P3D v4 Resample creates bgl with black border line

[Clutch Cargo]

This is an issue I just encountered that I don't believe I have seen for years. I created some ground imagery where the resulting .bgl is creating a single pixel-width black border on the right border side only. At further inspection the line is actually not black but water.

My process as follows:

1. Create GeoTIFFs using Global Mapper from USGS .sid files. The data completely covers areas where I am creating images

The resulting image very close up in photoshop of the lower right corner. All data there... no gaps. RGB image with no alpha.

2. Process the images using the following inf file. Very basic, no blendmask or watermask, just a simple day and night texture.

[Source]
Type                             = MultiSource
NumberOfSources       = 2


[Source1]
Type                  = GeoTIFF
Layer                 = Imagery
SourceDir          = "P:\AZ\Section_23\Images"
SourceFile         = "AZ_K21_CC.tif"
Variation           = DAY

[Source2]
Type                     = GeoTIFF
Layer                    = Imagery
SourceDir             = "P:\AZ\Section_23\Images"
SourceFile            = "AZ_K21_LM.tif"
Variation             = Night

[Destination]
DestDir                            = "P:\2_Output\Section_23"
DestBaseFileName         = "AZ_K21"
DestFileType                   = BGL
LOD                                 = Auto
UseSourceDimensions   = 1
CompressionQuality      = 85

The resulting image very close up in photoshop of the lower right corner. All data there... no gaps. RGB image with no alpha.




Here is what the resampled .bgl file looks like in TMV. In the View tab if I select 'Show Missing Data Mask' you can then see a gray single pixel width line appear.

So I suppose the issue is resample is not seeing that last line of image data? If you look closely between the photoshop image the TMV image you can see the last column of pixels is literally missing in TMV. Been researching the forums for a similar issue but have only come across one without really a conclusion. Other posts talk about a red line but that does not appear to be the same issue.

I have tried using NULL VALUE = 0,0,0 and NULL VALUE = 255,255,255 and NULL VALUE = 128,128,128 in the INF but same results. Has anyone come across this before in their endeavors?

Thanks,

Marcus

 

[Clutch Cargo]

Making headway. Usually, I will use Global Mapper to break up the image into workable parts. This time I created just one GeoTiff and resampled it and it come out looking correct... no black border. So the INF works as written so it must be with Global Mapper. So it's back to investigating. Fun!

 

[Clutch Cargo]

Hmmm... well done the exact same way with two other areas of imagery and they come out looking correct. This one no? So I'll throw it out there again to see if anyone has come across this before. Thx.

 

[Clutch Cargo]

If anyone reading this it appears after more testing, it is simply missing data for that one strip of pixels. As to why that strip of date is not read or processed is beyond me. I can fix it (work around?), by adding a a buffer of image data to extend the imagery beyond the missing pixels and that appears to work.