Stereo Tool

Yes, that is what I do. I'm using Winamp as the player with the Advanced mIRC Integration Plug-In. http://amip.tools-for.net/wiki/manual/amip


The AMIP plug-in can do many different things. I have it writing what's playing to a file and do the RDS character set translation.

strrep(ü,chr(153),strrep(Ü,chr(217),strrep(É,chr(194),strrep(é,chr(130),strrep(Ö,chr(215),strrep(Ä,chr(209),strrep(Å,chr(225),strrep(ö,chr(151),strrep(ä,chr(145),strrep(å,chr(241),strrep(',chr(39),strrep(á,chr(128),strrep(Á,chr(192),%name)))))))))))))

--
Ragnar

In a typing spree, I came up with this:

RDS_CHARSET = ur"""€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€€ !"#¤%&'()*+,-./0123456789:,<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]€_€abcdefghijklmnopqrstuvwxyz{|}€€áàéèíìóòúùÑÇŞßİIJâäêëîïôöûüñçşĝıijªα©‰Ğěňȍπ€£$←↑→↓º¹²³±€ńȕµ¿÷°¼½¾§ÁÀÉÈÍÌÓÒÚÙŘČŠŽĐĽÂÄÊËÎÏÔÖÛÜřčšžđŀÃÅÆŒŷÝÕØÞŊŔĆŚŹ€ðãåæœŵýõøþŋŕćśź€€"""

€ means either a) I didnt find the unicode character (because it does not exist, I was too lazy to search, or because I could not decipher it), b) the RDS codetable does not assign a character to that byte.

HTH. Hans, can you maybe include transcoding from UTF-8 into ST and make it optional (backwards compatibility)?

Sounds like a VERY useful option, I wanted to propose it yesterday already but I first wanted to wait if it works - I suppose since ragnar.jensen uses it it does. I'll have to dive into character maps first I suppose, since different PC character maps have different symbols at different locations

No, you only need my string from above I suggest two or three modes: a) old mode (RDS encoding) b) latin1 input c) UTF-8 input. Your code for mode b would convert the user input to UTF-8 first. Then and in mode c, the string is converted using my string to the RDS encoding.

Yes, I run this successfully with a small transmitter that is fed a Stereotool-generated MPX signal.

EDIT: Ooops! Not entirely true, I use an external RDS encoder. I'll make a test with the latest beta and without the external encoder tomorrow.

--
Ragnar

Tests are done. Umlauts seem to get through OK, but there are issues. Characters just before an umlaut are sometimes shifted down one ASCII code, but not always.

Test 1: Audio file is tagged with Artist: Nürnberg åäöÅÄÖ Title: Sämre
Tags are displayed OK in Winamp


The file written by AMIP in the RDS character set, viewed in Notepad. It looks OK, it is supposed to look like this.


Same file in a Hex editor, looks OK.


Stereotool does not display the umlauts.


What it looks like in a radio display.

Characters before an umlaut are changed. N has become M, the space before å is now A with cedilla, S has become R.

Test 2:Audio file is tagged with Artist: Nü MÜSÉk Title: Räksmörgås
Tags are displayed OK in Winamp


The file written by AMIP in the RDS character set, viewed in Notepad. It looks OK, it is supposed to look like this.


Same file in a Hex editor, looks OK.


Stereotool does not display the umlauts.


What it looks like in a radio display.

N in Nü has become M. The R before ä in Räksmörgås has actually survived, but the g is now an f.

Test 3: Audio file is tagged with Artist: Räksmörgås Title: Nü MÜSÉk; Same as Test 2 but Artist and Title tags are swapped and a semicolon added to Title.
Tags are displayed OK in Winamp


The file written by AMIP in the RDS character set, viewed in Notepad. It looks OK, it is supposed to look like this.


Same file in a Hex editor, looks OK.


Stereotool does not display the umlauts.


What it looks like in a radio display.

Some characters before umlauts are shifted one character code down. Although the words are the same as in test 2, the changes are not the same.

The results are consistent between different runs of the same test case.
I have tested with several different RDS receivers and they all show the same in their displays.

--
Ragnar

Ah! The characters that are changing are at odd places (1, 3, 5, ...) - which makes sense because in RDS always 2 characters are encoded together. So I think I have a type error (signed vs. unsigned char).

I'm building a fix now.

The latest version in the News section should fix the issue with changed characters before umlauts now.

Here's the RDS default character set (there are others for Greek characters etc.). My question: The spec mentions that some cheap receivers can only display the characters inside the thick black line. What happens if you send lower case (I suppose they get converted?) - but also: What happens if you send special characters with umlauts etc? Are they replaced by the same character without umlaut, or completely left out?

Where I live there are no stations that broadcast weird characters in their RDS texts, you may have more experience with this

Lower case might vanish as well (blank chars).
I think special unsupported will appear as blank chars instead. But I think this refers mainly to the PS name. Devices that support RadioText should support most of the charset.
Yes, thats what I typed above in my string.