The reason you get a decoding error is because the font is actually damaged. Running it through TTX (a utility for switching font representations between binary and XML form), we get the following:
> ttx FRE3OF9X.TTF Dumping "FRE3OF9X.TTF" to "FRE3OF9X.ttx"... Dumping 'GlyphOrder' table... Error: cmap subtable is reported as having zero length: platformID 1, platEncID 0, format 0 offset 20. Skipping table. Dumping 'head' table... Dumping 'hhea' table... Dumping 'maxp' table... Dumping 'OS/2' table... Dumping 'hmtx' table... Error: cmap subtable is reported as having zero length: platformID 1, platEncID 0, format 0 offset 20. Skipping table. Dumping 'cmap' table... Dumping 'loca' table... Dumping 'glyf' table... Dumping 'name' table... Dumping 'post' table...
Since the cmap structure contains information about which character the font supports, the broken cmap sub-table is the final error. You can try to output broken data from other cmap subtitles (which Photoshop probably does), but this does not guarantee the correctness of the action. Browsers err on the side of caution, so this font will simply be rejected.
What to do: 1. report this font as broken for the website so that it can either fix it or delete it, 2. select a new font (I see you already did it) and 3. probably grab a copy of TTX for yourself checks for bad fonts in the future.
source share