Heill Óskar:
I relabelled the mail, since your TABLE may be useful to
keep as reference for later.

Óskar's Table:

>
>SAMPA IPA character
>
>: : (lengthening marker)
>A script a
>E eta
>I capital i
>i i
>O inverted c
>u u
>Y capital y
>9 oe ligature
>
>p^h p with h superscript
>t^h t with h superscript
>c^h c with h superscript
>k^h k with h superscript
>p p
>t t
>c c
>k k
>v v (or u psilon)
>D ð
>z z
>j j
>G gamma
>f f
>T theta
>s s
>C ç, c-cedilla
>x chi
>h h
>l l
>m m
>n n
>N n with longer right foot (ng before g, k)
>J n with longer left foot (ñ in señor)
>r r
>^0 empty circle added to a symbol (to indicate devoicing)
>
>
>
>There, now you have all the standard sounds of Modern Icelandic. Now
>bring on that transcription :)

Now you saved me a lot of work. For now I shall be able to transcribe
IPA --> SAMPA without difficulty. Let me also give the reference here:
Magnús Pétursson, Lehrbuch der Isländischen Sprache. Mit Übungen
und Lösungen. 3. Auflage, BUSKE Verlag, Hamburg 1992. The reading
excercise called "The Northwind and the Sun" is on page 52 in the
book op.cit. It isn't very long. and the language is simple, since
it is the very first piece of Icelandic in the book., and it comes
immediately after he has discussed the phonetic alphabet that he calls
API mixed with some local Icelandic conventional signs, which I
understand is the norm on Iceland (?) But what I'll do is to
is to send it in a separate post with appropriate header. Since it
will be much easier for me to find it later in the archives then.

>> A thing I noticed was that MI seems to pronounce the "d" as
>> it occurs orthographically in Icelandic words, as the sound that
>> we call a "t". Also some b's get changed to p's. But apparently
>> not always. (maybe at the beginning of words?) But perhaps this
>> is only like a "gliding towards"? Later..
>
>That's because the orthographic <b> <d> <g> (lenis stops) are
>pronounced _unvoiced_ in Icelandic; in English, they're pronounced
>voiced, while <p> <t> <c/k> are pronounced unvoiced and aspirated.
>Icelandic only makes a distinction through aspiration, not voice.
>French, and other Romance languages, only makes distinction through
>voice, and do not have aspiration. <g> is pronounced "softer" (to use
>an untechnical term) in various in-word positions, corresponding to
>the relationship between <ð> and <d> (in fact, one could replace all
><ð> with <d> and still have perfect comprehensibility).
>

In my own language I just say the sound the way I think it sounds right.
And so I wouldn't be able to describe the mechanics very accurately.
But off hand I think that in Norwegian p, t and k are "wind sounds"
(no vibes of vocal chords) whereas b, d and g are their voiced counterparts.
You probably know that southern Scandinavian dialects (Denmark, Skåne,
South Norway) has seen a glide of p, t, k --> b, d, g which we
sometimes call softening ("den bløde kysstriben"). But disregarding this
(which only concerns the pronounciation of certain groups of words,
and has not eliminated the ptk sounds) I would say the distinction
voiced/unvoiced applies to the groups bdg/ptk in all of Scandinavia,
as well as Netherlands, Germany, yes even England. French is further
away, but I don't particularly recall any especial care had to be taken,
except switching on a general "french tone".

>Right, I'll just try to supply you with a list of the SAMPA versions
>of the IPA characters you have there (or that is, those that occur in
>Icelandic):

I take it API is the same as IPA? just like OTAN = NATO.

Keth