I created an image in Manifold 9 and exported it as a georeferenced tiff file. The projection is Washington State Plain North. Manifold 9 defaults to Lambert Conformal Conic as the type. When I import it into Manifold 8, the projection is identified as Lambert Conformal Conic and the registration is way off. I've tried to modify the imported image to replicate how the image is projected in Manifold 9, but I still end up with a misregistered image. That, and the image has artifacts to the right in the imported image that are not present in the exported image. If I create an image in Manifold 8 for import to Manifold 9, it comes in correctly registered. Is there a step I am missing when exporting georeferenced images from Manifold 9?

Attachments:
Image created in M9 (2023-03-10).jpg
Image Imported into M8 (2023-03-10).jpg

Both attached images are Release 9. Also, you need to provide way more information, like which builds you're using, exactly how you exported, the info on all details of the projection you're using in 9, how it comes in on 8, etc.

Thank you for catching that. I had inadvertently copied over the Manifold 8 image. Here are the comparison images again. I have also included the geotiff file created using Manifold 9 and imported into Manifold 8.

I haven't tested this hypothesis yet, but would it help if the image were projected as Lat.-Long. before exporting as a geotiff?

Attachments:
Image created in M9 (2023-03-10).jpg
Image Imported into M8 (2023-03-10).jpg
Test Geotiff.tiff

I haven't tested this hypothesis yet, but would it help if the image were projected as Lat.-Long. before exporting as a geotiff?

GeoTIFF can handle all sorts of different coordinate systems. It doesn't have to be in lat/lon.

Respectfully, you're not taking the right approach if you want help. The most likely problem is some error in workflow, although it might be something else. To find a possible error, anybody who wants to "look over your shoulder" needs details on what you're doing. Jumping around considering various hypotheses without describing the basics doesn't help.

There are three very important reasons to describe the basic, necessary info in complete detail: First, the process of simply stating everything that's going on in detail very often reveals an error in workflow, something overlooked, something that's been assumed to be a particular way when it isn't, etc. Second, you can't narrow possible issues down without knowing what's going on. Third, on rare occasions bugs do pop up, but those get exponentially harder to find if the basic info isn't there.

So start by saying what version/build of 9 you're using, how the image was created, exactly where it is supposed to be, and how it was exported. Verify there's nothing wrong with basic coordinate system management by doing things like displaying it in a map with "known good" layers like Bing streets or Google satellite, exporting it from 9 and then re-importing it into 9 and verifying it does come back in where it is supposed to be.

Once you have all that verified, if the issue is importing it into 8, say exactly how you're doing that, what version/build of 8 you're using, whether the various other files like .tifw and .prj files are in the folder from which you import, etc.

With all the details in hand, odds are great the forum can help out, and if not, those details will be essential for tech support to look into the matter to help you out, or if a new bug is in play to find and fix that bug rapidly.

Fair enough. I am using Manifold 9.0.169.4. and Manifold 8.0.34.0. My default maps for this project are projected as Washington State Plane in North using NAD83 as the datum. The initial Manifold 9 image showed colorized LiDAR terrain data. I turned that layer off and turned on Bing Satellite and created a new image. The point locations show that the LiDAR data corresponds well with Bing Satellite. I used the default settings for Save as Image. As suggested, I exported the image using both File/Export and right-click-Export,) and re-imported it to the same Manifold 9 map. The re-imported image falls where it should. Conversely, a similar image made using the same source data in Manifold 8 also imports into Manifold 9 where it should. The folder where I exported the images from Manifold 9 does .prj and .tiffw files.

I import images to Manifold 8 using the File/Import/Image command, which I have always used to import georeferenced images. Instead of importing at Latitude 47.424 N and Longitude -122.237 W, in comes in at Latitude 47.0442 N and Longitude -123.03326 with a projection of Lambert Conformal Conic. Re-assigning the projection to be Washington State Plane North makes things worse. Copying the coordinate information from Manifold 9 to Manifold 8 by hand still does not fix the issue.

One thing I did just notice is that the Manifold 9 base map is projected as State Plane Washington North. The image created by Save as Image is projected as State Plane Washington Northn (sic). I have no idea where this projection comes from since I could find no other instances of this projection when reviewing all available coordinate systems in Manifold 9.

Attachments:
Image created in M9 (2023-03-10).jpg
Image Imported into M8 (2023-03-10).jpg
M 9 Setup for Image Creation (2023-03-13).jpg
Manifold 8 upload to 9 (2023-03-13).jpg

It looks like you're doing everything right but that there is a problem with how 8 imports the tiff.

Is the Test Geotiff.tiff file you posted earlier the one that is created in 9, and which correctly imports back into 9 in the right position? I ask because it imports into a region of Washington state where Google satellite shows only trees and nothing that looks like the features in the image. It also imports as "Lambert Conformal Conic".

I imported that tiff into both 9.0.179.4 and also 8.0.35.0. It imported into 8 in a different position than in 9, so in 8 I right clicked on the Test Geotiff image in the project pane and used Assign Projection to see what the coordinate system parameters were. The false easting was wrong, but all the other parameters were correct. In 8 I set the False Easting to 500000, the same as reported by 9.

That put the image into the right location, as seen by putting it in a map using Google Maps Satellite Image as the base layer in both 8 and 9.

I suggest you send the image to tech support as a bug report, that it imports OK into 9 but when imported into 8 the false easting is wrong. Also, the tiff imports weird with some artifacts off to the side. Make sure to report in full detail how you created the image.

As to whether the image is being written by 9 with something off in the GeoTiff tags or whether 8 is getting something wrong on import, leave that up to tech support when they investigate your report.

To add to what Dimitri said:

One thing I did just notice is that the Manifold 9 base map is projected as State Plane Washington North. The image created by Save as Image is projected as State Plane Washington Northn (sic).

Can you post (or send to tech support, if you prefer that) a MAP file with an example component that does such renaming of coordinate systems during Save as Image? Because we cannot reproduce that and we obviously would like to fix it. (I suspect the component itself might contain nearly no data so you can delete nearly everything from it before sending, but if you go that road please check that the issue can still be reproduced on the data you are about to send.)

I sent the images I posted on this thread to tech support along with the offending tiff file. The Manifold 9 map used to create the image is very large (as in gigabytes) and contains proprietary information. Since the issue seems to be importing a tiff created in Manifold 9 into Manifold 8, it seemed more prudent to send the tiff causing the issue. Geotiffs created in Manifold 8 import into Manifold 9 correctly.