Save files.
Safe files.

Need turbo.
Neet turbo.

That's what I'm talking about. How much difference is there? The mouth and tongue movements are the same, the only difference depends on how soon you start devoicing the consonant.
Problem: that difference sounds too small too me. It might be easier to detect in some pairs (ex: v/f) and be more difficult with plosives (ex: d/t). In connected speech, unless it's slow, that difference becomes too small. Regional and personal variations make everything even more complicated. There might be no difference at all unless it's in very slow speech, and I might be fooling myself.

What do you say about this? Thank you.
Some people claim that the vowel that precedes the /f/ or /v/ (the /t/ or /d/) is different depending on which of the pair is used. They claim that a vowel is longer before an unvoiced consonant. So for some speakers, the preceding vowel may be the right thing to listen to for the difference. (I can never hear this difference, though.)

I don't devoice much, I guess, so I claim that I say those pairs differently. save and safe, as I say them, sound different to me, even before files. Same, only more so, with /d/ and /t/. Those really sound different to me, even before turbo -- unreleased /d/ vs. unreleased /t/ (followed by aspirated /t/).

CalifJim They claim that a vowel is longer before an unvoiced consonant.
Yes, but I noticed that vowel length only makes sense in stressed syllables, so as long as you stress the second word in each example this feature can be neglected.

Devoicing definitely varies from dialect to dialect, but... I can't "hear" all that difference. As soon as such cases happen in normal connected speech, the difference seems to disappear.

This is an example taken from the BBC Learning English on "gemination":

I'm a bit tired
We have a lot to do
Tell me what to say
She's slept for three hours
I've finished

I'm only interested in the last two examples in blue. I know that's British English, but... the topic is the same.
Teachers: We supply a list of EFL job vacancies
CalifJimThey claim that a vowel is longer before an unvoiced consonant.

Now, that's new to me Emotion: smile During my phonetic course I was told many times and this is the other way round, that is, vowels tend to be shortened (which is especially perceptible in the case of long vowels) before voiceless consonants. Emotion: smile

I'm just repeating what I heard. I don't know if it's true though.
Thanks for pointing out that mistake. I meant to say it the other way round! Whoops!

... longer before a voiced consonant.

I didn't notice the mistake either... I read it so fast I thought I had seen it right, lol. Yeah, longer before voiced consonants, and shorter before unvoiced consonants.... although I noticed that only applies to stressed syllables, and not always. I get the impression you can find long vowels before unvoiced consonants, but not short vowels before voiced consonants.

Anyway, this is probably off-topic, since vowel length doesn't seem to help distinguish the pairs we are discussing here.
Students: We have free audio pronunciation exercises.
KooyeenShe's slept for three hours
I've finished
I finally got around to listening to these examples, and I would be hard pressed to hear the difference between She slept and She's slept or between I finished and I've finished, given the way that that speaker says them. I think he's exaggerating the amount of assimilation there is between the two consonants purposely, making them sound like one consonant purely for illustrative purposes. The two consonants can merge into one, but (in my opinion) they don't always do so to such an extreme.

I usually hear (and say) the z sound, then quickly merge it into the s sound in She's slept. It may be a matter of milliseconds, but there is at least a tiny amount of time in which the z is present before the s bleeds in. (Likewise for the v to ftransition.) Anyway, that's my story, and I'm sticking to it! Emotion: smile

Sorry to be so late in responding. You're probably not very interested in the question by this time.

No, I'm still interested! I'm always interested, LOL.Thanks.
Yes, they are very devoiced in that clip, and I can't hear much gemination, so I think I'd have a lot of trouble to distinguish "She's slept" from "She slept", for example. Actually, I'm afraid I can't even pronounce them differently.

I might need to find a good book on English phonology if I want to know more. There are just too many weird linguistic features in English, and accent reduction courses are not comprehensive enough.