Mention of used camera and lens in panorama EXIF data

81 views
Skip to first unread message

Karsten Gieselmann

unread,
Aug 24, 2023, 9:55:11 AM8/24/23
to PTGui Support
As for EXIF data of the panorama output file I wonder why PTGUI maintains its own interpretation of the camera name (tag "Model"), put in square brackets, eg "[OM-1]" instead of "OM-1".

Also, why is the "LensModel" tag empty?

I'm using exiftool to fix these issues for every file generated by PTGUI but it would be nice to have this straight in output.

P.vR

unread,
Aug 25, 2023, 10:59:26 AM8/25/23
to PTGui Support
I for one am happy the camera model is not used for the panorama file.
Now I can easily filter the panoramas out based on the camera model name.

Please keep in mind that the original camera isn't the direct source of the panorama, since PTGui is used to stitch the various input files together.
The same applies to the LensModel. My fisheye doesn't produce a 360 degree picture. So, it wouldn't be correct to suggest it did.

On the other hand, I understand you want to save the reference to the camera/lens that was used for the input files.
It would probably be better to add that info in different tags, though.
Something like: "PanoInputLensModel" and "PanoInputCameraModel", or whatever makes sense here.
 
Just my 2 cents.

Peter

PTGui Support

unread,
Aug 26, 2023, 5:20:01 AM8/26/23
to pt...@googlegroups.com
Indeed, this was done by design. The idea is that the metadata should
match the image content. If you stitch two 28mm photos, the resulting
image has a wider field of view and should no longer have 28mm in the
metadata.

Since PTGui itself relies on the metadata being correct when stitching
images, it would just be wrong to add incorrect metadata. The camera
model is used to find out the sensor size. With a stitched image this is
no longer correct, so PTGui adds the brackets to indicate that it is no
longer truly an OM-1 image.

I know there are other ways to look at this, and I've learned that
metadata is a sensitive subject for some. Perhaps adding the original
metadata to a comment field would be useful, but then the question
becomes where to stop. If you need a true audit trail for your panoramic
image, the only true solution is to keep the original images and the
project file anyway.

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/34c70a47-1922-4957-9b12-0fa2105aefcan%40googlegroups.com <https://groups.google.com/d/msgid/ptgui/34c70a47-1922-4957-9b12-0fa2105aefcan%40googlegroups.com?utm_medium=email&utm_source=footer>.

Karsten Gieselmann

unread,
Aug 26, 2023, 9:03:12 AM8/26/23
to PTGui Support
Thank you for your explanation, Joost.

While I understand your motivation how to handle metadata, PTGui is the only image processing software I know that does it like that, as opposed to HDR, focus stacking, deep sky stacking, panorama (Adobe) and other programs that I'm using in my workflow.

I'm using this information eg. in Lightroom to distinguish images according to used cameras and lenses.
So it seems I need to continue with my exiftool script.

Greetings to NL
Karsten Gieselmann

Mike Cowlishaw

unread,
Aug 29, 2023, 1:14:54 PM8/29/23
to PTGui Support
For information, PanGazer also does this.   If you 'snapshot' a view (i.e., crop an image) the lens information is adjusted to match the image saved.   This means that when viewing the cropped image you will continue to get correct (that is, the same) pan and tilt angles for a given feature in the image.  If the lens data were not updated to match the crop, the pan and tilt coordinates of a given feature would change (be wrong).

I based this feature on other software I've used, so I believe this is quite common in image-processing software.

Mike
Reply all
Reply to author
Forward
0 new messages