R1 Nikon Settings ?
I just bought the R1 ( jan/12) .. and since then i am trying to set up the R1 without sucess ... I am using Sigma 8mm f/3,5 for Nikon ... 04 shots ... at 0 degree ..
In the user guide manual ( http://www.nodalninja.com/Manuals/R1_QRG.pdf ) it is wrote to use NPP value 1.2 ...... but in this post ( http://www.nodalninja.com/forum/show...sigma-8mm-F3.5 ) it is wrote to use 2.25 and in the same post Heinz wrote that he used 2.2
I tried all the settings above ( 1.2 / 2.25 / 2.2 ) and also 1.2 / 1.3 / 1.4 / 1.5 / 1.6 / 1.7 / 1.8 / 1.9 / 2.0 / 2.1 / 2.2 / 2.3 / 2.4 and 2.5 ... and all settings i have stitching errors !!
Any idea ?
Liao, Stitching errors aren't necessarily due to poor setup of the head; your use of the stitching software might well be the problem. You can set up the R1 head accurately and easily in no more than 5-10 minutes, without shooting any images. See http://www.johnhpanos.com/epcalib.htm .
I use an R1 with a Sigma 4.5mm and stitch with PTGui - 4 shots round with Nikon DX, so with a Sigma 8mm 4 shots round should be OK for FX, but 6 shots round for DX (personally I would use 8 shots round for DX as it gives more overlap for using the Mask tool).
For a Sigma 8mm I would use the gold ring for the NNP
Hope this helps, Hugh.
Hi John and Hugh ...
I calibrated my R1 according to http://www.johnhpanos.com/epcalib.htm .... and i got the NPP at 3,40 ...( my lens is sigma 8mm f/3,5 and Fuji S5 -based on nikon d200 ) .... so i shooted 4 and 6 shots ... and i stitched with PTgui ... even 4 or 6 shots i got stitch errors ..
the photo are here ....( sorry .. i do not know how to attach photo here ) ... http://www.mediafire.com/?je7mfsiq6j47a17 and http://www.mediafire.com/?w7b0w90fiusob1i ... if you use the devalvr to see it .. you will notice that the first one the nadir is a perfect square and the second one is perfect six sided ... but in the zenith i have stitching errors ...
Any suggestions ?
Thanks a lot .
Add control points at the top of each image pair. I seldom get a great stitch relying only on automatic generated control points.
Also, delete all control points that are on image pairs that are not next to each other. That can throw off a stitch.
I usually do not see a panorama stitch properly on the very first try. It usually takes a couple of extra control points to set things right.
It would also help if you provided the full set of images so we can do a test stitch.
You can tilt this camera up 5 degrees and remove the need for the Zenith shot. 4 around + a hand held Nadir patch shot and you have captured the entire sphere.
Originally Posted by DennisS
Hi Dennis ..
In the PTgui i use automatic control points .. and after that i add some couple extra points .. but still i get stitching errors ..
I will try 5 degree up ..
anyway .. the original photos are here ... http://www.mediafire.com/?7342b848n9vb2h7 and here http://www.mediafire.com/?jf026o9uwsv69a8 .... 04 and 06 photos..
Thanks a lot ..
P.S. I will be off town this week .. i will back here to the forum in the next week .
Both sets stitched just fine on the very first try with no extra control points needed.
I opened PTGui, dragged over the image set, went to the crop tab and set the crop circle, Project Assistant tab, align images, deleted worst control point (only 1) and got a maximum error of 1.4 with the 4 pictures and a maximum error of 2.2 with the 6 pictures.
I would say your issues are not hardware. You are doing something when you stitch. There is a very slight parallax issue, but not enough to cause the stitching errors you are seeing. Try following the few steps I did.
Like Dennis, I see no stitching errors.
I am using PTGui Pro 9.1.3 and I used the following process with both the 4 and 6 image sets, then with all 10 images together:
1. Load images...
2. Align images...
Click the "Optimizer" tab and use "Heavy + lens shift"
3. Create panorama...
Check that the Width is 2 x the Height and enter a file name at "Output file:" and then click the "Output Panorama" button
I used no other steps and did not even look at the Control Points or Crop because the "Optimizer" results suggested that these were fine.
The results running the "Optimizer"were all "very good" and as follows:
AverageControl Point Distance 4 images (0.86) 6 images (1.04) 10 images (1.06)
MinimumControl Point Distance 4 images (0.06) 6 images (0.02) 10 images (0.07)
MaximumControl Point Distance 4 images (1.88) 6 images (2.51) 10 images (2.89)
It appears to me that your setup is "spot on" looking at the nadir.
Personally I would go for the 6 images round.
I did go back and look at the Control Points after creating the Equiractangulars and it is clear that the coverage is much better with the 6 images than the 4, but both cases have worked fine here.
I can email the Equirectangular and PTS files to you if you wish, or send them by WeTransfer if you prefer.
I hope this helps, but, like Dennis, I cannot see that there is anything wrong with your set up.
All the best, Hugh.
Hi Hugh ,
Thanks for replying ... I did step by step what you described and i still getting stiching error ... i will write you a PM with my email addr ... please , can you send me the files ?
Thanks a lot
Originally Posted by Hugh
Hi Dennis ,
thanks for replying ... i did not get what you did ... please , can you describe step by step ?
Originally Posted by DennisS
Drag and drop the image set,
Set the crop circle,
Project Assistant tab,
Select "Control Points" on the main menu,
Delete worst control points.
To respond to the original question about the NPP for the R1 with a Sigma 8, it would seem that the recommended setting of 1.2 is just incorrect. I just recently got the R1/Sigma8 and in my tests it seems like closer to 3 gets the results we are looking for. Here is a quick test image I did, shooting 3 shots for each setting (1.2 and 3). I used the my Pole Series A on a tripod in my office about 3 feet in front of camera, aligned this with a straight light out side the window at about 5o feet and then the garage on the house across the street. Center shot is straight on, then left and right are from peripheral view from 45deg each way. Clearly we can see that 1.2 is not what we are looking for. At 3 its still not perfect but a hell of a lot closer.
A few more tests and 2.9 is the sweet spot for me. (at 0 deg)
Nick or Bill, can you comment on why my results differ so much from the recommended setting of 1.2 that is here. http://www.nodalninja.com/qr/r1-r10/...ltimate-r1/17/ (Sigma 8mm f/3.5 DG for Nikon - 4 shots around)
Hi Hugh ..
Sorry for the latte reply ... i was off the town ... i came back these days ...
Thanks for your email and the attached files ...
I opened your file ( 4 images ) ... and i done step by step what you did .. but i still getting stitching errors ...
I sent you a email with attached my pts file and the print screen of your file settings and my file settings .