Posts: 37
Location: Austin, TX, United States
Registered: 24 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 17:42 GMT
|
|
Why does the tripod, as well as the head, need to be leveled? Isn't it enough that the head/bracket is leveled?
- David
|
|
alert moderator
|
|
DorinDXN
Posts: 2802
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 18:03 GMT
updated: 18 Jan 2007 at 18:04 GMT
|
|
Good question David, for some tripods, at the cheeper side, you can't move your tripod head in any direction easy without wobbling so if you ajust the legs length to have good leveled tripod it helps on leveling the head easier. At least in my case if I don't tight secure the "yaw" on tripod its woobling, then after leveling the tripod, by ajusting the "pitch" I level the bracket
Dorin
|
|
alert moderator
|
|
spyboy
Posts: 287
Location: New Hampton, NH, United States
Registered: 7 Oct 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 20:30 GMT
updated: 18 Jan 2007 at 20:31 GMT
|
|
the one step I forget all the time is to set my Sigma fisheye to infinity. Usually I remember about 3 pictures into the pano, and then have to start over.
I found an easy way to "mark" your panos if you screw up or want to separate them (if you're shooting in the same spot) is just to put your hand out in front of the camera and shoot, then you know the next shot is a "new" set.
I didn't do that on my shoot yesterday, so now I have a directory full of thumbnails that look very similar (a few times I accidentally kicked my tripod while rotating around, thanks to wind and really cold temperatures..my brain wasn't working).
Ah, live and learn.
|
|
alert moderator
|
|
Peter Patricelli
Posts: 154
Location: Eugene, OR, United States
Registered: 20 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 20:41 GMT
|
|
David, I am mounting my pano head on the fast-detachable feature of a regular head which has it's own 3 D adjustments. It is a very good head whereas my pano head is Panosaurus, relatively cheap. I find it easier and more secure to rotate the tripod head rather than rotate the Panosaurus So, after leveling the tripod itself, I must check to be sure my pitch and roll of the tripod head are not off from some other use. The head has built-in pitch and roll bubble-levelers.
Dorin,
Lots of good suggestions there. I'm sure I will incorporate some.
A practice pano I shot in a windsurfing shop in Hood River, Oregon has been incorporated into their website.
www.bigwinds.com
It has one significant stitching error which I didn't notice until now. Was done with the Peleng 8mm. Now using the Nikon 10.5mm and realizing that some of my issues with the Peleng may have been technical camera issues and not the Peleng. As soon as I get a good chance I am going to do head-to-head panos with both lens equally processed and see what the resolution differences are.
Peter Patricelli
|
|
alert moderator
|
|
FourPiSr
Posts: 107
Location: Stockholm, Sweden,
Registered: 17 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 20:42 GMT
updated: 18 Jan 2007 at 20:46 GMT
|
|
And if you use a zoom-lens then don't forget to zoom out. When I take my camera out of the case the 18-55 mm lens is always pushed in to 30 mm. I don't know how many times I have had to start over because of that.
Lennart
|
|
alert moderator
|
|
Peter Patricelli
Posts: 154
Location: Eugene, OR, United States
Registered: 20 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 20:56 GMT
|
|
Dorin,
Something about your suggestion for closing the viewfinder just occured to me. Since I almost never, until now, shot off-camera I have never consciously had a light-into-viewfinder problem. In this application I will be determining exposure primarily with a light meter and also spot metering with the handheld camera, so exposure issues won't occur there. THEN the exposure is set and locked on MANUAL, so ambient light into the viewfinder cannot knock off the exposure. Can ambient viewfinder light actually get to the sensor during an exposure?
Peter Patricelli
|
|
alert moderator
|
|
DorinDXN
Posts: 2802
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 22:14 GMT
updated: 18 Jan 2007 at 22:16 GMT
|
|
Peter
Here is some extract from Canon 350D user guide page 50
-----8<----------------
Using the Eyepiece Cover
During self-timer or remonte control when your eyes does not cover the viewfinder eyepiece, stray light may enter the eyepiece and throw off the exposure. to prevent this, use the eyepiece cover (attached to the neck strap) to cover the eyepiece
-----8<----------------
Sure if your eye cover the the eyepiece or you use standalone light meter to determine the exposure may not be needed at metering, I do cover every time because I only take measurement with camera on tripod and I watch indicators on the display.
I'm also sure that when the sun is directly on eyepiece the light could actually get to the sensor, and when we taking pano is a chance for sun to be like that and not only outdoor, indoor when you have bright windows, in my last pano was needed for 3 seconds exposure, still having bright windows behind the camera and to avoid casting shadow in panos I do not stay on windows->camera or sun->camera axis to shadow the viewfinder from sun or bright windows
Sure depends on camera used. But having such cover on camera's belt make one think that maybe needed.
Dorin
|
|
alert moderator
|
|
John Houghton
Posts: 3412
Location: Hitchin, United Kingdom
Registered: 17 Jan 2005
|
Re: Work sequence for shooting VR panorama
Posted: 18 Jan 2007 at 23:02 GMT
|
|
Peter Patricelli said: It has one significant stitching error which I didn't notice until now.
It has 7 or 8 significant errors. I think it is a mistake to rotate the tripod head rather than the Panosaurus. If you rotate the Panosaurus, it doesn't matter at all if the panosaurus and the tripod head are not levelled. You can level the panorama during the stitching. If you rotate by the tripod head, you are unlikely to set the entrance pupil of the lens on the rotational axis as accurately as when you rotate the Panosaurus.
John
|
|
alert moderator
|
|
Peter Patricelli
Posts: 154
Location: Eugene, OR, United States
Registered: 20 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 19 Jan 2007 at 0:26 GMT
|
|
John,
First of all I greatly appreciate (as I am sure everyone here does)your willingness to share your expertise. Unfortunately I am so far down on the learning curve that I can only incorpor.ate a little of what you are saying at any one sitting. If there was a class, I would take it But I am reading your tutorials and working hard at it. Who knows when I will go beyond what PTGui does autopmaticlly and actually try the "advanced" mode, set control points, etc.. So far I'm just trying to optimize what it will do automatically, mostly by putting great effort into zero-ing the level and the entrance point. Imagine my horror when it was stated somewhere yesterday that fisheye lenses do not have a single nodal point and it is not IN the lens but in front of it. Every day is a revelation here.
Yes, I am aware of many errors in that pan. To me a "significant" error is 2(!) fin racks. To you a significant error is a visible misalignment in any straight line stitch. It was only my second pan and was not meant for commercial use but the store owner wanted it.
Your point about rotating the Panosaurus rather than the tripod is interesting. It never occured to me that the detachable mount might not be rotating on the same precise axis, but you are right. It might not. Certainly rotating the Panosaurus gaurantees that it will and removes any doubt. I'll do it that way so long as there are obvious Houghton-significant" stitching errors and when it gets perfect, try one the other way as well. It is much EASIER to rotate the head. But by then I suspect I will have invested in a much better head. Too bad NN does not fit either the Kodak DCS SLR or the D2X.
Thanks again and in advance.
Peter Patricelli
|
|
alert moderator
|
|
John Houghton
Posts: 3412
Location: Hitchin, United Kingdom
Registered: 17 Jan 2005
|
Re: Work sequence for shooting VR panorama
Posted: 19 Jan 2007 at 8:11 GMT
|
|
Peter Patricelli said: Imagine my horror when it was stated somewhere yesterday that fisheye lenses do not have a single nodal point and it is not IN the lens but in front of it.
Nodal points may be inside or outside the lens, but the no-parallax point is located at the entrance pupil, not the nodal point, and that is behind the front surface of the lens. The entrance pupil position is not at a single location; it varies according to the angle an incoming light ray makes to the lens axis. This is nicely illustrated in the photo by Michel Thoby at tinyurl.com/dy2f6 . So if you are taking 4 images around, the seams will be at 45 degrees to the lens axis, and it is the entrance pupil position at this angle that is important for avoiding parallax. Fortunately, you can see the entrance pupil as a bright spot if you look into the lens with it stopped down to f/22, say, and with the camera's viewfinder pointing at a bright source of light. If you move round from the front to look into the lens from the side, you will see the entrance pupil move forwards, but never outside the lens!
John
|
|
alert moderator
|
|
Peter Patricelli
Posts: 154
Location: Eugene, OR, United States
Registered: 20 Dec 2006
|
Re: Work sequence for shooting VR panorama
Posted: 19 Jan 2007 at 13:55 GMT
|
|
John, sorry to keep picking your brain but you keep throwing out curveballs.
"The entrance pupil position is not at a single location; it varies according to the angle an incoming light ray makes to the lens axis. This is nicely illustrated in the photo by Michel Thoby at tinyurl.com/dy2f6 . So if you are taking 4 images around, the seams will be at 45 degrees to the lens axis, and it is the entrance pupil position at this angle that is important for avoiding parallax"
So, functionally, the camera position on the head setting non-parallax point would be different for a 4 image rotation compared to a 6 image rotation? So, in setting non-parallax point one should use alignment test points that coincide, angularly, with the angle of the seams. So that if I am shooting 6 around with the 10.5mm that is a 60 degree FOV per stop and the seams will occur 30 degrees on either side so I should do my parallax test shots at +30 and -30 and NOT at the far edges of the lens?
Peter Patricelli
|
|
alert moderator
|
|
DorinDXN
Posts: 2802
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Work sequence for shooting VR panorama
Posted: 19 Jan 2007 at 15:59 GMT
|
|
Sorry that I'm interferring but for people like me who need to read two times in English to understand corectly (but having only one chance to write) I want to simplify the problem here.
The parallax is causing only stitching problems so the parallax must be corrected only for the objects in regions where one anticipates the seams, otherwise parallax is good, making us to realise our 3d world
So for us, the answer is yes, could be different NP for different number of shoots for a pano, so test and correct parallax better from now on.
Dorin
|
|
alert moderator
|
|
John Houghton
Posts: 3412
Location: Hitchin, United Kingdom
Registered: 17 Jan 2005
|
Re: Work sequence for shooting VR panorama
Posted: 19 Jan 2007 at 16:16 GMT
|
|
Peter Patricelli said: So that if I am shooting 6 around with the 10.5mm that is a 60 degree FOV per stop and the seams will occur 30 degrees on either side so I should do my parallax test shots at +30 and -30 and NOT at the far edges of the lens?
Yes, you've got the right idea. When testing for parallax, you rotate the camera just as you would when taking a panorama, and arrange for there to be something close the camera in the middle of the overlap area. It's common sense really.
John
|
|
alert moderator
|
|
back to threads list |
this thread is closed
Prev
