I really like the new audio for the example sentences. One thing though: when I push the speaker button the audio plays several times in a row. Sometimes it will play it once the first time and then it will go multiple times after that.

So yeah, I'm alpha testing this; it's a text to speech solution that we're testing out for the Chinese example sentences, provided for free for now by another Skritter user to see if we like it. Feedback is welcome on its usefulness. I will try to fix the strange bugs with it--what browser are you using, JimAndress, and are you sure you didn't click the button multiple times while waiting for it to load?

It's cool, but I don't think I'll use it on any regular basis.

I doubt I would use it on a regular basis either. At the moment, I can't get it to work at all. I click on the button, and it attempts to connect to their website. However, it just gets hung up there.

Firefox 3.6.13
Windows Vista

I like it, it is bit overlapping with iself sometimes, and a bit too fast for my level, however i really love it,

I'm seeing the same thing with the endless loop playback. I won't use it a lot, but I'd be awfully grateful to have it there. Thanks to whoever provided it!

I've made a tweak that might fix the repeats; let me know if you still see problems. We might be able to get them to slow down, too.

One question is, do you think that manually clicking a button to hear the audio when desired is best, or is there another solution that might be more useful, like having it automatically play (but cut itself off when moving to the next word)?

1.
IMHO, she speaks definitely to fast. Isn't it possible make it slower?
2.
I think manually clicking is better. Furtermore, I would suggest you define a keyboard shortcut for it.

Thanks :)

I think the speed now is more adequate, at least for my newebe level.

I wouldn't make it playing automatically (like words): sometimes I just want to go fast through parts studying; but sometimes I like the additional option of earing characters and words in a context and to try to repeat the example sentence.

I'm late to comment on this, having just relocated my audio volume slider - thanks, Nick:). I had turned it off b/c I don't want it on all the time (trying to more discreetly skritter in public).

Seems to me the speed is about right, kinda slow but not awkwardly. The overlapping and computer-generated-ness is not ideal but the most important thing is that all the tones are right so I think it's helpful. I would definitely not want to hear every sentence; I'd stop listening or turn off the audio and forget how to turn it back on again. I don't look at the sentence for every entry, either.

About the sentences: some have commented that the current selection is poor, that they're not useful. Maybe not useful in conversation (I'd never actually say that stuff) but very useful to demonstrate the proper usage, fine points of meaning and so on. Perhaps for the simpler vocabulary more usable sentences would be preferable but I like the current range. I do have a question - are some of them really legitimate sentences? For example the first 4 or so of 行为 say "the act or process of xxx" - to my native English ear those don't sound like sentences, rather phrases. I keep meaning to ask my teacher.

In the end, another useful feature your team has made available. Thanks for continued good work.

P.S. It seems that the volume of the char/word audio isn't equivalent to the sentences. The sentence volume is lower, maybe 30% lower. So if I have it adjusted to my liking for words, I cannot easily hear the sentences. Just me?

What the example sentence audio irritates me most is that it doesn’t seem to understand the sandhi rules. Take the following sentence (from 暖烘烘) as an example:
狗倚着暖烘烘的散热器.
gǒu yǐ zhe nuǎnhōnghōng de sànrèqì.

The first character 狗 should be changed to 2nd tone because it is followed by 倚, which is also a 3rd tone. Then the character 暖 should be half 3rd tone instead of the full 3rd tone. Try copying the sentence into Nciku and hear how it is read there.

Then, the tones will fluctuate for some unknown reasons. Let’s look at the examples from 一会儿. (I’m not talking about 儿化. I understand that they are too difficult for synthesizers.)

来访者在中午前一会儿离开了.
炸弹一会儿就爆炸了.
It seems that 一会儿 is read as [yìhuǐr].

Then in 医生一会儿就到, it is [yíhuǐr].

But, actually, it can read it as [yíhuìr] as can be seen in:
所有的灯都闪了一会儿
他每次外出都会顺道去那家书店待一会儿.

And that’s why I’ll usually use Nciku to read them out instead despite all the troubles involved.