DennisS
Posts: 1210
Location: Los Anglels, United States
Registered: 1 Sep 2007
|
Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 0:02 GMT
|
|
I am interested only in the finished product as published on the internet. Each of the following three panoramas were shot with the exact same camera/lens/pano head, processed exactly the same with PTGui, had "Unsharpen Mask" done in Photoshop and used the exact same project file in Pano2VR. The final file sizes are right at 2.3 megs. 6, 12, and 24 are the number of shots taken around. No Zenith or Nadir shots taken here.
www.dlsphoto.net/PanoTest/PanoTest.htm
If someone has a different combination of hardware that yields different results, please post them so we can all see the differences and contribute to the knowledge base.
You can decide for yourself if there is an advantage in taking more pictures when there is going to be so much detail removed for publication.
|
|
alert moderator
|
|
David Tan
Posts: 155
Location: Kualalumpur, Malaysia
Registered: 31 Mar 2010
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 3:12 GMT
|
|
Dennis,
Thanks for taking this kind of experiement. I have been wondering the same thing also and finally somebody did it! 
Here is my observation:
Pano A = 6 shots
Pano B & C = either 12 or 24
Both B and C look sharper than A
There is no noticeable difference between B and C
I think the result from 12/24 shots is better than 6 shots.
However somebody has to pay attention to details to see the difference between A and B/C.
I think 12 shots will be beneficial to produce higher detailed of image, especially in a bookstore:
www.mosscreekmedia.com/2010/05/timothy-whelan-pho...
But for outdoor scenery or real estate, I think most people don't pay attention to detail, thus 6 shots are sufficient.
Please correct me if I am wrong. Thanks
|
|
alert moderator
|
|
|
Guest
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 3:21 GMT
updated: 14 Mar 2011 at 16:13 GMT
|
|
My guess:
A=6
B=12
C=24
|
|
|
|
Nick Fan
[NodalNinja]
Posts: 696
Location: Hong Kong
Registered: 26 May 2006
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 6:35 GMT
|
|
Hi Dennis,
is the pano at max resolution? what is the resolution?
nick
|
|
alert moderator
|
|
Juergen Schrader
Posts: 217
Location: Germany
Registered: 14 Jul 2006
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 6:36 GMT
updated: 31 Aug 2010 at 6:39 GMT
|
|
B, A, C (is worst)
But the most obvious thing is, that someone in this house needs some driving practice.
SCNR 
Jürgen
|
|
alert moderator
|
|
klaus mayer
Posts: 140
Location: Australia
Registered: 15 Jan 2008
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 6:45 GMT
|
|
A is not as sharp as B and C.
B and C are similar but C has better colours which is quite obvious around the front tyre area of the Toyota. And you should get those panels fixed before the rust takes hold of them
A is probably 6 images. Did you have to increase the size of A or was this the calculated maximum size?
My preference is to stitch at maximum size. The final result can always be resampled for the internet but bandwith and speed get better every year and the big file may come in handy in coming years.
|
|
alert moderator
|
|
Smooth
Posts: 3729
Location: Mount Panorama, Australia
Registered: 21 Jul 2004
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 7:40 GMT
|
|
OK, I'll bite!
B, C, A.
B being the sharpest, A being the fuzziest. I guess I'm saying B=24 C=12 and A=6
I'm not sure this is a fair test though, as you haven't really given us much detail. We "assume" all used the same focus, same shutter speeds, aperture and of course same camera and lens. Then we consider the if it was the same settings in PTGui and Pano2VR and all other post processing. There are noticeable warping differences in some areas between these 3 panoramas and this will effect how the blending takes place of those areas.
But yes, there isn't a huge amount of difference between them all. Still, it's that 1% we are all looking for!
Regards, Smooth 
|
|
alert moderator
|
|
enbilaman
Posts: 119
Location:
Registered: 3 Mar 2006
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 7:53 GMT
updated: 31 Aug 2010 at 8:01 GMT
|
|
Aging vision was not helping me but C looked slightly sharper than B. A looked the softest of all.
Then I tried to guess the number of stitched images:
I was first tempted to give a grade to the 3 images by counting the number of green flare spots that hide in the leaves of the trees against sunlight.
But I could not decide for sure and I changed my mind several times without having a firm opinion.
Anyhow, instead of using binoculars, I have attempted to analyze the file in another way: assuming that the source file Bytes count would be bigger as the sharpness increases, I looked for the exact numbers in the cache of Safari. I have posted an illustrated overview where I also show the green flare spots as well as some other specifics of the image.
B>C>A:
I then propose A=6 and B=24
Michel
|
|
alert moderator
|
|
Nick Fan
[NodalNinja]
Posts: 696
Location: Hong Kong
Registered: 26 May 2006
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 9:59 GMT
|
|
DennisS said: I am interested only in the finished product as published on the internet. Each of the following three panoramas were shot with the exact same camera/lens/pano head, processed exactly the same with PTGui, had "Unsharpen Mask" done in Photoshop and used the exact same project file in Pano2VR. The final file sizes are right at 2.3 megs. 6, 12, and 24 are the number of shots taken around. No Zenith or Nadir shots taken here.
Hi Dennis,
Let me know if I get it right.
you just took 24 shots around, then you stitched them in 3 ways.
1. use all 24 images.
2. use 12 images by skipping one image.
3. use 6 images by skipping 3 images.
am I right?
Nick
|
|
alert moderator
|
|
Hans Nyberg
Posts: 2760
Location: Denmark
Registered: 28 Aug 2005
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 11:03 GMT
|
|
I am sorry to say but this is a ridiculous test which just shows that you have not understood how it works.
I assume as you have not done a 4 images test that this is the Nikkor 10.5mm.
But you have restricted the view to around 65 degree vertically.
Of course that means that you will not get any difference in quality.
The area which might show a difference (if you had done it in max resolution which this is not) is now out of the view.
The area you have to look at in a pano from a fullframe fisheye is at +- 65 degree. Thats were you blend the zenith.
If you shoot 6, 8, 10 or 12 around does not really give much difference unless you also shoot a row at +-45 degrees.
But then I could just as well use a 28mm and get the extra resolution.
When I do churches were I actually have some important data at the upper areas I often shoot the zenith with 3 images at 45 degrees. This helps especially for glass mosaic windows.
Hans
|
|
alert moderator
|
|
DorinDXN
Posts: 2802
Location: Timisoara, Romania
Registered: 14 Nov 2006
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 11:18 GMT
updated: 31 Aug 2010 at 11:20 GMT
|
|
I would add that the images looks like bit out of focus, still some unsharpen mask are stated as was used too.
Maybe Dennis can explain this, i.e. how he see the focus of the source images used comparred with other images he done before.
Dorin
|
|
alert moderator
|
|
David Tan
Posts: 155
Location: Kualalumpur, Malaysia
Registered: 31 Mar 2010
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 12:28 GMT
|
|
Hans Nyberg said: But then I could just as well use a 28mm and get the extra resolution.
Hans, you mean Canon EF 28mm f/1.8?
Hans Nyberg said: When I do churches were I actually have some important data at the upper areas I often shoot the zenith with 3 images at 45 degrees. This helps especially for glass mosaic windows.
I didn't know this trick. I think this makes sense and I will try the 45 degrees. Thanks for sharing!
|
|
alert moderator
|
|
Hans Nyberg
Posts: 2760
Location: Denmark
Registered: 28 Aug 2005
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 12:34 GMT
|
|
David Tan said: Hans Nyberg said: But then I could just as well use a 28mm and get the extra resolution.
Hans, you mean Canon EF 28mm f/1.8?
For example. I use an old Nikkor 28mm from the 70s with adapter on Canon 5D Mark II.
10+10+10+1
Hans
|
|
alert moderator
|
|
Sid
Posts: 97
Location: Athens, Greece
Registered: 4 Nov 2009
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 13:01 GMT
|
|
This is a test that i have been planning on doing myself.
thanks for putting the question out there.
looking at the 3 choices..
I cannot see a difference between the 3, they look identical.. (yes i double checked to make sure i hadn't opened the same link 3 times  )
|
|
alert moderator
|
|
erik leeman
Posts: 144
Location: Netherlands
Registered: 24 Aug 2007
|
Re: Do more pictures make for a sharper panorama?
Posted: 31 Aug 2010 at 13:10 GMT
updated: 31 Aug 2010 at 13:12 GMT
|
|
Actually it was using the EF 24mm 2.8 on my then new 5D in 2006 that sparked the idea of using only the high quality areas of each shot for a composite.
That 24mm was (and still is) absolutely brilliant edge to edge on my 20D, but the edges were absolutely horrible on the 5D!
However, those bad image edges proved themselves to be good enough (and convenient) for aligning shots for stitching, so I fine tuned overlap to only get the good bits into my composites.
When I started using the 15mm fisheye I simply adopted the same method for it.
I sometimes shoot four to six extra shots 45 degrees up or down, but only when distribution of interesting detail makes it worthwile.
I recently posted an example in another thread; in a pano from a high cliff you really need lots of detail down there : )
Erik
|
|
alert moderator
|
|
back to threads list |
this thread is closed
Prev
