On Fri, 2 Nov 2007, pondlife wrote:
> Hi Stéphane,
> I was actually thinking of a generic keyclick - i.e a sound made whenever a
> key is pressed (as quickly as possible, so you have feedback that your
> button press was successful).
As I said last week, I'd like to try it, but I'm not convinced. The
clicking might be annoying to some, and the wait/loading splash really
signifies a longer wait than usual, which I think is a different
> A while back I made an attempt at this:
> http://www.rockbox.org/tracker/task/7307 It needs syncing though.
Oh. OK I've updated it and tried it out. I'll upload my version and
comments shortly. It's not working too reliably for me though :-). I like
some aspects of it, but I don't feel it really replaces the loading splash
> Then remove the Loading splash completely.
Now do you mean remove the voicing of the loading splash, or remove the
> I don't particularly like the approach of #8067 as this adds an exception
> case, and deviates the voice UI from the displayed UI. (I feel strongly
> that the voice UI and displayed UI should always match wherever at all
I agree that the non-visual interface should try to convey the same
information as the on-screen UI. Using the same strings is often the best
approach. But I do feel that some creativity is called for in some cases
to provide a smoother non-visual UI.
Now if you're saying that the entire loading splash is annoying to you,
then that's another story. But your initial post said you were annoyed
specifically by its voicing. Seems to me speaking "loading" IS an exact
match to the displayed UI, and that's what's annoying you. But that splash
is indeed somewhat special, among other reasons because it's a zero
P#8067 adds an exception case in the voice processing, but it at least
conveys similar info to the displayed UI, and the exception makes it
smoother. Removing the voicing of the splash while keeping the splash on
screen would be a much bigger deviation IMO.
Speaking things outloud allows to convey a great variety of info and is
easily interpreted by new users, but it's not necessarily the only means
available in a non-visual UI. Sound effects are a bit like icons on a
screen: they're require less attention to interpret, although they work
only in cases where the interpretation is obvious. And of course sound is
a fairly linear medium compared to a screen, so things can't always match
If you mean to take out the wait splash completely and replace it with the
keyclick, you should probably ask sighted users for their opinion :-).
IMO hearing the click on each key press builds an expectation of
instantaneous response: you're so used to it that you stil feel things are
stuck when one particular key press suddenly seems to have no effect and
has you waiting for a few seconds. As I said before, I find the loading
splash most appropriate on disk spin ups. Conceivably it might be taken
out altogether on flash players. Oh and also, it could be argued that the
playback beep and the loading splash are two inconsistent ways of
conveying the same info.
> p.s. It's quite a nice noise you found there, but I'd expect the keyclick
> will have to use lower-level beep code (i.e. pcmbuf or piezo) though, for
Definitely. But OTOH a disk spinup takes long enough that there is plenty
of time to play a more indicative sound.
Now I've tried and updated your patch, and will go post comments on the
tracker. Will you at least try mine out please?
Received on 2007-11-02