Getting different results when stitching the ‘same’ source images?

44 views
Skip to first unread message

Philip Giles

unread,
Nov 8, 2022, 1:46:59 PM11/8/22
to PTGui Support

Hello, 

I have created 3 sets of tiffs from a single set of RAW files. There are slight variations in the RAW conversations like saturation, colour temperature, contrast etc, but they are basically 3 sets of the same images. From each set of tiffs, I want to create 3 panoramas with EXACTLY the same stitching output, so I can then take them into Photoshop, overlay one on top of the other and selectively blend them to create the finished panorama. 

For this to work the stitch for each panorama needs to be identical, and for some reason, they’re not. There are small stitching differences in each version of the panorama and I can't work out why.

Using the same .pts file ( I copied it from one set of tiffs to the others) and the same set of images (albeit with small variations in colour, saturation, etc) I presumed the stitching seams would be in exactly the same place. Are there certain settings in PTGui to avoid if I want to create the EXACT same stitch from a set of source images?

Any help will be much appreciated. 

Regards,

Phil

PTGui Support

unread,
Nov 8, 2022, 2:48:37 PM11/8/22
to pt...@googlegroups.com
Hi Philip,

Sounds like you would need to disable 'Find Optimum Seams' in the
Blending side bar in the panorama editor. The seam finding algorithm
will find different optimum seams if you feed it with different images.

Kind regards,

Joost Nieuwenhuijse
www.ptgui.com
> --
> You received this message because you are subscribed to the Google
> Groups "PTGui Support" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ptgui+un...@googlegroups.com
> <mailto:ptgui+un...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com <https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com?utm_medium=email&utm_source=footer>.

John Houghton

unread,
Nov 8, 2022, 3:27:40 PM11/8/22
to PTGui Support
If it's important to retain the use of "Find optimum Seams", could you perhaps process all the images together as HDR by linking the sets of images manually and generate three blend planes?  (No actual HDR blending).  Would the same seam lines then be used?

John

Philip Giles

unread,
Nov 8, 2022, 3:41:21 PM11/8/22
to PTGui Support
Thanks Joost, that's solved the problem. I assumed the 'find optimum seams' would just be the same, even if there were little variations in what are basically the same sets of images. 

Thanks also for your input John, but the stitch is good in this instance so the find optimum seams is not that important with this particular pano. I'll certainly keep it in mind if I come across the problem again as it sounds like an interesting workaround. 

Best regards,

Phil

PTGui Support

unread,
Nov 9, 2022, 2:05:47 AM11/9/22
to pt...@googlegroups.com
Usually there are multiple 'almost optimum' seams and they can be quite
different. A small change in the images can cause one of the other
candidates to become the optimum.

The same problem occurs when stitching time lapses: the seam will move
around when the frames are played in sequence.

Kind regards,

Joost Nieuwenhuijse
www.ptgui.com

> www.ptgui.com <http://www.ptgui.com>
> https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com <https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com> <https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com?utm_medium=email&utm_source=footer <https://groups.google.com/d/msgid/ptgui/87528a1f-d58e-4dfb-9f8e-e1677bfa779en%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
> --
> You received this message because you are subscribed to the Google
> Groups "PTGui Support" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ptgui+un...@googlegroups.com
> <mailto:ptgui+un...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ptgui/e8434cf3-9649-44a4-a3f0-b76434914ff9n%40googlegroups.com <https://groups.google.com/d/msgid/ptgui/e8434cf3-9649-44a4-a3f0-b76434914ff9n%40googlegroups.com?utm_medium=email&utm_source=footer>.
Reply all
Reply to author
Forward
0 new messages