Forum/thread	Message
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 9 Oct 2008 at 12:37 GMT updated: 9 Oct 2008 at 12:55 GMT go to thread Hello Hans, Nice to read that you did your homework so well But I guess your tests are a little bit off topic. It is not about the specific sharpness, or CA, or fisheye mapping of lens A, B or C. All lenses you mentioned are very close and comparable so lets not argue about if a specific lens is sharper or not, has more CA or not, and if you are sure that your Tokina 10-17 is mapping close to the edge exactly the same as your Sigma, then so be it. You can be wrong and so can I and that doesn't matter at all. The only thing what really matters is if a lens, camera, panohead, etc. is capable for doing the task you, we, I want to use it for. I use the 4 shot 60 degree slanted pano system with the Sigma 8mm and a EOS 450 D for snapshot panos and I think the results are fine and because you didn't argue this I guess you agree. Another person pointed to the fact that when shooting full spherical pole panoramas with this specific lens and camera combination he has no choice and so he must use a slanted system. So as it turns out there are more motives for using a slanted pano system then just for making snapshot panos as I do. Hans when you take a good look at the Old Harbor of Ouddorp demo again I am sure you will see some CA in the panos, I did But the little amount of CA that is visible doesn't bother me at all. And in case the CA in a specific pano is really horrible I can always process the RAW's, remove the CA to get a better pano. Hans, happy slanted shooting greetings Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 9 Oct 2008 at 5:53 GMT updated: 9 Oct 2008 at 6:00 GMT go to thread @ Bigwade, Thanks for posting the nadir and zenith images. It is easy to see why a little tilt of the Sigma8 f.3.5 is needed to fill the hole in zenith. And you are right about the offset of the image projection on the sensor. When shooting non slanted and without a tilt then also on my system the black whole is a little larger in nadir then in zenith, not much but clearly visible. Perhaps this has something to do with the free play of 0.2 mm of the lens mount of the Sigma8 in the EOS body. @ Hans, First thanks for the detail images of the comparison of slanted and not slanted panos. The reason why you are making such strong statements about the lesser quality of slanted systems is clear. Besides the Sigma 8mm f3.5 and EOS 450D I also have a shaved Tokina 10-17mm and a EOS 5D. When comparing both lenses it is clear to me that the Tokina is not as sharp as the Sigma8 at the outer edge of the circle, the difference is very, very small but visible. And the distance setting of the Tokina is done precise The Tokina also has a little more CA at the outer edge of the circle, not much but also visible. And then there is the issue that the Sigma 8 is a true fisheye lens from the center of the image to the outer circle while the Tokina 10-17, and so is the Nikon 10.5, is not. The last two lenses behave like a normal fish lens up to approx. 175 degree and from 175 to the outer circle the mapping is different and that is why it is never a good idea to put CP's to close to the edge. But I am sure you already know this BTW, Michel Thoby has published some very good articles about the different fish eye lens behavior of the Nikon 10.5 and Tokina 10-17 on his website. So for me it is clear that you can not for 100% compare the Sigma 8mm f3.5 and EOS 450D with the Tokina 10-17mm @ 12.5mm and EOS 5D. All three issues mentioned above, the sharpness, the CA, and the different mapping at the outer circle are making the Tokina 10-17 less suited for a slanted pano system. That said, when the source images are pre processed and CA and sharpness are both corrected as much as possible then even with the Tokina 10-17 it is possible to make very nice snapshot panos on a slanted pano system. A small sidestep. When reading the text above one could think that the Sigma8 is a "better" lens compared to the Tokina 10-17 but this is not true. The flair control of the Tokina is far superior, due to the special coating the lens surface is much easier to clean and the zoom is a real bonus. So both lenses has there strong and weak points. All the best, Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 8 Oct 2008 at 13:34 GMT go to thread Hello Andrew, The cheap rotator is made by Feisol, model PB 70. It is a model without clickstops and without roller bearings but nevertheless a very good product. www.feisol.com/english/enconnect-n.htm BTW, on the side of Feisol the product is named "Panning base" The tripod on the overview is from the same manufacturer. Best, Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 8 Oct 2008 at 13:28 GMT go to thread Hello Hans, I know that there are many issues involved with the use of a slanted pano system and that the quality may decrease a little (depends on the scene if this is visible or not) but nevertheless the slanted system is very suited for snapshot panos. If you have looked and inspected all 5 panos of the Old Harbor of Ouddorp then I think you will agree that these panos are not so bad at all despite that they are shot slanted The slanted system is parallel invented by Luca Vascon, one of the most skilled panographers, and by Michel Thoby, also a very experienced person so I guess that justify the use of it And keep in mind that for this demo none of the images is pre or post processed, that even the logo is fully batch patched in the pano and all HTML files are batch generated in Pano2VR. And with the use of 4 instead of 3 images I keep on the safe side of CA troubles.... Hans thanks for your comment, it is good that people will be aware of the fact that a slanted pano system is not the holy grale for all panos. Best, Wim alert moderator
Forum: Panoguide website feedback Thread: Open link in new window.	Re: Open link in new window. Posted: 8 Oct 2008 at 12:00 GMT go to thread Hi folks, Just like many other forum members I like to see that an embedded URL in a posting opens a new window for the content of the link. That way when a visitor clicks on the link and later close the new window the visitor is back at the posting. I am pretty sure it is possible to set the forum software to act like this by default... Anyway, with a piece of HTML code < target="_blank" > a link can be set to open always a new window. I don't have any experience with the code options of this forum and I don't think it is wise to experiment with it so perhaps the forum master can explain me how to use the html code < target="_blank" > in a posting. Thanks a lot for your help and all the best, Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 8 Oct 2008 at 11:26 GMT updated: 8 Oct 2008 at 11:27 GMT go to thread Hi Folks, Good to read that more people are using a slanted system As you perhaps know the 60 degree slanted panohead was introduced on the market by AGNOS some time ago and it was based on a concept of Luca Vascon. I have the original AGNOS CP and it worked fine with the older Sigma 8mm f4 on my EOS 350D but not on my EOS 450D with Sigma f 3.5. To use a Sigma 8mm f3.5 on a 1.6 cropped Canon camera I found out that an angle of 62 degree and a lens tilt of 2.5 degree gives a a little better overlap for shooting just 3 images. I like to use my gear as flexible as possible and that's why most of my stuff is transformed into modular pieces. Here is small a video of the 60 degree pano system. It shows the system in use with a very expensive rotator, with a very cheap one, and even without a rotator: www.dmmdh.nl/zoomify/60degree_slanted_system/movi... As you can see it doesn't matter what kind of rotator is used. Of course for batch stitching I definitely need a precise rotator and even then it is not necessary to use the Seitz VR Drive, as long as the rotator is precise and can be provided with a Giottos 200H quickmount system it suits me fine. Here is a zoomify image that shows most parts of the 60 degree slanted pano system: www.dmmdh.nl/zoomify/60degree_slanted_system/over... I guess by now you will have a pretty good idea of the stuff I use and the flexibility of it. Best, Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Re: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 7 Oct 2008 at 20:14 GMT updated: 7 Oct 2008 at 20:19 GMT go to thread Hello Andrew, Indeed I use a custom made bracket for the slanted technique (with 4, not 40 shots The reason for not using f.e. a lens ring is that it is not so easy to switch between lenses, cameras and panoheads with a fixed lensring and I don't think there is a lensring on the market yet that is capable of providing a lens tilt and at the same time is easy to mount in seconds.... There are several advantages when using a slanted system, and some dis-advantages that needs a work around. One advantage of using a slanted system for short fish eye lenses is that the overlap in the center part of the image is huge compared to an image that is shot in a portrait position. Due to this larger overlap placing CP's automatically in PTGui is very succesfull and with a good spread. Usually PTGui will very rarely place CP's close to nadir or zenith in the images of short fisheye lenses but it has a very good spread in the other parts of the image and a slanted system provides more of this "better to handle" part for PTGui. A very big advantage is the fact that the FOV of the source image is fully used so even without zenith or nadir shots the image is by default more then 180 degree. Only the size of the panohead is the restricting factor of the footprint and if the rotator is slim then so is the footprint. Another advantage is that a fixed bracket can be smaller then a panohead and when the camera is mounted to it it the combination can be easily transported in a normal sized photo equipment bag. So a slanted system has more then one advantage. That said there are also some dis-advantages. At the seem parts of the source images with different NPP properties have to blend so theoretically a stitch will give more parallax problems. Luckily in real life situations this is seldom a real problem. I have experience with slanted systems for many years and I found that the only serious problem is when I have to shoot in crowded and dynamic scenes. In such scenes every seem in all pano systems can be a potential "problem zone". Because the length of the seem of a slanted system is longer compared to a "normal" system the change that moving objects at the seems will cause problems is larger. I found out the hard way that I really had to fine tune my shooting method for dynamic scenes to avoid extra problems and because I prefer to shoot without the need of a lot of Photoshop work in those scenes I prefer to shoot with a panohead instead of a slanted system. And of course because the bracket is "fixed" making a topshot in case you need to catch all the details of the ceiling of that fine painted cathedral is not possible. End if you need a nadirshot with a long exposure time in the same cathedral and you want to use an easy method (see below) it is also is not possible nodalninja.com/forum/index.php?topic=353.0 For this reason I will never use a slanted bracket for a commercial panoshoot but for some other purposes, f.e. pano snapshots with a good image quality, it serves me well. Best, Wim alert moderator
Forum: Q & A Thread: Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens	Demo of a 2 and 4 shot pano system with a Sigma 8 mm fish eye lens Posted: 7 Oct 2008 at 15:32 GMT updated: 7 Oct 2008 at 15:36 GMT go to thread Hello Forum, A few weeks ago I reported about the possibilities of the Sigma 8 mm lens for making panos with a 2 shot pano system: www.panoguide.com/forums/tipsntricks/5296/ Since I have read a lot on this forum about the options for a 2, 3 or 4 shot system with fish eye lenses. Yesterday I shot a few demo panos at the old harbor of Ouddorp with my 4 shot 60 degree slanted bracket system (EOS 450D+Sigma 8mm): www.dmmdh.nl/panos/ouddorp_harbor_06102008/image_... And today I shot some panos on the market in my home town with the 2 shot pano system (EOS 5D+Sigma 8mm): www.dmmdh.nl/panos/ouddorp_market_07102008/image_... About the 4 shot 60 degree slanted bracket system: The size of the equirectangular (8800x4400) px is sufficient for a good full screen pano with a fine quality. Although it is also possible to get the same size with just 3 instead of 4 images around I prefer to use 4 images for a pano. With 3 images around the fringing and CA of the lens is much more visible in the pano then with 4 around and so with 3 around I must use RAW's (I always shoot RAW+JPEG) to correct the CA and fringing. With 4 around I can use the JPEG's of the camera. About the 2 shot pano system: In my first posting I said that don't think the Sigma 8mm lens with it's FOV of 182.6+ degree is suited for normal panos because there are many issues with the shooting and the stitching. After having some more experience with the 2 shot pano system since I can tell that the pano shoot and the stitching isn't so hard to do at all. With just a few CP's the batch template of PTGui can be fine tuned for the lens shift parameters d and e. Fine tuning of the batch template is needed each time the lens is changed because the mount of the Sigma 8 mm lens has a free play in the Canon body of approx. 0.2 mm, the resulting image shift on the sensor depends on the sensor resolution but is to much to ignore. I tried it but I found out that is impossible to mount the lens each time in the same position on the body. When you look at the seem between the images it is obvious that there is a lot of annoying blue fringing and CA. All images for the demo are processed from RAW to correct as much as possible CA and fringing but still there is this annoying blue fringing that I can't get rid off. Besides the fringing there is also the issue of loss of sharpness at the image seem and the issue of the low resolution of the pano. My conclusion is that although the 2 shot pano system is not so hard to use at all the overall quality of the panos is not good enough for me. The 4 shot 60 degree slanted bracket system is a different story. The Sigma 8 lens is tilted +2.5 degree and together with the FOV of 182.6+ degree this is sufficient to get a full 360x180 pano without any black whole and without to much blue fringing and CA. The small logo covers the footprint of < 23 degree so with just 4 shots around I can very fast and very easy shoot and batch process fine panos. Untill there is an affordable 190+ degree fish eye lens on the market for use on a 20+ MPX full size sensor camera there is no need for me to use a 2 shot pano sytem for normal panos and I am pretty sure that untill then the 4 shot 60 degree slanted bracket system will do fine whenever I want to make a fast and easy to process pano. Best, Wim alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	Re: 2 shot pano system with a Sigma 8mm f 3.5 Posted: 17 Sep 2008 at 8:51 GMT updated: 17 Sep 2008 at 8:51 GMT go to thread enbilaman said: FYI: the FOV of the (CP P5100 +FC-E8) combo is much larger than 185° with the Nikon UE-R20 original adapter. Anyhow, I have shortened the Agnos "clone adapter" and the resulting FOV is larger than 192° michel.thoby.free.fr/DuoPix/DuoPix_solution.html Hello Michel, Thanks for posting the link to your article. I was thinking about the same solution but decided not to built one when I heard about the irritant long shutter delay of the Nikon camera and the lack of a remote control port. And besided all the issues you mentioned there is also the huge(*) amount of flare of the FC-E8 lens. (*) compared to the Tokina 10-17. Best, Wim alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	Re: 2 shot pano system with a Sigma 8mm f 3.5 Posted: 13 Sep 2008 at 19:28 GMT updated: 13 Sep 2008 at 19:31 GMT go to thread Hello Doug, To make a good comparison possible I converted your equirectangular to a flash pano: www.dmmdh.nl/panos/2shot_panos/5D+Sigma8_C6.html The most visible difference between the output of PTGui and IPIX Image Studio is the oval shape of the footprint of the IPIX software output. It is obvious that the lens correction params of the IPIX software are not OK but you already explained that you need images with straight lines to make better corrections. I used 4 images around for a lens calibration and I used the lens params for stitching and that is why the footprint in my output is OK. Thanks for your efforts and the explanation of the Nikon Coolpix system. All the best, Wim alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	Re: 2 shot pano system with a Sigma 8mm f 3.5 Posted: 13 Sep 2008 at 10:42 GMT go to thread Hello Doug, Good to read that you have success with iPIX Interactive Studio. You mentioned stitching errors, I am curious about the differences in output, can you post a link to your output ? BTW, when you compare the overall output quality of the 2shot system with your Nikon set do you think there is a noticable difference in quality ? Best, Wim. alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	Re: 2 shot pano system with a Sigma 8mm f 3.5 Posted: 12 Sep 2008 at 8:17 GMT go to thread Hello Doug, Here is a set of 2 full circle fish eye images in full resolution and high quality (18 MB): www.dmmdh.nl/panos/2shot_panos/copyright_dmmdh_pr... You and others can use this set for educational purpose only, but please, no publishing on a website, no posting in fora, no hot linking, etc. etc. The set is slightly different from the previous posted ones, since I adjusted the 2 shot system bracket and in this set the NP is very, very precise and the batch template has CP errors less then 0,5px around the crop circle. The images are shot on a Feisol carbon tripod at 2 m heigth. The footprint is extremely small due to the fact that the legs are only 50 cm apart. For stabilizing the system a small powerlift dumbbell with a weight of 5 kg is hanging at the hook of the center column of the tripod. You can compare your own results with my batch stitched output here: www.dmmdh.nl/panos/2shot_panos/2shot_pano_system.... Let me know if you have any success with iPIX Interactive Studio. Best, Wim alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	Re: 2 shot pano system with a Sigma 8mm f 3.5 Posted: 11 Sep 2008 at 14:35 GMT updated: 11 Sep 2008 at 14:44 GMT go to thread Hello Marc, I use the VR Drive in almost any situation, for normal panos, for sub GigaPan resolution panos for timelapse panos, on tripods, on a high pole or on a table support. I use the device in combination with self made brackets and customized panoheads of manufacturers, f.e. the Nodal Ninja NN5 panohead. Making a timelapse pano is in fact making a series of panos from the same position over a period of time. For ease of use I customized a Canon cable remote timer to fit on the VR Drive. The number of panos I can shoot for a timelapse is limited to 99 by the timer. The time interval can be programmed from a few second up to many hours, depending on the scene I usually shoot a pano between 1 and 20 minutes. Each pano is stitched in PTGui with a individual template that is created from a single batch template. First I make 2 or 3 individual templates from 2 or 3 random chosen set of images, then I combine the CP's of the templates into 1 template to get an optimized combined batch template. By combining the CP's of multiple panos I get a more useful batch template, little errors in the params are flatten out. With a self written HyperCard script (runs in SheepShaver) I use the batch template to create a template for every individual pano and with the Batch stitcher of PTGui I output all panos as equirectangulars. The equirectangulars are transformed in Pano2VR into the 3x2 Special format for spherical video. The 3x2 Special format images are then imported in QuickTime Pro and exported as H.264 encoded MP4 movie. Pano2VR also creates the necessary HTML file and adds some other files to make it possible to view the timelapse panorama in any webbrowser with flash version 9.0.64 or higher. Here you can see a timelapse of the sunrise of the Gornergrat I shot last year at an excursion with the PanoTools Meeting group in Swiss, for this one I didn't used the VR Drive, no timer, no advanced rotator and all was done bij hand and without a watch ! www.dmmdh.nl/panos/timelapse/flash_timelapse-mp4.... BTW, loading and processing can take some time and all you see while waiting is a white screen, this is normal and after a minute or so you get a rotating pano. You can stop the rotation and the time with the buttons in the panorama. Perhaps you see some familiar faces in the Gornergrat scene. More faces can be seen in a timelapse of the Welcome Dinner of the last PanoTools Meeting in Prague. The link to that timelapse is posted in the PanoTools Meeting Group forum and available for members. This timelapse pano is made and created with the VR Drive and Canon timer as described above. Making a timelapse is not something you do in a spare minute, it takes a lot of time for preparation and processing and because the number of panos is to large for retouching in post processing you have zero tolerance so many errors are visible. That said, what counts in the end is the viewing experience and for that purpose a timelapse is a wonderful thing. Happy timelapse shooting, Wim alert moderator
Forum: Tips & Tricks Thread: 2 shot pano system with a Sigma 8mm f 3.5	2 shot pano system with a Sigma 8mm f 3.5 Posted: 11 Sep 2008 at 6:48 GMT go to thread Hello Forum, When I recently tested a custom made pano bracket for shooting 4 images with a Sigma8 f3.5 lens on a Canon 450D camera in a 60 degree slanted position (the AGNO way) I was happy to see that the results are fine. Finally the black wholes in nadir and zenith are gone. So the need for an extra nadir or zenith shot when shooting slanted are gone (unless I want to patch the footprint of course). The above is only an introduction for this topic, when testing the slanted system I also noticed that the angle of the crop circle of the Sigma8 f3.5 as measured in PTGui is a constant 182.6+ degree so I decided to explore the possibilities of a 2 shot pano system with the Sigma 8 f3.5 lens and a full size sensor camera. I managed to get decent results (*) with a Canon 5D and the Sigma 5 f3.5 lens mounted on a custom bracket on top of my Seitz VR drive. (*) The definition "decent" is a personal one. You can see the results here: www.dmmdh.nl/panos/2shot_panos/image_001.html An image of the 2 shot system: www.dmmdh.nl/panos/2shot_panos/2shot_system.html There are lots of issues that makes it hard to use the 2 shot system: *) I have to shoot under controlled conditions, it is not possible to retouch moving objects etc. on the seem, *) I always have to use a very precise panohead and tripod because the lap is only 2.6 degree !! *) Only in a controlled environment it is possible to set CP's around the image circle to make a stitch template so it is really necessary to stitch images that are shot in the real world in batch mode. Lens calibration has to be done on another system with 3 or even better 4 images. *) Because the size of the equirectangular is limited the output isn't really suited for viewing full screen on a 24" monitor. The 2 shot pano system is more useful when making timelapse panos. The main advantage is that the system reduces the number of seems from 3 to 2 so that means less destructed heads and limbs of people in the scene because in that kind of panos no retouching is done due to the large number of panos that are involved. For timelapse panos based on flash I am limited to output a cube size of max 960 px so then the limited output of the 2 shot system isn't an issue. I hope you like the examples and that they inspire you to do some experiments with the Sigma 8mm yourself. Best, Wim alert moderator
Forum: Q & A Thread: Stitch Speed Test on different machines	Re: Stitch Speed Test on different machines Posted: 24 Aug 2008 at 7:55 GMT go to thread Hello Henri, Nice test to see the varies speed of different machines. Om my 2,5 yr old MBP, first generation, 2 GHz+2GB paired RAM+ original slow 100GB 5400 rpm disk it took 70 seconds to get the final output with the latest PTGui 8b8. Enblend 3 (from the combined Enfuse+Enblend distribution) took 40 seconds so I guess thats pretty good for this oldy. Another oldy performed much less, my iMacG5 2 GB+2GB paired RAM + 500GB HD 7200 rpm needed 175 seconds to finish the job with the same software. BTW, anti virus software was disabled. Best, Wim alert moderator