|
admin
|
 |
« on: July 16, 2009, 08:38:20 PM » |
|
http://www.ddisoftware.com/qimagev2009.267 07/16/09 Priority: Low v2009.267 improves raw support for the Panasonic GH1 by automatically correcting barrel distortion at wide angles when using the 14-140 kit lens. v2009.267 also corrects a bug that was causing the bottom row of thumbnails to not update properly after a thumbnail rebuild under certain conditions. Mike
|
|
|
|
|
Logged
|
|
|
|
|
Terry-M
|
|
« Reply #1 on: July 16, 2009, 09:03:03 PM » |
|
Mike, automatically correcting barrel distortion at wide angles That sounds very clever to me  What about other cameras and lenses, or do you need a database of lens characteristics like PTLens? I have a Canon 17-85 IS lens that has poor aberration performance at wide angle both with barrel distortion and chromatic aberration. Auto correction of both would be very nice.  Terry. |
|
|
|
|
Logged
|
|
|
|
|
Fred A
|
|
« Reply #2 on: July 16, 2009, 09:48:58 PM » |
|
|
|
|
|
|
Logged
|
|
|
|
|
admin
|
|
« Reply #3 on: July 16, 2009, 10:19:36 PM » |
|
Mike, automatically correcting barrel distortion at wide angles That sounds very clever to me What about other cameras and lenses, or do you need a database of lens characteristics like PTLens? I have a Canon 17-85 IS lens that has poor aberration performance at wide angle both with barrel distortion and chromatic aberration. Auto correction of both would be very nice. Terry. Ah you see into future oh great one!  That's the next step. Now that I've written my own code to correct lens distortions, you'll see it in the batch filter in the future and I may even make plugins for Qimage that will allow automatic correction based on lens models. This change was to support new technology by Panasonic that was required to properly decode G1 and GH1 raw photos. Panasonic has done something interesting in their latest lenses/software. They've made their lenses simpler and the software more complex to compensate. Instead of making expensive lenses that correct distortion optically, they can make lenses with less elements and make them cheaper. Then they remove the barrel distortion at wide angle via software, so software removal of lens distortion is a requirement for raw decoding on their new micro four thirds cameras! Unfortunately dcraw doesn't perform this operation like other raw converters like the Panasonic converter, SilkyPix, and Adobe ACR so I had to write my own code to fix the flaw in dcraw. This whole concept of shifting some of the lens optics to software is interesting enough that it will be next months article here. Stay tuned. Mike |
|
|
|
|
Logged
|
|
|
|
|
Terry-M
|
|
« Reply #4 on: July 17, 2009, 10:38:44 AM » |
|
That's the next step. Now that I've written my own code to correct lens distortions, you'll see it in the batch filter in the future That would be enough for me. I'm already trying to imagine how it would be implemented, something like the Straighten Tool on a distorted element, and click on a fringe in the enlarged pre-view. That's just guessing and I don't expect an answer now I'll read your next article with interest  Terry. |
|
|
|
|
Logged
|
|
|
|
tomc
Jr. Member
Posts: 52
|
|
« Reply #5 on: July 17, 2009, 04:21:35 PM » |
|
Now that I've written my own code to correct lens distortions, you'll see it in the batch filter in the future and I may even make plugins for Qimage that will allow automatic correction based on lens models. Interesting. You could develop a new product to go along with this feature that would supply a target and software to create the lens model based on images of the target. A geometry version of Profile Prism. Tom |
|
|
|
|
Logged
|
|
|
|
|
admin
|
|
« Reply #6 on: July 17, 2009, 08:00:38 PM » |
|
Now that I've written my own code to correct lens distortions, you'll see it in the batch filter in the future and I may even make plugins for Qimage that will allow automatic correction based on lens models. Interesting. You could develop a new product to go along with this feature that would supply a target and software to create the lens model based on images of the target. A geometry version of Profile Prism. Tom That's one of the first things I thought about while working on this project. It wouldn't be that difficult to do but the process itself is tedious. There are also a number of issues such as the need to take multiple pictures of the target at different focal lengths while still getting the target to cover the entire frame: that can be difficult at the wide angle end in particular. All of these can be overcome of course, with the right amount of time and development. It's a possibility. Mike |
|
|
|
|
Logged
|
|
|
|
|
Fred A
|
|
« Reply #7 on: July 19, 2009, 11:26:32 AM » |
|
So I decided to have a look at the barrel distortion on my Canon "L" 17-40mm lens. (750.00). Not that bad, but it is there, and undeniably so. The write-ups on that lens indicated that there was plenty. Distortion ranges from moderately strong barrel distortion at 17mm to mild pincushion at 40mm with 24mm being the approximate point of minimal distortion. As usual, closer distances resulted in stronger distortion and significantly less distortion will be noticeable on 1.6x FOVCF bodies. You will likely notice the barrel distortion in the full-frame 17mm ISO 12233 resolution chart sample crops. I guess I look forward to the distortion correction feature too. Fred |
|
|
|
|
Logged
|
|
|
|
|
Terry-M
|
|
« Reply #8 on: July 19, 2009, 02:04:48 PM » |
|
barrel distortion on my Canon "L" 17-40mm lens. (750.00). Not that bad, Want to see it really bad, look below from a Canon 17-85 f4-f5.6 at 18mm. Chromatic aberration is not good at the wide end either. Things do improve significantly at 20mm and above. The Right Hand image was corrected with PTLens using their database of lens characteristics. Terry |
|
|
|
|
Logged
|
|
|
|
|
admin
|
|
« Reply #9 on: July 19, 2009, 04:31:16 PM » |
|
Want to see it really bad, look below from a Canon 17-85 f4-f5.6 at 18mm.
Look, a bay window! Oh wait... NOT! I'm looking into making a lens database in Qimage but for sure, you'll be able to do these corrections manually in Qimage soon. The database may come later. Mike |
|
|
|
« Last Edit: July 19, 2009, 04:33:09 PM by Mike Chaney »
|
Logged
|
|
|
|
|
Terry-M
|
|
« Reply #10 on: July 19, 2009, 04:37:38 PM » |
|
you'll be able to do these corrections manually in Qimage soon. That'll suit me; I can always make a little library of filters to save time, a personal database. That's assuming it'll be in the batch Filter. Terry |
|
|
|
|
Logged
|
|
|
|
Peter_Corser
Newbie
Posts: 15
|
|
« Reply #11 on: August 01, 2009, 04:39:12 PM » |
|
Mike
Just caught up with this thread.
Tom Niemann's PTLens basically does this - the problem of doing all the testing and targets is basically covered by specifying the requirements for test files (on the web site) and requesting users to generate and transfer them to him. Would this approach help you?
PTLens is pretty good at correcting the distortions (not quite as good as DxOptic Pro can produce) but having this facility in QI would be really useful.
Peter
|
|
|
|
|
Logged
|
|
|
|
|
Qimage registration expired? New lifetime licenses are only $59.99!
