GeoTIFFs in ArcGIS 10

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

GeoTIFFs in ArcGIS 10

April Davis

All,

 

We have recently upgraded to Esri ArcGIS 10, and have started pulling in  4 band (red,green,blue,IR) GeoTIFFs.  We began to see areas of “no data”.  When looking at the image in other software platforms, we do not see the areas of “no data”.  All the pixels have a value.  We sent an image to Esri technical support and they replied by stating “The tiff has a mask tag that ArcGIS didn’t use in the past.  At ArcGIS 10.0 we read and apply this tag.  In this case, the tag specifies band 4 to be a nodata mask – any place where band 4 has the value 0 will show as nodata.  This is a setting on the tiff file and not in ArcGIS.  ArcGIS just started to honor this tag.”

 

Does anyone know what tag Esri may be referring to?  If I could figure out what the tag is or what exactly it specified, then I could remove the tag or change it.  But I am at a loss as to what to look for.

 

Any help would be appreciated.

 

April H. Davis, GISP

GIS Analyst III

Southwest Florida Water Management District

Mapping and GIS Section

2379 Broad Street

Brooksville, FL 34604

(v): 352-796-7211 x 4216

(f): 352-540-6018

www.watermatters.org

 

IMPORTANT NOTICE:  All E-mail sent to or from this address are public record and archived.  The Southwest Florida Water Management District does not allow use of District equipment and E-mail facilities for non-District business purposes.


_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: GeoTIFFs in ArcGIS 10

Max Martinez

April,

 

There are at least two possibilities.

 

First, the mask could be a whole other image in the TIFF file. This image would have a NewSubfileType tag value with the second bit turned on. You could disable this bit (then the mask image would either be ignored or treated as another image in the dataset) or you could remove the whole mask image by relinking the IFD chain without including the mask IFD.

 

The other possibility is that the mask is an additional sample included with the 4 band image (so the image has actually 5 samples). In this case the ExtraSamples tag would likely have a value of 2 for “Unassociated alpha data” for the sample representing the mask. If you changed this value to 0, the mask would possibly be treated as another band and you might then be able to avoid displaying it.

 

See the TIFF spec for more info. You may also want to investigate the source of the images (why is that mask being put in there in the first place?).

 

Max

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of April Davis
Sent: Thursday, April 07, 2011 2:20 PM
To: [hidden email]
Subject: [Geotiff] GeoTIFFs in ArcGIS 10

 

All,

 

We have recently upgraded to Esri ArcGIS 10, and have started pulling in  4 band (red,green,blue,IR) GeoTIFFs.  We began to see areas of “no data”.  When looking at the image in other software platforms, we do not see the areas of “no data”.  All the pixels have a value.  We sent an image to Esri technical support and they replied by stating “The tiff has a mask tag that ArcGIS didn’t use in the past.  At ArcGIS 10.0 we read and apply this tag.  In this case, the tag specifies band 4 to be a nodata mask – any place where band 4 has the value 0 will show as nodata.  This is a setting on the tiff file and not in ArcGIS.  ArcGIS just started to honor this tag.”

 

Does anyone know what tag Esri may be referring to?  If I could figure out what the tag is or what exactly it specified, then I could remove the tag or change it.  But I am at a loss as to what to look for.

 

Any help would be appreciated.

 

April H. Davis, GISP

GIS Analyst III

Southwest Florida Water Management District

Mapping and GIS Section

2379 Broad Street

Brooksville, FL 34604

(v): 352-796-7211 x 4216

(f): 352-540-6018

www.watermatters.org

 

IMPORTANT NOTICE:  All E-mail sent to or from this address are public record and archived.  The Southwest Florida Water Management District does not allow use of District equipment and E-mail facilities for non-District business purposes.
 

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: GeoTIFFs in ArcGIS 10

Even Rouault
Selon Max Martinez <[hidden email]>:

Running tiffinfo (utility from libtiff distribution) on your file might give you
more hints on what is going on.

> April,
>
>
>
> There are at least two possibilities.
>
>
>
> First, the mask could be a whole other image in the TIFF file. This
> image would have a NewSubfileType tag value with the second bit turned
> on. You could disable this bit (then the mask image would either be
> ignored or treated as another image in the dataset) or you could remove
> the whole mask image by relinking the IFD chain without including the
> mask IFD.
>
>
>
> The other possibility is that the mask is an additional sample included
> with the 4 band image (so the image has actually 5 samples). In this
> case the ExtraSamples tag would likely have a value of 2 for
> "Unassociated alpha data" for the sample representing the mask. If you
> changed this value to 0, the mask would possibly be treated as another
> band and you might then be able to avoid displaying it.
>
>
>
> See the TIFF spec
> <http://partners.adobe.com/public/developer/en/tiff/TIFF6.pdf>  for more
> info. You may also want to investigate the source of the images (why is
> that mask being put in there in the first place?).
>
>
>
> Max
>
>
>
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of April Davis
> Sent: Thursday, April 07, 2011 2:20 PM
> To: [hidden email]
> Subject: [Geotiff] GeoTIFFs in ArcGIS 10
>
>
>
> All,
>
>
>
> We have recently upgraded to Esri ArcGIS 10, and have started pulling in
> 4 band (red,green,blue,IR) GeoTIFFs.  We began to see areas of "no
> data".  When looking at the image in other software platforms, we do not
> see the areas of "no data".  All the pixels have a value.  We sent an
> image to Esri technical support and they replied by stating "The tiff
> has a mask tag that ArcGIS didn't use in the past.  At ArcGIS 10.0 we
> read and apply this tag.  In this case, the tag specifies band 4 to be a
> nodata mask - any place where band 4 has the value 0 will show as
> nodata.  This is a setting on the tiff file and not in ArcGIS.  ArcGIS
> just started to honor this tag."
>
>
>
> Does anyone know what tag Esri may be referring to?  If I could figure
> out what the tag is or what exactly it specified, then I could remove
> the tag or change it.  But I am at a loss as to what to look for.
>
>
>
> Any help would be appreciated.
>
>
>
> April H. Davis, GISP
>
> GIS Analyst III
>
> Southwest Florida Water Management District
>
> Mapping and GIS Section
>
> 2379 Broad Street
>
> Brooksville, FL 34604
>
> (v): 352-796-7211 x 4216
>
> (f): 352-540-6018
>
> www.watermatters.org
>
>
>
> IMPORTANT NOTICE:  All E-mail sent to or from this address are public
> record and archived.  The Southwest Florida Water Management District
> does not allow use of District equipment and E-mail facilities for
> non-District business purposes.
>
>


_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: GeoTIFFs in ArcGIS 10

Frank Warmerdam
In reply to this post by Max Martinez
On 11-04-07 09:18 PM, Max Martinez wrote:

> April,
>
> There are at least two possibilities.
>
> First, the mask could be a whole other image in the TIFF file. This image would
> have a NewSubfileType tag value with the second bit turned on. You could
> disable this bit (then the mask image would either be ignored or treated as
> another image in the dataset) or you could remove the whole mask image by
> relinking the IFD chain without including the mask IFD.
>
> The other possibility is that the mask is an additional sample included with
> the 4 band image (so the image has actually 5 samples). In this case the
> ExtraSamples tag would likely have a value of 2 for “Unassociated alpha data”
> for the sample representing the mask. If you changed this value to 0, the mask
> would possibly be treated as another band and you might then be able to avoid
> displaying it.

April,

I would add that it would be helpful to see a report on the contents of the
file from the tiffinfo utility.  I believe it is available as part of
the FWTools for windows kit at:

   http://fwtools.maptools.org/

If you install the software, then launch the FWTools shell and enter a
command like:

   tiffinfo C:\path\to\your.tif > C:\path\to\your.txt

Then send the .txt file to the list.  That will give more concrete
information on how the file is setup.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, [hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff