Post by hydrophilic on Jan 13, 2016 0:14:44 GMT
While optimizing the source code for SAM 128 (to make it compact to allow room for new features), I discovered an undocumented phoneme for SAM (but not Reciter).
The phoneme is "*" (an asterisk), and sounds about half-way between the SH and CH phonemes. I like to think of it as a short-duration "SH".
At first I thought this was a bug I created while optimizing SAM 128, but then I tested it on original C64 SAM and there it is!
Off-topic, I also discovered that Reciter treats a question mark, ?, different than SAM. Again, I thought I created a bug during optimization, but the same mis-feature exists in the original. In detail, Reciter will render a question mark the same as a period (or semi-colon). There is no direct mention of this in the original documentation (I checked both the C64 and Apple II dox)... however the original documentation does say that not all English sentences which end with a question mark should have a rising pitch (this is SAM behavior to a ?), but only questions which require a yes/no answer. So I'm theorizing that Reciter is too stupid to know when a question should have a rising pitch (or not) and thus defaults to substituting a period in its place (this results in reduced pitch at the end of a question).
As the ? behavior is undocumented, it is not completely off-topic
The phoneme is "*" (an asterisk), and sounds about half-way between the SH and CH phonemes. I like to think of it as a short-duration "SH".
At first I thought this was a bug I created while optimizing SAM 128, but then I tested it on original C64 SAM and there it is!
Off-topic, I also discovered that Reciter treats a question mark, ?, different than SAM. Again, I thought I created a bug during optimization, but the same mis-feature exists in the original. In detail, Reciter will render a question mark the same as a period (or semi-colon). There is no direct mention of this in the original documentation (I checked both the C64 and Apple II dox)... however the original documentation does say that not all English sentences which end with a question mark should have a rising pitch (this is SAM behavior to a ?), but only questions which require a yes/no answer. So I'm theorizing that Reciter is too stupid to know when a question should have a rising pitch (or not) and thus defaults to substituting a period in its place (this results in reduced pitch at the end of a question).
As the ? behavior is undocumented, it is not completely off-topic
