spectrum element in python

[Christian Leichsenring]

On 08/13/2015 01:37 PM, Christian Leichsenring wrote:
> The second problem is that with to_string() I can see that the data is a
> 2×128 array (with the number of bands being the default of 128), even
> though there are 48 channels in the audio stream. So I should really be
> getting a 48×128 array (with multi-channel set to true of course).

Ok, I fixed this second problem. Turns out I had to force the number of
channels in the caps filter to something larger than 2, not only set the
channel-mask to 0. Why the audio source caps lied to me and said it had
configured 48 channels after the pipeline went into PLAYING still
mystifies me but whatever.

Now is there likely to be a fix for the first problem as well or will I
have to parse the string?