Betty Botter had some butter,
"But," she said, "This butter is bitter!
"If I put it in my batter,
"It will make my batter bitter!
"But a bit of better butter,
"That will make my batter better."
So Betty Botter bought some butter
Better than her bitter butter,
And she put it in her batter,
And her batter was not bitter.
(So, it was better Betty Botter
(Bought a bit of better butter.)
Can somebody submit the phonetic transcription of this tongue twister,
cuz i don't know what words i have to stress and what monosyllables i have to substitute for a schwa.
That's a nice transcription of Mxsmanic. Of course, it doesn't fit my pronunciation at all. Here's how I'd pronounce it.
Sorry about the crude ASCII approximation to the IPA. Note this is a phonemic not a phonetic transcription (and would thus also work for RP)
Also note that unless I were being careful I'd probably end up flapping those "t" as indicated in Mxsmanic's IPA transcription.
The "flap t" sounds much closer to a 'd' than to the true quality of a 't'. Is the difference just a matter of rhythm?
"Bedder" and "better" (with a "flap t") sound identical, to me at least. Maybe for Americans, the 'e' before "tt" is shorter than that before "dd", and that's about it?
The "flap t" sound is actually an 'r' sound, identical to the Spanish flapped 'r' (the r that _isn't_ rolled). It is indeed closer to d than to t, and it's a very common allophone for intervocalic /d/ in GAE.
The transcription I gave was General American English and fairly narrow (that is, more precise than a phonemic transcription).
Bedder and better are usually the same in GAE.
Your transcription of "so" is interesting. You don't actually pronounce it /sO/ instead of /sOu/, do you?
What is that concave line under /ai/? I don't believe it's part of the IPA, is it?
If you have a look at the IPA chart, you'll see that it's voiced alveolar tap or flap. You also notice that [d] (as in "did") is voiced alveolar plosive and that [t] (as in "tit") is a voiceless alveolar plosive. You'll also see that, as it is voiced, it naturally sound be closer to [d] than to [t]. So, yes, they are distinct but can you really call this flapped "t" a kind of "r" sound? The "r" sound in English, [r\] (in X-SAMPA, as in "ray") is a voiced (post-)alveolar approximant. Again, this is a different sound.
Is it right to call this flapped "t" a kind of "r" sound? It would probably make sense for the Spanish but, whilst "Betty" and "beddy" sound similar to me, "berry" sounds quite distinct. What does it mean to call it an "r" sound? If all it boils down to is that the Spanish write it with an "r", then fine, but for English speakers (who don't know Spanish) this might not fit. If you compare the way your tongue moves for [r\], [d] and  ( is X-SAMPA for this voiced alveolar tap), you'll find that the similarity of [d] and  lies in the fact that your tongue moves up to the top of your mouth but not for [r\].
Mxsmanic uses those concave lines to show that it's a diphthøng rather thæn two short vowels.
I realised that wasn't a broad transcription.
The flap t occurs regularly in Australian English too, so much that there's no consistency, as to a consensus on 't'. It only seems consistent in formal readings where the regular t (as appropriate) is often employed, but even then, newsreaders for example, often adopt the flap t. Having said that, it's not a regional thing, as it might be in the US.
Tom, a quick check of a spectrogram indicates that I pronounce /o/ as a very weak diphthong, namely, [oU]. Non-phonemic diphthongs tend to be very weak in GAE, based on spectrograms I've seen (and my own speech). I often transcribe them as monophthong allophones that most ESL students find easier (at least if they are coming from languages with only monophthongs, like French). Pronouncing them as monophthongs is no less incorrect than overpronouncing them as exaggerated dipththongs (which some students will do through overcorrection).
The phonemic dipthongs are far more distinct, as you might expect. In my speech and in other spectrograms I checked they are unmistakable. I always transcribe these, as they are essential to comprehension and they are very strongly marked in all variants of English.
I do usually transcribe /e/ as [eI], but I debate over it constantly, since it's pretty weak in GAE. For French students with their monophthongs it seems to be better if I point it out (but then I still have to have them back off a bit so that they aren't pronouncing two distinct vowels). I guess it's no stronger than [oU], and I usually don't transcribe that. The IPA doesn't allow for an indication of how marked a diphthong is, so for non-phonemic diphthongs I hesitate between overlooking the off-glide and transcribing it, because neither usually represents it exactly (it's there, but it's quite weak).
for a spectrogram showing my pronunciations of these diphthongs (plus [ju], which I don't actually consider a diphthong).
Tom, as Jim pointed out, those lines underneath the vowels show that they are to be pronounced as one dipthong (the second vowel is not independent of the first), rather than as two completely distinct vowels. Most transcriptions don't show this, but I do, because technically without that line, the vowels must be pronounced distinctly (as in Noël), and for students who speak languages in which such distinct vowel pairs exist, it's ambiguous to leave out this indication when transcribing diphthongs (English speakers do it a lot, but that's because they are lazy or don't know any better).
There are other, alternate ways of transcribing diphthongs in the IPA but this is the one that I think most explicitly shows the actual pronunciation.
I also transcribe the 'd' in bed correctly ([E], not [e] as so many English texts do), and the vowels [i] and [I] correctly (not [i:] and [i] as many English texts do). Sometimes I think that English lexicographers and other scholars never study any other languages, because they are very careless and incorrect in their transcriptions. This causes a lot of confusion for ESL students, so I take care to transcribe everything strictly as it should appear.