With devices called "Echo" it should be easy to create echo like effects for spoken voice. But surprisingly it turns out that within SSML there is no out of the box support for such a feature. Using some tricks with APLA it is possible though to create various echo and distortion effects for spoken text.
The above example code "echo-and-distortion.json" demonstrates one possible approach how to do it.
The basic idea is to use APLA to say an utterance multiple times in parallel with increasing small delays.
In the example code you can create various echo and distortion effects by modifying the parameters "offset" (any integer describing delays in milliseconds) and "echoNum" (from 1 to 10) at the very end of the file within the datasources segment of the json.
You can play around, experiment and listen to the effects in the developer console using the "Multimodal Responses" option which you can find as an option under the "Build" tab.