Confused about double letters.

[Puer Pedens]

can anyone give me a run down on how all the double letters are pronounced?
I know, "rr" is just "r" but extended, but "tt" is actually two "t" sounds. How about two c's like in ecclesia or two s's like in posset?
Thanks

 

[cinefactus]

Try something like, "this seat", or "black cat"

 

[Anbrutal Russicus]

Example 1, Example 2, Example 3, Example 4, even word-finally. The point is that you prolong a sound in such a way as to make it finish the previous syllable and also begin the next. The only special case is the double LL, which is clear like in Italian, Hungarian and the Turkish example, and not dark like in English or Finnish (single L is generally dark).

 

[Hector]

Basically, a double consonant letter sounds like a longer version of the single consonant sound. It should probably be at least twice as long, but possibly even more than twice as long (based on comparisons to modern languages that contrast long and short versions of consonants). As Anbrutal Russicus says, LL is a special case where there was a further difference in the quality of the sound.

I'm not sure exactly what "two 't' sounds" means. In the Latin sound system, long/double consonants are treated as if they are split between two syllables (and so behave similarly to a consonant cluster like pt or ct). However, syllable breaks aren't actually directly audible. Linguists refer to sounds like p, t, k, b, d, g as "plosives", and describe them as having multiple stages; the last stage of a plosive is called "release". In English, a sequence of two /t/ sounds between separate words might in very careful pronunciation have two separate releases, like if you pronounce "rat tail" in a way that sounds like it has a "pause" between the words. But I don't think we have any reason to suppose Latin speakers used two separate releases when pronouncing pp, tt, cc, bb, dd, gg. It's possible, but I think it's more likely they held the plosive for twice as long as usual, and then ended with a single release. Likewise, I think ss was just an extended version of a short s sound.

 

[Anbrutal Russicus]

That is precisely the problem I saw in the OP - the distinction they make only makes sense if they're describing double release for /tt/ in contrast to /rr/. And the English examples would canonically be pronounced with double release, which wouldn't be correct for Latin. Even so, I didn't think an exposition like the above would be as useful to the OP as examples of actual speech. A considerable experience has convinced me that talking about phonetic matters on the internet without providing audio recordings is generally undesirable. Here's an example of /nn/ with two releases, and here's /tt/ - both in Polish. This is specifically how they weren't pronounced in Latin - even the practice of spelling them double is a relatively late development.

As to the exact duration, judging by several studies it varies between 2-3 times the singleton duration in different languages, speech styles etc. The type of sound has a big influence on this - sonorants (L, N) have a significantly lower ratio than stops do, and for this reason geminate stops are much more common.