Canoris Beta

From 10.00 to 12.00 this morning the Canoris team has been busy upgrading the production servers to bring you some new features and a more stable service. We've bumped the version number to 0.10!

For all newly uploaded content you will be able to not only retrieve the statistics of the analysis, but all the analysis data for each audio frame as well. The format of the data is JSON and can be retrieved from the /files/*filekey*/analysis_frames/ resource. Happy synthesizing!

We added one more descriptor to the highlevel analysis results: the music vs. speech classifier!

Instead of uploading complete files you can now also post URLs to publicly accessible files on the web. Canoris will download the file for you and process it like if you had uploaded it. See the documentation for the files resource.

Stelios Togias and Ciska Vriezenga created two new client libraries for Canoris, the canoris-java and the canoris-javascript libs. Also, be sure to update the python client library if you were using it before.

Other minor changes include a new uniform paging interface over the different REST collections, error responses as JSON instead of plain text. Also, temporary files (those that are the output of a processing task) will be deleted after two hours if you don't mark them as temporary.

Finally, we've decided to impose some limits on the API use. Please see the documentation on what these new limits are. We know that there are some applications out there for which these limits might be too strict. The good news is that if you contact us, we can easily help you out!