Need help / Feedback on my workflow

[Sukima]

I need some feed back on how I work with making panoramas. I documented my work on my blog and copy/pasted it here. I would really appreciate any advice. Thanks.

I often thought my workflow was a bit unique. Though recently I found that it is not that unusual. I wanted to document the workflow for reference and because I've been running into some difficult results in my final panoramas and having the steps I took to get there might allow myself to get better feedback from others. At the bottom of this post is the final output. For those interested I am also offering my original images (in RAW format) to see if anyone might be willing to offer advise. I don't know if my troubles stem from my Lens, my Camera, my software, or my workflow. Any advice would be greatly appreciated.

Taking the pictures

I am using a Canon EOS Rebel XS (A.K.A. EOS 1000D) with a Bower 8mm Fisheye lens (A.K.A Pro-Optic, Samyang, or Rokinon). I mount my camera on a Panosaurus panoramic head on a tripod. I believe I've calibrated the parallax correctly to the best I could through the view finder and by taking several pictures with a close object and far object and rotating back and forth.

I setup the tripod and start taking photos every 60 degrees (six images) around, one zenith, two nadir images 180 degrees from each other, and a final hand held nadir just in case.

I use a wired remote shutter to lessen the jitter. I try to sample several angles before picking the final settings I use in manual mode. Since the lens doesn't have any communication with the camera I manually dial in the F-stop. I set the focus to infinity. I use the Raw format.

Pre-processing photos

I used RawTherapee to pre-process the Raw files I downloaded from the camera. Since the auto settings seemed pretty good I kept them except I set the white balance to Shade. I batch saved all the images to 16 bit TIFFs (Uncompressed).

Stitching

I used Hugin set to the Expert mode. I'm using version 2014 on the Mac because (at the time of this article) it is the only version available for the Mac. In Expert mode I add all the images including the hand held nadir. I used the Full Frame Fisheye with the Focal Length of '8' which in the past seemed like it worked. Mostly guess work here. Anyway Hugin auto populates the field of view for me and I go with it.

Than I opened up the mask section and carefully draw exclusion masks around the tripod in all three nadir shots. Finally I click optimize and then open the GL Previw window. And everything looked great. I then opened the stitcher view and selected a width of 1024 so the rendering would be super fast while I previewed. I found that the GL Preview glossed over some artifacts that you don't notice till the final renders.

I noticed that the hand held nadir shot (although almost aligned) had parallax issues. Mainly since the nadir was a set of wood boards. So I gave up on the idea that the hand help nadir image was worth anything and deleted it from the image list in Hugin. I went back to the stitcher and selected Calculate optimal size and created a final version. Now with a gaping hole where the mask for the tripod was.

Editing the nadir

Since I use the GIMP to edit photos I have to convert the 16 bit TIFF to an 8 bit TIFF using GraphicsMagick:

$ gm convert MillPano.tiff -depth 8 MillPano8.tif

Then I use the Panotools scripts to create my set of cubes:

$ erect2cubic --erect=MillPano8.tif --ptofile=cubic.pto
$ nona -o cube cubic.pto

In the GIMP I edited cube0005.tif and used the clone tool to carefully reconstruct the wood planks from the sides (this was challenging using a touchpad). Saved it to cube0005a.tif.

Finally I reconstructed the panorama:

$ cubic2erect cube000{0,1,2,3,4,5a}.tif MillPanoFinal.tif

And this is the result:

Ye Olde Town Mill

Questions

1. Is there a way to get Hugin to use a hand held nadir shot when the shot doesn't line up well with the shots taken via the tripod head? Would make nadir editing easier.
2. Why is the final output blurry? Is it chromatic abrasion? Bad stitching? Crappy camera/lens?

Download the Original Images. (100MB Zip)

[dgjohnston]

My first recommendation would be for you to find out the hyperfocal focus distance for your aperture, lens, and camera combination; then use that to set your focus rather then infinity. For example your XS with an 8mm lens at f16 would use a focus distance of 8 meters.

Don Johnston

Sent from my Samsung device

--
A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ
---
You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hugin-ptx+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/9f5afc45-afe8-4f5b-b868-83796c8aba5e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[panostar]

On Wednesday, April 20, 2016 at 4:58:40 PM UTC+1, Donald Johnston wrote:

My first recommendation would be for you to find out the hyperfocal focus distance for your aperture, lens, and camera combination; then use that to set your focus rather then infinity. For example your XS with an 8mm lens at f16 would use a focus distance of 8 meters.

I don't think that can be optimum.  People commonly set the focus at 1m for an 8mm lens.  F/16 won't necessarily give the sharpest images either, as smaller apertures suffer from diffraction effects.  Its best to do some careful tests to establish the best focus distance and aperture combination.  I had the Falcon version of this lens and found f/11 worked best for me.

The setup of the panorama head is good, but more care needs to be taken with the handheld nadir shot to make sure it captures a large enough area.  Here, the tips of two two feet leave holes at the nadir that have to be patched.  I did a stitch with PTGui using viewpoint correction for the nadir, but Hugin has a similar feature that can be used (see http://wiki.panotools.org/Image_positioning_model ).  FWIW, my stitch is uploaded to https://www.sendspace.com/file/gajmwr .

John

 

[Donald Johnston]

Thanks John. Sorry, I read the wrong figure for the hyperfocal distance. With the numbers quoted it is 0.233m (8m seemed high but it didn’t click in my brain properly). At f/11 it moves to 0.326m.

The point being that infinity certainly isn’t the right number!  And you’re more likely to run into diffraction problems at f/22 then f/16. If you want to set the focus at 1m you can use an aperture setting as low as f/3.5 and still have good focus from 0.5m out to infinity (according to the calculators anyway; try out various settings yourself and see what sharpness you get with far away objects).

--
A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ
---
You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hugin-ptx+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hugin-ptx/03619f1b-e72c-4579-839a-2b95e16e6b92%40googlegroups.com.

[Stefan Peter]

Hi Sukima,

On 20.04.2016 15:44, Sukima wrote:
<snip>
> I used RawTherapee <https://tritarget.org/#RawTherapee> to pre-process

> the Raw files I downloaded from the camera. Since the auto settings
> seemed pretty good I kept them except I set the white balance to

> *Shade*. I batch saved all the images to *16 bit TIFF*s (Uncompressed).

If you do full sphere panos, make sure to disable lens corrections and
vignetting correction. Hugin will do a much better job for your specific
lens/camera.

> the images including the hand held nadir. I used the *Full Frame
> Fisheye* with the Focal Length of '8' which in the past seemed like it

The Samyang 8mm lens is a stereographic projection lens, see
http://michel.thoby.free.fr/SAMYANG/Early%20test%20report.html
for more information. And, according to my experience, the projection
selection influences the control point detection, so just changing the
projection in a existing pto won't help.

Make sure you set the "focal length multiplier" correctly four your
camera, that should be around 1.6 for the 1000D.

> around the tripod in all three nadir shots. Finally I click optimize and

> then open the /GL Previw/ window.

What exactly did you optimize for? In expert mode, you can choose
between several optimization levels. I normally start with "Positions
(Incremental from anchor)" and work my way down to "Everything without
translation". And then I "Show control points" icon in the top bar, sort
the control points by distance and inspect all the points with a
distance >10. Or, if I'm lacy (so most of the time) I use the "select by
Distance" button, select everything above let's say 10, and then delete
them. Then I redo the last optimization, and do some more pruning of
control points. Rinse and repeat to your liking, but be aware that you
may end up with unconnected images if you overdo this, but practise
makes perfect ;)

Another option would be the "Actions->Control Points->keep 5 CPs per
image pair" menu entry. But this uses the Phyton plugin interface and I
don't know if this has been packaged in the 2014 Mac version you use.

> And everything looked great. I then

> opened the stitcher view and selected a width of *1024* so the rendering

> would be super fast while I previewed. I found that the GL Preview
> glossed over some artifacts that you don't notice till the final renders.

You can not really trust the preview if you are aiming for quality: If
it would be possible to assemble a quality panorama in realtime, we
would not need the PTBatcher. On the other hand, if your preview just
shows a crazy tangle of lines, the final image will not be any better.

And if you want a high quality, sharp image, don't limit the output to
1024 pixels. In my setup the final image width is around 6000 pixels and
by condensing this to 1024 pixels by combining roughly 6 pixels to one,
you definitely will loose sharpness. If I need a smaller image, I
normally use the panotools-scipt erect2cubic script from
http://search.cpan.org/dist/Panotools-Script/
resize and sharpen thcubic images in gimp and then reassemble them with
cubic2erect.

The setting of File->Preferences->Programs->Nona->Default
interpolator(i) may play a role in this, too. But I'm not versed well
enough to give a recommendation here.

>
> I noticed that the hand held nadir shot (although almost aligned) had
> parallax issues. Mainly since the nadir was a set of wood boards. So I

Most probably your hand held shot was not from the exactly same point as
the ones from the tripod. In this case, you would need to apply a second
lens to your nadir shot and optimize this image for xyz. But I have
never been able to pull this off successfully, so your mileage may vary.


I hope this gives you some starting points.


With kind regards

Stefan Peter


--
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
(See https://en.wikipedia.org/wiki/Posting_style for details)

[Markku Kolkka]

20.4.2016, 19:53, panostar kirjoitti:
> On Wednesday, April 20, 2016 at 4:58:40 PM UTC+1, Donald Johnston wrote:
>>
>> My first recommendation would be for you to find out the hyperfocal focus
>> distance for your aperture, lens, and camera combination; then use that to
>> set your focus rather then infinity. For example your XS with an 8mm lens
>> at f16 would use a focus distance of 8 meters.
>>
>
> I don't think that can be optimum. People commonly set the focus at 1m for
> an 8mm lens. F/16 won't necessarily give the sharpest images either, as
> smaller apertures suffer from diffraction effects. Its best to do some
> careful tests to establish the best focus distance and aperture
> combination.

One more thing: the focusing scale on Samyang 8mm lenses is often
(usually?) incorrectly calibrated so you can't just set it to 1m or
infinity and trust that it's actually focused at that distance. You need
to focus each shot using liveview or calibrate the distance scale
yourself. You can find the instructions for calibration here:
https://hdrtist.wordpress.com/2011/11/11/rokinon-8mm-fisheye-lens-review/

[Sukima]

On Wednesday, April 20, 2016 at 5:15:15 PM UTC-4, markku...@iki.fi wrote:

One more thing: the focusing scale on Samyang 8mm lenses is often
(usually?) incorrectly calibrated so you can't just set it to 1m or
infinity and trust that it's actually focused at that distance.

This would explain why the panos I do indoors look good while many of my outdoor ones are blurry. And here I thought it was the stitching process and/or abrasion since the lens and the camera did not cost a small fortune. I'm going to try the calibration tonight and see if that was the culprit. Thank you!

[dkloi]

My 2 cents:

First of all, you don't need to use manual exposure. Use AE + bracketing if required. Hugin will photometrically optimise to adjust relative brightness of each iamge.

I find that a setting of 9mm works out well as the initial setting of the focal length (this will be optimised later).

For the 8mm/3.5, I used f/8. The field of focus can be quite curved so I use liveview to focus at max aperture so that the ground near the nadir is in sharpest focus, then stop down. This is because when I blend in the nadir shot, the sharp central part of the nadir shot needs to blend in nicely with the bottom of the horizontally taken shots so I want this region to be as sharp as possible. Stopping down to f/8 will mean that infinity objects will be sufficiently sharp even if the focus is 1m to 2m.

Pitch your lens down so that you have a smaller nadir hole. You can lower it until the panohead just comes into view on the bottom of the frame. The Panosaurus does have a large nadir footprint unfortunately.

Processing, I set WB to be the same for all images. I'll tweak the exposure of each image to minimise the under or overexposed parts of the image. In non-overlapping regions I may do some dodging and burning. I'll do vignetting correction at this stage.

In Hugin, I give the zenith and nadir shots their own lens (3 lenses in total). I mask out the tripod and panohead and set crop circles for zenith and nadir. Mask out moving objects if required. I'll usually add CPs by hand. Optimise position, then view and position and barrel, then everything except translation. Check CP errors, tweak, reoptimise. Next is optimization of exposure, usually just exposure and WB without vignetting or EMOR.

The nadir is added just like any other image. I connect it to the horizontally taken shots with manual CPs. During geometry optimization, you might get large optimization parameters for the nadir lens but don't worry about this. Usually I don't need to use shear correction.

Your setup will support 10K x 5K equirectangular panos. I use Sinc256 interpolation in Nona.

IMG1095 is unsharp.

Blown highlights, need to braccket to capture them. Could also bracket to get dark areas.

I attach my stitch and the PTO file (I used LR to process the raw files so it may not work perfectly well with Rawtherapy output files due to differing pixel dimensions). I have not edited the nadir outside of Hugin. I have done some minor post-processing for contrast curve, sharpening and saturation boost. You can see the holes left by your feet in the nadir. The part of the pano where IMG1095 covers is blurry, the rest of it is much sharper.

Daniel.

[Sukima]

On Wednesday, April 20, 2016 at 10:36:10 PM UTC-4, dkloi wrote:

For the 8mm/3.5, I used f/8. The field of focus can be quite curved so I use liveview to focus at max aperture so that the ground near the nadir is in sharpest focus, then stop down. This is because when I blend in the nadir shot, the sharp central part of the nadir shot needs to blend in nicely with the bottom of the horizontally taken shots so I want this region to be as sharp as possible. Stopping down to f/8 will mean that infinity objects will be sufficiently sharp even if the focus is 1m to 2m.

I've tried several times to figure out the focus with the fisheye. Every time I can't seem to tell the difference between infinity and 0.4 ft! Maybe I need my eyes checked.
 

Pitch your lens down so that you have a smaller nadir hole. You can lower it until the panohead just comes into view on the bottom of the frame.

 
Do you mean when I'm hand holding the tripod assembly to get the portion of the nadir that is covered with the tripod footprint into the center of the fisheye lens that I should lower the camera closer to the ground? Doesn't this make parallax impossible to manage? I thought all shots had to have nearly the same or super close fixed point in space to prevent stitching problems. If the camera lowers the scale of the shot will be all distorted compared to the rest of the shots. Also don't I have to have the camera be perpendicular to the ground. Shooting at an angle to get my feet out of the way would make stiching impossible also. Usually my feet are there because I don't have the dexterity to hold the tripod assembly at arms length in the air perpendicular to the ground at the same height as the previous shots while avoiding any moving or wind.

The YouTube tutorials make it look so easy. the guy spins around once click click with his arm extended and amazing it all works in one go. I always assumed he had massive equipment (thousands of dollars on cameras and lenses and software) and massive upper arm strength to stay still. My I'm poor and all I can do is free software and the lowest end DSLR from Best Buy. I was lucky to have the fisheye I have.

The Panosaurus does have a large nadir footprint unfortunately.

 
Yeah. That is why I take as many nadir shots as I can.

Processing, I set WB to be the same for all images. I'll tweak the exposure of each image to minimise the under or overexposed parts of the image.

You can have differing exposures per image?! I was always under the impression this was a huge enblend no-no.
 

I'll do vignetting correction at this stage.

What is vignetting correction? How do I know my images are vignetted since I don't see any black hazy borders in my fisheye photos.
 

In Hugin, I give the zenith and nadir shots their own lens (3 lenses in total).

 
Three different lenses same settings though?

I mask out the tripod and panohead and set crop circles for zenith and nadir.

I do that. Hadn't thought about the cropping though. Will try that next.

 

I'll usually add CPs by hand.

Any advice on this. Most attempts I've done result in lots of parallax so I depend almost exclusively on the auto finder. I had always thought more is better so I would try to generate a lot.

 

Optimise position, then view and position and barrel, then everything except translation. Check CP errors, tweak, reoptimise. Next is optimization of exposure, usually just exposure and WB without vignetting or EMOR.

Whoa that is a lot. Do you know any resources that would help train novice Hugin users on these techniques?
 

Blown highlights, need to braccket to capture them. Could also bracket to get dark areas.

Does saving multiple exposures from a single Raw file count or is bracketed shots the way to go here? I tend to have trouble getting the stacks to work well in Hugin any advice?
 

I have not edited the nadir outside of Hugin. I have done some minor post-processing for contrast curve, sharpening and saturation boost. You can see the holes left by your feet in the nadir.

Yours came out amazing! I didn't know you could do so well with what I had. Thank you so much. Now I know that it is possible just need to learn and practice.

The part of the pano where IMG1095 covers is blurry, the rest of it is much sharper.

I think that was the wind. I should have taken more shots.

[dkloi]

On Friday, 22 April 2016 00:13:52 UTC+1, Sukima wrote:

I've tried several times to figure out the focus with the fisheye. Every time I can't seem to tell the difference between infinity and 0.4 ft! Maybe I need my eyes checked.

If you camera doesn't have liveview, then you'll can find the proper focus distance the long and tedious way by taking a series of shots at different focus settings (make a note of the settings for each photo) and then examine them on your computer (zoom) to figure out which one had the best focus. It's practically impossible to critically focus a fisheye lens using a standard optical viewfinder.

  

Pitch your lens down so that you have a smaller nadir hole. You can lower it until the panohead just comes into view on the bottom of the frame.

 
Do you mean when I'm hand holding the tripod assembly to get the portion of the nadir that is covered with the tripod footprint into the center of the fisheye lens that I should lower the camera closer to the ground? Doesn't this make parallax impossible to manage? I thought all shots had to have nearly the same or super close fixed point in space to prevent stitching problems. If the camera lowers the scale of the shot will be all distorted compared to the rest of the shots. Also don't I have to have the camera be perpendicular to the ground. Shooting at an angle to get my feet out of the way would make stiching impossible also. Usually my feet are there because I don't have the dexterity to hold the tripod assembly at arms length in the air perpendicular to the ground at the same height as the previous shots while avoiding any moving or wind.

This isn't about the nadir. What I meant was that instead of taking your 6 shots around at 0 degree pitch angle (i.e. shooting perfectly horizontally), tilt the lens down a few degrees.

For nadir shots, there are various ways of using the tripod to support the camera so that even long exposures are possible. E.g. http://www.rosaurophotography.com/html/technical6.html For flat surfaces, you can get away with not having the nadir shot at the same height. For a nadir where parallax may be a problem, you can raise the camera to the correct height by extending the centre column whilst the tripod is positioned to the side and the camera pointed down.

 

You can have differing exposures per image?! I was always under the impression this was a huge enblend no-no.

It's an "old wives tale" that you have to lock exposure for taking good panos, not with today's software including Hugin. Hugin can optimise the brightness of each image before feeding them to enblend. A little bit of shadow and highlight recovery in Lightroom usually doesn't present a problem.

 

I'll do vignetting correction at this stage.

What is vignetting correction? How do I know my images are vignetted since I don't see any black hazy borders in my fisheye photos.

There can be gradual light fall-off as you go away from the centre of the image. According to http://www.photozone.de/canon-eos/526-samyang8f35eos?start=1 at f/5.6 or smaller, the corners are 0.7EV darker than the centre. Have a look here for examples of various types of vignetting http://www.dxo.com/us/photography/community/tutorials/correcting-vignetting-dxo-optics-pro-9

 

In Hugin, I give the zenith and nadir shots their own lens (3 lenses in total).

 
Three different lenses same settings though?

Have a look at the PTO file to see what the lens parameters are for the different lenses. This is especially important for getting a good stitch of the nadir. It sometimes helps for the zenith since the seams between it and the horizontally taken images are at different parts of the image field compared with seams between the horizontally taken photos themselves. In your example, I used 4 lenses with optimised FoV 97.9, 97.1, 99.7, and 407.9 degrees. The last one was the nadir shot and the other lens parameters are also quite extreme.

 

I mask out the tripod and panohead and set crop circles for zenith and nadir.

I do that. Hadn't thought about the cropping though. Will try that next.

I forgot to mention I mask out the corners of the horizontally taken photos as they are usually a lot less sharp than the central portion of the image that gets stitched. It also helps control where the seams gets placed.

  

I'll usually add CPs by hand.

Any advice on this. Most attempts I've done result in lots of parallax so I depend almost exclusively on the auto finder. I had always thought more is better so I would try to generate a lot.

Simply place a string of CPs along where you think the seam would be between overlapping images. Pay special attention to lines or patterns that would should discontinuities due to alignment errors. Give a good spread over different radii with respect to the lens axis.

 

Optimise position, then view and position and barrel, then everything except translation. Check CP errors, tweak, reoptimise. Next is optimization of exposure, usually just exposure and WB without vignetting or EMOR.

Whoa that is a lot. Do you know any resources that would help train novice Hugin users on these techniques?

Practice :-). Hanging out here or other pano forums helps too.

 

Blown highlights, need to braccket to capture them. Could also bracket to get dark areas.

Does saving multiple exposures from a single Raw file count or is bracketed shots the way to go here? I tend to have trouble getting the stacks to work well in Hugin any advice?

You can try saving multiple versions from a single raw file but bracketing usually works better. I found it a bit difficult to recover much highlight detail from your raw files unfortunately, hence the burnt out clouds. Canon had been lagging some of the other manufacturers when it comes dynamic range (though apparently the 80D has now caught up somewhat with its peers). I tend to use exposure blending to cope with high DR scenes. I'll take a normal set of images, then I'll only take images of the bright or dark areas with exposure compensation to capture the detail missing from the normal set. In Hugin, I'll stitch together the normal set and produce a normally exposed equirect. I'll then mask in the over or under exposed areas, and then produce dark and light versions of the panos with the mission detail in the highlights and shadows respectively. The light, mid and dark versions of the pano are then combined with enfuse.

 

I have not edited the nadir outside of Hugin. I have done some minor post-processing for contrast curve, sharpening and saturation boost. You can see the holes left by your feet in the nadir.

Yours came out amazing! I didn't know you could do so well with what I had. Thank you so much. Now I know that it is possible just need to learn and practice.

Examine the PTO file to see how I've defined lenses, did the masking, set control points, and the optimized geometry and exposures. The first thing is to get the mechanics of taking the source images and stitching them. Afterwards, you'll encounter different situations that will raise issues with exposure, shadows, movement, colour balance, extreme contrast, flare, etc. That's how you build up experience. I'm still learning new ways of handling problems or else pushing myself to take trickier panos.

The part of the pano where IMG1095 covers is blurry, the rest of it is much sharper.

I think that was the wind. I should have taken more shots.

One thing you'll learn to do is "chimp" your shots :-). There's nothing worse than to spend a lot of time taking source images and later finding out when you get back home that you missed some little thing. I've managed to forget taking zeniths or nadirs, or found out that a shot was blurry, or I'd forgotten to change the ISO to minimum after using the camera for something else, or a whole host of ways you can mess up a pano.

[Carl von Einem]

Hi Sukima,

Sukima wrote on 22.04.16 01:13:

> I've tried several times to figure out the focus with the fisheye. Every
> time I can't seem to tell the difference between infinity and 0.4 ft!
> Maybe I need my eyes checked.

I use the same type of lens (branded as "Walimex pro" AE 8mm 1:3.5
Fish-eye CS) and prior to shooting panoramas I made some test shots. I
was aware that the manufacturer put only little effort into adjusting
the flange focal distance (see
<https://en.wikipedia.org/wiki/Flange_focal_distance>). I found that for
my lens the actual position for infinity is when the index just hits the
0.4m mark on the scale... instead of adjusting the flange distance
myself I just lazily use that 0.4 mark, it works for me :-)

Your test images are a bit soft compared to what I get from my lens. I
always keep the aperture between 8 and 5.6 since smaller apertures (like
11 or 16) tend to produce soft images. Also keep in mind that for the
hand held nadir shot you will probably want to use a shorter speed than
the 1/15 sec. you used.

>> Pitch your lens down so that you have a smaller nadir hole. You can
>> lower it until the panohead just comes into view on the bottom of
>> the frame.
>
> Do you mean when I'm hand holding the tripod assembly to get the portion
> of the nadir that is covered with the tripod footprint into the center
> of the fisheye lens that I should lower the camera closer to the ground?
> Doesn't this make parallax impossible to manage? I thought all shots had
> to have nearly the same or super close fixed point in space to prevent
> stitching problems. If the camera lowers the scale of the shot will be
> all distorted compared to the rest of the shots. Also don't I have to
> have the camera be perpendicular to the ground. Shooting at an angle to
> get my feet out of the way would make stiching impossible also. Usually
> my feet are there because I don't have the dexterity to hold the tripod
> assembly at arms length in the air perpendicular to the ground at the
> same height as the previous shots while avoiding any moving or wind.

It really helps to train shooting with a philopod. Depending on the
situation also the shadow of your camera can be used as a great virtual
"mark" for a close to perfect position of your camera.

> The YouTube tutorials make it look so easy. the guy spins around once
> click click with his arm extended and amazing it all works in one go. I
> always assumed he had massive equipment (thousands of dollars on cameras
> and lenses and software) and massive upper arm strength to stay still.
> My I'm poor and all I can do is free software and the lowest end DSLR
> from Best Buy. I was lucky to have the fisheye I have.

It's a great and compact lens! The tutorials you see just show that
those guys have some experience, and when you have a nice workflow you
will see that holding a camera for a short moment doesn't need too much
strength.

> You can have differing exposures per image?! I was always under the
> impression this was a huge enblend no-no.

I also shoot full spherical panoramas with a fixed (manual) exposure,
mostly using bracketing (+/0/- 1 f-stop) and at home selecting the best
exposure series from my raw files.

Carl

[Carlos Eduardo G. Carvalho (Cartola)]

2016-04-22 6:56 GMT-03:00 Carl von Einem <ca...@einem.net>:

Do you mean when I'm hand holding the tripod assembly to get the portion
of the nadir that is covered with the tripod footprint into the center
of the fisheye lens that I should lower the camera closer to the ground?
Doesn't this make parallax impossible to manage? I thought all shots had
to have nearly the same or super close fixed point in space to prevent
stitching problems. If the camera lowers the scale of the shot will be
all distorted compared to the rest of the shots. Also don't I have to
have the camera be perpendicular to the ground. Shooting at an angle to
get my feet out of the way would make stiching impossible also. Usually
my feet are there because I don't have the dexterity to hold the tripod
assembly at arms length in the air perpendicular to the ground at the
same height as the previous shots while avoiding any moving or wind.

It really helps to train shooting with a philopod. Depending on the situation also the shadow of your camera can be used as a great virtual "mark" for a close to perfect position of your camera.

I would also recommend a philopod for the nadir. In fact nowadays I practically only shoot with it, without tripod at all, unless when it is totally necessary. You can try to use it only for the nadir, and you don't need to turn the camera down, just turn it enough to an angle that reaches the nadir and make 2 or 3 pictures around.

The workflow could be like: when shooting with the tripod, use a philopod to mark the nadir point with a coin or any small object. Do your pictures with the tripod, then do the nadir pictures with the tripod.

Some videos:

https://www.youtube.com/watch?v=ouOEM4cKKGc

https://www.youtube.com/watch?v=znT-kzOXf-A

And here a forum post where I've suggested the use of a philopod + tripod in cases where you may need long expositions, for example (it is in Portuguese, but images can illustrate the idea): http://www.panoforum.com.br/forum/viewtopic.php?f=13&t=506

Carlos E G Carvalho (Cartola)
http://cartola.org/360
http://www.panoforum.com.br/