[IDV #PUJ-510012]: IDV - Failure to load DEM files

Hi Steve-

> Don, you will probably not know that I ran into load problems with larger
> DEM files and have been working with Stuart to resolve the issue. This email
> is to bring everyone up to date.
> 
> First a little background info. I create my base grids in Surfer (GRD),
> transfer the GRDs to Globalmapper and output DEMs for import into IDV. The
> GRDs are okay - no problems other than errors I introduce in my processing.
> 
> What I have discovered is that Globalmapper produces DEMs that have some
> columns shuffled by one or two grid nodes compared to the original GRDs
> (doesn't seem to do the same with rows). All apps (Surfer being one) that I
> have tried show the DEMs with this oddity. I don't know what causes it,
> don't even know if it is a real error. Don spotted these offsets in the
> columns (used the term faults) when he first viewed my DEM but I assumed he
> was referring to my own processing errors that I know exist in the original
> GRD. Also, it seems that the large number of 'blank areas', null-data areas,
> in my datasets might be contributing to these offsets.

Glad to know it's not just my software. ;-)

> Anyway, what I have discovered is that some DEMs will load to IDV, others
> will not. Big DEMs, 40megs or so, will always fail; smaller DEMs 2-15megs
> might load but they have the offsets. I've tried all sorts of tricks in
> Globalmapper to overcome the issue but to no effect. As you know
> Globalmapper was formerly owned and written by USGS so I'm assuming the DEM
> output is correct.

When you say they fail, do they give an error message or do you just run
out of memory?  If the latter, I can understand that.  We are not very
efficient in some areas and have no way of subsetting/decimating when
we read in the grid.  If the former, could you provide a sample file
for me to test with?

> So, I'm side-stepping the issue and using ARC Ascii grids instead, i.e.
> taking an original GRD file, importing it into Globalmapper, converting it
> to ARC ASCII grid and importing that into IDV. That works. The file is 90
> megs -  bit big for IDV! Sucks up all 1500megs of RAM I've assigned to IDV
> and then IDV starts caching to disk. Shrinking the geographical area I
> examine is something I'll have to do.

Same issue with memory here.

> So, I don't think I can do anything more with the DEM problem but I can make
> good use of GEON-IDV, especially since Stuart has helped with earthquake
> data formats.
> 
> Thanks to you both for assisting. If I can assist you, by testing further,
> let me know.

Thanks for your input and bug finding!

Don


Ticket Details
===================
Ticket ID: PUJ-510012
Department: Support IDV
Priority: Normal
Status: Open