Crystal Reports for VS 2010 corrupts "some" Thai chars on PDF Export

Question

Any Thai character that contains a symbol above it (like an umlat in English) gets corrupted when exporting a report to PDF, yet the same report can be exported to Excel or Word without any issue.
See the following character - once with a symbol above one without:
ล้ ล
The PDF file will not display the first character - instead it displays the second character overlayed on a box (like you get when the font does not contain the required character)
If I create a text field in the Crystal designer, it does the same thing - but when I right click and edit, it displays correctly - until I click out of the field when the problem returns. So this suggests the problem is before the PDF export and maybe with the designer - but then why does it work if i export to Word?
I have tried may different fonts - including those specifically recommended and used by the Thais, but all have the same issue.
My work around at the moment is to use Word instead of PDF, as I do not believe there is a configuration change I can make that will resolve the issue, but would love to hear otherwise.
I am using Crystal Reports for Visual Studio 2010 version 13 running on Windows 7 x64. I have a beta 2 download of v14, but it requires a key code which i do not have, so as yet ahave been unable to verify if this issue is resolved in that version.
if anyone knows otherwise, please let me know
Cheers
Jed

Answer 1

Thanks Don,
(Good to know that version 13 I am running is the correct one)
I have tried many UTF-8 fonts, with largely similar results, but this testing here is with Cordia New - a font largely aimed at Thai Language and the one our Thai customers always use.
I have confirmed that right click and export from the IDE still shows the box on certain characters - but note that the box does not replace these characters (as would be the case if they were missing) but rather overlays them - which is why my subject includes the word 'corrupt'
All testing so far in on my Windows 7 development machine - so no user issues are involved
The code if fairly simple at this stage - define and load a data set - then link it to the report and use the built in export to disk functions:
                OutboundSummary rpt = new OutboundSummary();
                rpt.SetDataSource(outboundSummary);
                string tempDir = Environment.GetEnvironmentVariable("TEMP") + "
                string filename = tempDir + DateTime.Now.ToString("yyyy-MM-dd HH-mm-ss");
                rpt.ExportToDisk(ExportFormatType.PortableDocFormat, filename + ".pdf");
                rpt.ExportToDisk(ExportFormatType.WordForWindows, filename + ".doc");
If anyone wanted to verify, here is a piece of Thai text that will show the problem
ผลรวมแฟกซ์ที่ยังอยู่ในคิว
To better illustrate the problem, the attached image will help
[IMG]http://i55.tinypic.com/hweh42.jpg[/IMG]
I am coming to the conclusion that perhaps these characters are not single characters - at least do not appear as such with each font - and perhaps Windows does some magic to make them appear correctly. I can see that Word automatically converts them to a suitable font - i.e. a mixed language text file imports into word with different fonts for different languages - and you cannot change the font to one that does not support the language. I can specify Arial from Crystal which does not contain thai characters at all - but export to word and they show correctly.
Crystal spits the data out and Word converts it correctly, but acrobat does not - so maybe it is their fault?
I have all language packs on my machine - and doing Chinese next suggests that even if I find a thai font that worked, it would not work in chinese, so am thinking word or rtf might be the winner.
Cheers
Jed