Rubbersheeting produces staggered lines and letters

[Anonymous]

We have our old hand drafted Plat of Surveys (land surveys) scanned in .cal format. When rubbersheeting them to conform them to boundary lines in cad they unexpectedly get 2nd kind of distortion: lines and letters became zigzagging and staggered they loose the "smoothness" and are unacceptable as final plat. We didn't have this problem in LD3. Any help will be greatly appreciated, Thank you, Przemek Panek

[Anonymous]

On Wed, 9 Jun 2004 09:15:56 -0500, "ppanek" wrote: >We have our old hand drafted Plat of Surveys (land surveys) scanned in .cal >format. >When rubbersheeting them to conform them to boundary lines in cad they >unexpectedly get 2nd kind of distortion: >lines and letters became zigzagging and staggered they loose the >"smoothness" and are unacceptable as final plat. >We didn't have this problem in LD3. Newer versions of Raster Design have a choice of Rubber Sheeting methods. "Triangular" is the new method, it is more accurate than the method that Raster Design used to use, "Polynomial". Perhaps the old method will work better for your images. Phil Daley

[Anonymous]

On Wed, 9 Jun 2004 09:15:56 -0500, "ppanek" wrote: >We have our old hand drafted Plat of Surveys (land surveys) scanned in .cal >format. >When rubbersheeting them to conform them to boundary lines in cad they >unexpectedly get 2nd kind of distortion: >lines and letters became zigzagging and staggered they loose the >"smoothness" and are unacceptable as final plat. If you want, post an image and a drawing with marked rubber sheet points to autodesk.autocad.customer-files and I will take a look at it. Phil Daley

[Anonymous]

I've seen someing like this in 2005 too... to get better quality I had to promote the binary image to Grayscale using the IDEPTH command. Then convert it back to binary after the rubbersheet. "ppanek" wrote in message news:40c71b95$1_3@newsprd01... > We have our old hand drafted Plat of Surveys (land surveys) scanned in .cal > format. > When rubbersheeting them to conform them to boundary lines in cad they > unexpectedly get 2nd kind of distortion: > lines and letters became zigzagging and staggered they loose the > "smoothness" and are unacceptable as final plat. > We didn't have this problem in LD3. > Any help will be greatly appreciated, > Thank you, > Przemek Panek > >

[Anonymous]

Thanks for the input Chris. I have posted 2 files: Before.cal and After.cal example files of the distortion in Autodesk.autocad.curtomer-files. All the LD2005 stations in office produce the same error while rubbersheeting. We used triangular rubbersheeting in this example but polynomial method yields the same error. Before.cal - raster before rubbersheeting After.cal - raster after rubbersheeting, 4 anchor points at 4 border corners 2 anchor points at North and South end of line labeled 97.37' 1 source - destination point at the East end of line labeled 125.18' 1 source - destination point at the East end of line labeled 124.17' Requested shift at the 2 source - destination points was no bigger than 1% of the length of entire raster image in this example. We tested much smaller shifts with exactly same very bad, unacceptable result. Any help will be greatly appreciated, Przemek Panek

[Anonymous]

I seem to be able to duplicate the staggered lines with any binary image, with any set of points. Grayscale and color images rubbersheet fine. The only solution I have found so far is to temporarily convert the image to grayscale (8-bit/palette) using IDEPTH, rubbersheet it, then convert it back to binary (1-bit). It seems to be a quick conversion, and provides good quality output from what I have tested. "ppanek" wrote in message news:40d362ea$1_3@newsprd01... > Thanks for the input Chris. > I have posted 2 files: Before.cal and After.cal example files of the > distortion in Autodesk.autocad.curtomer-files. All the LD2005 stations in > office produce the same error while rubbersheeting. We used triangular > rubbersheeting in this example but polynomial method yields the same error. > > Before.cal - raster before rubbersheeting > After.cal - raster after rubbersheeting, > > 4 anchor points at 4 border corners > 2 anchor points at North and South end of line labeled 97.37' > 1 source - destination point at the East end of line labeled 125.18' > 1 source - destination point at the East end of line labeled 124.17' > Requested shift at the 2 source - destination points was no bigger than 1% > of the length of entire raster image in this example. > We tested much smaller shifts with exactly same very bad, unacceptable > result. > Any help will be greatly appreciated, > Przemek Panek > >

[Anonymous]

Thanks to these posts I was able to reproduce the bug and fix it. On Tue, 22 Jun 2004 17:44:55 -0400, "Chris Dodge [Microdesk]" wrote: >I seem to be able to duplicate the staggered lines with any binary image, >with any set of points. Grayscale and color images rubbersheet fine. > >The only solution I have found so far is to temporarily convert the image to >grayscale (8-bit/palette) using IDEPTH, rubbersheet it, then convert it back >to binary (1-bit). It seems to be a quick conversion, and provides good >quality output from what I have tested. > >"ppanek" wrote in message news:40d362ea$1_3@newsprd01... >> Thanks for the input Chris. >> I have posted 2 files: Before.cal and After.cal example files of the >> distortion in Autodesk.autocad.curtomer-files. All the LD2005 stations in >> office produce the same error while rubbersheeting. We used triangular >> rubbersheeting in this example but polynomial method yields the same >error. >> >> Before.cal - raster before rubbersheeting >> After.cal - raster after rubbersheeting, >> >> 4 anchor points at 4 border corners >> 2 anchor points at North and South end of line labeled 97.37' >> 1 source - destination point at the East end of line labeled 125.18' >> 1 source - destination point at the East end of line labeled 124.17' >> Requested shift at the 2 source - destination points was no bigger than 1% >> of the length of entire raster image in this example. >> We tested much smaller shifts with exactly same very bad, unacceptable >> result. Phil Daley

[Anonymous]

Where is the fix available? "Phil Daley" wrote in message news:tld5e016v5oeucsqgf5dqds726q2ilceha@4ax.com... > Thanks to these posts I was able to reproduce the bug and fix it. > > > On Tue, 22 Jun 2004 17:44:55 -0400, "Chris Dodge [Microdesk]" > wrote: > > >I seem to be able to duplicate the staggered lines with any binary image, > >with any set of points. Grayscale and color images rubbersheet fine. > > > >The only solution I have found so far is to temporarily convert the image to > >grayscale (8-bit/palette) using IDEPTH, rubbersheet it, then convert it back > >to binary (1-bit). It seems to be a quick conversion, and provides good > >quality output from what I have tested. > > > >"ppanek" wrote in message news:40d362ea$1_3@newsprd01... > >> Thanks for the input Chris. > >> I have posted 2 files: Before.cal and After.cal example files of the > >> distortion in Autodesk.autocad.curtomer-files. All the LD2005 stations in > >> office produce the same error while rubbersheeting. We used triangular > >> rubbersheeting in this example but polynomial method yields the same > >error. > >> > >> Before.cal - raster before rubbersheeting > >> After.cal - raster after rubbersheeting, > >> > >> 4 anchor points at 4 border corners > >> 2 anchor points at North and South end of line labeled 97.37' > >> 1 source - destination point at the East end of line labeled 125.18' > >> 1 source - destination point at the East end of line labeled 124.17' > >> Requested shift at the 2 source - destination points was no bigger than 1% > >> of the length of entire raster image in this example. > >> We tested much smaller shifts with exactly same very bad, unacceptable > >> result. > > Phil Daley

[ANDREAMARKAC3430]

In connection with the topic already discussed few months ago, I just want to know if the problem of rubbersheeting bitonal b/n images and creating staggering lines has been solved.
It seems to me that in some way the image risulted is divided in vertical areas and for some of them an invertion happened.
Anyway it is very strange to get this problem in Raster Design 2005 when in Autocad Overlay it worke properly.

With Kind Regards

A. Markac

[Anonymous]

On Mon, 20 Sep 2004 07:19:11 GMT, SOGEPI wrote: >In connection with the topic already discussed few months ago, I just want to know if the problem of rubbersheeting bitonal b/n images and creating staggering lines has been solved. >It seems to me that in some way the image risulted is divided in vertical areas and for some of them an invertion happened. >Anyway it is very strange to get this problem in Raster Design 2005 when in Autocad Overlay it worke properly. This was fixed in SP1. Phil Daley