Table of Gigapixel Panaorama  

Share what you've done with Autopano Pro.
Montrez nous vos oeuvres !
User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

Table of Gigapixel Panaorama

by gkaefer » Sat Jul 24, 2010 10:32 pm

under autopano wiki-> Resources
http://www.autopano.net/wiki-en/action/view/Main_Page
I added a table with the gigapixel pano records...

direct link: http://www.autopano.net/wiki-en/action/view/Gigapixel_Records

Liebe Gruesse,
Georg


EDIT: after relaunch of kolor site the entry page is found here: http://www.autopano.net/wiki-en/action/view/Projects_and_Community
Last edited by gkaefer on Fri Apr 01, 2011 3:06 pm, edited 1 time in total.

no avatar
gddxb
Member
 
Posts: 219
Likes: 0 post
Liked in: 0 post
Joined: Tue Apr 27, 2010 4:07 am
Info

by gddxb » Sun Jul 25, 2010 5:08 am

You missed one out - there was a 31GP pano done of the Swiss Alps that came out just a week before my Dubai one.

Wikipedia page here:

http://en.wikipedia.org/wiki/Largest_photographs_in_the_world

(Going by the publication dates, I don't think the Marburg and Vienna images ever actually held the record, so shouldn't be listed here.)
Last edited by gddxb on Sun Jul 25, 2010 5:10 am, edited 1 time in total.

no avatar
julian kalmar
Member
 
Posts: 34
Likes: 0 post
Liked in: 0 post
Joined: Fri May 15, 2009 6:21 pm
Info

by julian kalmar » Sun Jul 25, 2010 7:45 am

gkaefer wrote:under autopano wiki-> Resources
http://www.autopano.net/wiki-en/action/view/Main_Page
I added a table with the gigapixel pano records...

direct link: http://www.autopano.net/wiki-en/action/view/Gigapixel_Records

Liebe Gruesse,
Georg

There is a mistake in your table: Vienna 50 Gigapixel is online since 21st of July.
and Rio is online since 23 of JULY- not June
Last edited by julian kalmar on Sun Jul 25, 2010 7:46 am, edited 1 time in total.

User avatar
fma38
Moderator
 
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Sun Jul 25, 2010 8:11 am

Frédéric

no avatar
gddxb
Member
 
Posts: 219
Likes: 0 post
Liked in: 0 post
Joined: Tue Apr 27, 2010 4:07 am
Info

by gddxb » Sun Jul 25, 2010 9:09 am

julian kalmar wrote:
gkaefer wrote:under autopano wiki-> Resources
http://www.autopano.net/wiki-en/action/view/Main_Page
I added a table with the gigapixel pano records...

direct link: http://www.autopano.net/wiki-en/action/view/Gigapixel_Records

Liebe Gruesse,
Georg

There is a mistake in your table: Vienna 50 Gigapixel is online since 21st of July.
and Rio is online since 23 of JULY- not June

I stand corrected.

You might want to go and edit the Wikipedia page to reflect this :)

User avatar
panoguy
Member
 
Posts: 106
Likes: 1 post
Liked in: 0 post
Joined: Mon Jan 25, 2010 3:55 am
Location: Bavaria
Info

by panoguy » Sun Jul 25, 2010 11:25 am

I see only size matters BUT quality suffers in most of this examples ...
Specially the last 67Gp is totally blurred.
What about letting members vote on picture quality or some other factors.

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Sun Jul 25, 2010 4:00 pm

;-)))
thanks for all replies.

I dont want to play court... so I add all panos I know to the list according the stated Gigapixel achieved.
If listed I add creation and going online date. So if or if not one pano hold the record for 5 seconds or not .... does not finally care...
(and if you compare the dates you can see that vienna pano never hold the record ;-( )

Finally interesting is to compare date and size and get the feeling for the fact thats absolute nonsense to hunt for the record, when within 6 months the size is more than doubling... ;-)

everybody can add items to the list, thats the reason I've choosen the wiki and not the forum itself...
just on top right side create your own account and just press edit ;-)

I'll try to catch up all listed panos and to correct/add the wiki table...

Liebe Gruesse,
Georg

User avatar
taf
Member
 
Posts: 2684
Likes: 1 post
Liked in: 0 post
Joined: Tue Dec 20, 2005 12:13 am
Location: Paaaaaaris !
Info

by taf » Mon Jul 26, 2010 10:02 am

One of the first gigapixel pano created :

http://www.tawbaware.com/maxlyons/gigapixel.htm
November 28, 2003

It's was a real performance. I think it's very important to have this link in the record table...
Look. There's a rhythmic ceremonial ritual coming up !

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Tue Jul 27, 2010 12:23 am

I added some panos (Delft, Bruce Canyon and Pague)
and I also added to special cases:

the terapixel project (Astronomiical pano where images are taken since 50 years... - so biggest possible timelapse pano ;-)
also some additional infos about terapixel project: http://www.physorg.com/news198754211.html

I added "Torrey Pines City Beach" - not really a pano per definition - but its a single film/analog image with 4 gigapixel

and finally I added the index page of Julian Kahlmas Gigapixel panos because its showing some records like worlds first fullsperical gigapixel pano or the biggest (6 gigapixel)pano using focus stacking and bracketing

Liebe Gruesse,
Georg

no avatar
gibie76
Member
 
Posts: 91
Likes: 0 post
Liked in: 0 post
Joined: Mon Dec 17, 2007 7:43 pm
Info

by gibie76 » Tue Jul 27, 2010 9:11 pm

A new gigapixel pano "Budapest 70 gigapixel" : http://70gigapixel.cloudapp.net/

User avatar
panoguy
Member
 
Posts: 106
Likes: 1 post
Liked in: 0 post
Joined: Mon Jan 25, 2010 3:55 am
Location: Bavaria
Info

by panoguy » Tue Jul 27, 2010 9:31 pm

In case you don't have Silverlight installed, you may have a look at the making of panorama here http://www.360cities.net/image/budapest-janos-hegy-elizabeth-lookout

no avatar
gddxb
Member
 
Posts: 219
Likes: 0 post
Liked in: 0 post
Joined: Tue Apr 27, 2010 4:07 am
Info

by gddxb » Wed Jul 28, 2010 4:54 am

gibie76 wrote:A new gigapixel pano "Budapest 70 gigapixel" : http://70gigapixel.cloudapp.net/

35 billion pixels of sky.

What's to stop someone taking that 67GP shot of Rio and photoshopping in another 50GP of sky?

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Fri Jan 28, 2011 12:48 pm

Can someone explain to me what the difference is between the "Total pixels" and the "estimated optical pixels"?

on gigapans site of the Shanghai Pano its is said...:

Total pixels: 272.31 G
Horizontal viewing angle: 175 degree
Vertical viewing angle: 65 degree (+5 degree to -60 degree)
Total shots: 12,000 (150 columns and 80 rows)
Lens: Canon 400mm F5.6 and 2X tele-converter
Camera: Canon 7D (18M pixels per image)
Estimated overlap: 28% (set to 25%)
Average time per image: 2.4s
Estimated optical pixels: 112G pixels

source: http://gigapan.org/gigapans/66626/

Liebe Gruesse,
Georg
Last edited by gkaefer on Fri Jan 28, 2011 12:49 pm, edited 1 time in total.

no avatar
Apapane
Member
 
Posts: 140
Likes: 0 post
Liked in: 0 post
Joined: Mon Sep 01, 2008 2:04 am
Location: Pittsburgh, Pennsylvania, USA
Info

by Apapane » Sat Jan 29, 2011 10:41 am

gkaefer wrote:Can someone explain to me what the difference is between the "Total pixels" and the "estimated optical pixels"?

on gigapans site of the Shanghai Pano its is said...:

Total pixels: 272.31 G
Horizontal viewing angle: 175 degree
Vertical viewing angle: 65 degree (+5 degree to -60 degree)
Total shots: 12,000 (150 columns and 80 rows)
Lens: Canon 400mm F5.6 and 2X tele-converter
Camera: Canon 7D (18M pixels per image)
Estimated overlap: 28% (set to 25%)
Average time per image: 2.4s
Estimated optical pixels: 112G pixels

source: http://gigapan.org/gigapans/66626/

Liebe Gruesse,
Georg

Here's how I calculated "estimated optical pixels" - with the assumption that, in overlapped areas, each frame would contribute either 1/2 or 1/4 of the pixels - see figure.
Full frame 18 Mpixels with 3:2 ratio


Interior frame:
Unique pixels - no overlap
sides of rectangle = 1.5 x 1 = 1.5 Units squared
1.5 units squared / 6 units squared (full frame, 3:2 ratio) = 25%
18 Mpixels x .25 = 4.5 Mpixels contributed by the non-overlapped area. 4.50 Mpixels

Left & Right 2 frame overlap:
.75 wide x 1 high = (.75/6) x 18 = 2.25
divide by 2 to get unique pixels from interior frame = 1.125
2 sides = 2.25 Mpixels 2.25 Mpixels

Top & Bottom 2 frame overlap
1.5 wide x 0.5 high = (.75/6) x 18 = 2.25
Unique pixels = 1.125
2 sides = 2.25 Mpixels 2.25 Mpixels

4 frame overlap
0.75 wide x 0.5 high = 1.125 Mpixels
unique pixels = 1.125/4 = 0.28125
4 corners = 0.28125 = 1.125 Mpixels 1.125 Mpixels
_______________
total unique Mpixels for interior frame 10.125 Mpixels

Using the same procedure:
Edge frames = ~12.656 unique Mpixels
Corner frames = ~14.625 unique Mpixels

For the 12,000 frames (80x150)
Corners = 4 x 14.625 Mpixels 58.5 Mpixels
Edges = 150 + 150 + 80 + 80 - 4 = 456 edges x 12.656 Mpixels 5,899.5 Mpixels
Interior = 11,540 x 10.125 Mpixels 116,883 Mpixels
__________________
total 122,841 Mpixels
122.841 Gpixels

For 28% overlap. the calculated optical pixels = 86.53 Gpixels

The other pixels of the 272 Gpixels come from projection distortion???

Richard




Last edited by Apapane on Sat Jan 29, 2011 10:54 am, edited 1 time in total.

no avatar
Apapane
Member
 
Posts: 140
Likes: 0 post
Liked in: 0 post
Joined: Mon Sep 01, 2008 2:04 am
Location: Pittsburgh, Pennsylvania, USA
Info

by Apapane » Sun Jan 30, 2011 9:11 pm

upload try #25

750 x 500 pixels
1440 Kb uncompressed .tif
280 Kb .jpg

Whoops, I've been trying to upload the CMYK version --- DUH!

RP


Last edited by Apapane on Sun Jan 30, 2011 9:13 pm, edited 1 time in total.

no avatar
GURL
Member
 
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Mon Jan 31, 2011 6:44 pm

How much pixels per degrees are displayed and how large is the field of view would be more meaningful:
- pixels per degrees = visibility of details.
- depending of the subject a large horizontal and/or vertical FOV is meaningful or not...

For Guiness World Records it seems that a print is required (?) http://www.guinnessworldrecords.com/Search/Details/Largest-compound-digital-photograph/104900.htm
Georges

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Tue Feb 01, 2011 1:48 am

what I dont understand is why all this calculations at all.
total pixels - X images multiplied with y pixels per image
but the pixels of the pano dont need to be calculated manually, After rendering with APG the pixels of the final image is displayed... or not?

and in case of using Merlin: the images are not absolutley 100% equal taken. theres a difference some more or less. so even if I say I want 20% overlap its never exactly 20% overlap for each image...

Liebe Grusse,
Georg
Last edited by gkaefer on Tue Feb 01, 2011 1:51 am, edited 1 time in total.

no avatar
Apapane
Member
 
Posts: 140
Likes: 0 post
Liked in: 0 post
Joined: Mon Sep 01, 2008 2:04 am
Location: Pittsburgh, Pennsylvania, USA
Info

by Apapane » Tue Feb 01, 2011 6:46 am

Total estimated pixels, 12,000 frames x 18,000,000 pixels = 216 Gpixels - only if there is NO overlap.

Total Pixels as per pano data = 272 Gpixels - How does this occur?

Estimated optical pixels noted in pano data (surely taking an "average" overlap into account) = 112 Gpixels

Richard's calculated estimated optical pixels with 25% overlap = ~ 123 Gpixels
28% overlap = ~ 86.5 Gpixels

Yes, there is a range of overlap between images. My calculation assumed no variation in overlap - a hypothetical occurrence based on no distortion during image capture and projection.

Overlap DOES make a difference in estimating optical pixels.

Richard

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Tue Feb 01, 2011 9:27 am

Richard,

before starting with the pano you've source images with x megapixel, each image should have the same dimensions so exact the same pixel in total.
this number is finally for the resulting pano not meaningful. clear...

and before doing your pano your formula calculates you an estimated number of resulting pixel. also clear...

but what I dont understand is why having an estimated number of pixel when the pano is finished?
each software can exactly say you the width and height pixel of you image. And Autopano does this also before you start rendering the pano

so why all this effort of listing an estimated number of pixel? each final pano has an exact number of pixels. why not listing this number?

Liebe Gruesse,
Georg
Last edited by gkaefer on Tue Feb 01, 2011 9:27 am, edited 1 time in total.

no avatar
Apapane
Member
 
Posts: 140
Likes: 0 post
Liked in: 0 post
Joined: Mon Sep 01, 2008 2:04 am
Location: Pittsburgh, Pennsylvania, USA
Info

by Apapane » Thu Feb 03, 2011 5:48 am

gkaefer wrote:Richard,

before starting with the pano you've source images with x megapixel, each image should have the same dimensions so exact the same pixel in total.
this number is finally for the resulting pano not meaningful. clear...

and before doing your pano your formula calculates you an estimated number of resulting pixel. also clear...

but what I dont understand is why having an estimated number of pixel when the pano is finished?
each software can exactly say you the width and height pixel of you image. And Autopano does this also before you start rendering the pano

so why all this effort of listing an estimated number of pixel? each final pano has an exact number of pixels. why not listing this number?

Liebe Gruesse,
Georg

Georg,
Your question from earlier in this thread:
Can someone explain to me what the difference is between the "Total pixels" and the "estimated optical pixels"?

In order to understand the difference between "total" and "estimated" pixels, one must first define them. I showed how "estimated optical pixels" may be derived from a given set of images with known overlap.

What is more curious to me is: How does a stitched pano (in this case, a gigapixel panorama) end up with more pixels rendered than the maximum possible "optical" pixels (frames x Mpixels per frame)? To me, it suggests that there is distortion due to the projection of the stitched images (I probably missed a thread about this), or there is some mis-calculation by the stitching program. Any other ideas?

Richard

no avatar
gddxb
Member
 
Posts: 219
Likes: 0 post
Liked in: 0 post
Joined: Tue Apr 27, 2010 4:07 am
Info

by gddxb » Sat Feb 12, 2011 6:27 pm

Hi Richard -

Just going to chip in with a couple of points here.

I'd certainly agree with GURL regarding a starting point for calculating the "true" number of pixels.

If you know the focal length used, sensor size, and pixel density of the sensor, then it's a relatively trivial matter to work out the effective number of pixels for a given field of view of the resulting panorama. You don't have to worry about number of individual photos and overlap percentages at all.

However, I believe (and I don't profess to be an expert on this) this result would be for a spherical projection. If you flatten the panorama, then you're going to introduce interpolated pixels as you move out from the panorama center.

The issue is, how many.

I suspect there's some maths that could be used here, but to be honest, I don't really have the inclination to work it all out.

One thing that I can comment on though, is that if your panorama isn't close to being centered on the center of the image, then AutoPano can wildly inflate the size of it. I've taken panoramas of the Burj Khalifa (similar to this one - http://gigapan.org/gigapans/65156/ ) of just a few hundred images that AutoPano told me would result in a >200GP image.

Obviously in that example, the vast majority of the image is above the center point of the panorama (I'm referring to the center point as where the sensor plane is vertical). In the Shanghai example, I'd guess something like 80% of the image is below the horizontal.

I've not witnessed other programs (PTGui, ICE & one more I can't remember the name of at the moment) doing this. It is something peculiar to AutoPano.

Regards,

Gerald.

Apapane wrote:
gkaefer wrote:Richard,

before starting with the pano you've source images with x megapixel, each image should have the same dimensions so exact the same pixel in total.
this number is finally for the resulting pano not meaningful. clear...

and before doing your pano your formula calculates you an estimated number of resulting pixel. also clear...

but what I dont understand is why having an estimated number of pixel when the pano is finished?
each software can exactly say you the width and height pixel of you image. And Autopano does this also before you start rendering the pano

so why all this effort of listing an estimated number of pixel? each final pano has an exact number of pixels. why not listing this number?

Liebe Gruesse,
Georg

Georg,
Your question from earlier in this thread:
Can someone explain to me what the difference is between the "Total pixels" and the "estimated optical pixels"?

In order to understand the difference between "total" and "estimated" pixels, one must first define them. I showed how "estimated optical pixels" may be derived from a given set of images with known overlap.

What is more curious to me is: How does a stitched pano (in this case, a gigapixel panorama) end up with more pixels rendered than the maximum possible "optical" pixels (frames x Mpixels per frame)? To me, it suggests that there is distortion due to the projection of the stitched images (I probably missed a thread about this), or there is some mis-calculation by the stitching program. Any other ideas?

Richard

no avatar
Guillaume Roumestan
Member
 
Posts: 24
Likes: 0 post
Liked in: 0 post
Joined: Sun Feb 22, 2009 1:45 pm
Info

by Guillaume Roumestan » Thu Jan 05, 2012 9:09 am

Hello,

I would like to present you some of my gigapixels which could be listed in your page : www.gigapixeltour.com

Let me know about it.

Happy new year everybody!!!

User avatar
gkaefer
Member
 
Topic author
Posts: 3549
Likes: 0 post
Liked in: 15 posts
Joined: Tue Jun 09, 2009 1:01 pm
Location: Salzburg
Info

by gkaefer » Thu Jan 05, 2012 11:51 am

Guillaume Roumestan wrote:Hello,

I would like to present you some of my gigapixels which could be listed in your page : www.gigapixeltour.com

Let me know about it.

Happy new year everybody!!!

welcome :cool:
its a wiki, so take an account (login top rigth and than choose an id, fill in email and choose a password - than you can edit the wiki page and add your panos. As I can see the Cannes 65gigPano is listed since July...
... of course I can do it too..:D

I added your site in the listing on the bottom...

Liebe Gruesse,
Georg
Last edited by gkaefer on Thu Jan 05, 2012 12:01 pm, edited 1 time in total.


Who is online

Users browsing this forum: No registered users and 2 guests