I’ve noticed that when ever Alexa speaks on my Echo Dot the way she enunciates is completely different and sounds robotic when listing out the weather metrics.
This is not the case with the larger Echo I have.
At first I thought its because of the lower quality speaker but having played different music tracks, and had Alexa speak using different skills etc. This has nothing to do with the speaker as it has to do with how the skill has been formatted.
May I ask the Weather Flow team to review this document which indicates many attributes can be changed to make her sound more natural.
I’ve got two Echo Dots and the WF report sounds ok to me with my Dots. I have noted though in the last couple of weeks, that Alexa seems to have undergone an upgrade with some pronunciations, as she would not say my home towns name properly, but she does now.
This is interesting. We deploy one skill to Amazon for all devices. So, there are no differences on our end from device to device. We do use SSML in the skill to improve the speech. We could make use of it a bit more, but from our perspective everything is the same across devices.
