Troubleshooting & Advanced Notes
I’m not seeing as many results as expected / I’m seeing too many results!
The way results are displayed depends on the various parameter settings. Depending on how you need to see the results, adjusting the minimum confidence and marker interval plus whether to allow or disallow repeat matches can drastically change the output. When building out AI workflows, pick some solid short test clips and try reprocessing the results with different combinations of parameters until you’re happy with the output.
My worker action failed!
The failed action should indicate if it’s an Amazon error or something else. If it’s an Amazon error your job results may have expired, or there may be a problem with the input. It’s also worth checking over the parameters you’ve entered in the worker config to make sure they aren’t the cause of the issue. If you see anything around ‘Error 403: Access Denied’ the most likely issue is that the IAM credentials provided do not have the correct permissions for the requested task. Speaking of permissions…
Which permissions do I need on AWS?
The AWS IAM permissions or policies are required for whichever account is used are:
For the S3 remote volume configuration, LIST, GET, PUT and DELETE permissions on the applicable bucket should do the trick.
Which CatDV fields are required?
The clip fields created and used by the plugin are:
It is easiest to create these fields before running a job, but the plugin will check and create them on the first run if they do not exist. The only field which end users should be concerned with on a day to day basis is the job status field which shows “IN_PROGRESS” or “SUCCEEDED” depending on the state of the job.
I hit some other bug/issue!
Please report any unrecoverable task failures to Square Box Support. Please note that support is available to solve bugs in the software and assist with software installation / upgrades - workflow design consultation would typically constitute chargeable work, available via Square Box Professional Services.
For system builders and power users, here are some additional plugin options which can be set in the Worker Node Advanced Properties box:
amazonAI.transcribe.markerSeparator = word / , ! ? (change the default marker separator of “.” to something else)
amazonAI.transcribe.maxMarkerLength = # (change the built-in marker length limit of 160 chars)
amazonAI.transcodeSizeThreshold = # (# = megabytes; use 0 to always transcode media for analysis)
amazonAI.outputRawResponseDataToField = <field identifer> (writes a JSON object to the field of your choosing e.g. clip[ai.json] - note these responses can be half a megabyte or larger which doesn’t sound like much but can crowd your database if you’re not careful, so it is recommended only to use this feature in a transient nature and then remove the extra data when finished.)
amazonAI.outputRawResponseDataToFile = <path> (same as above, but to an external text file (e.g. /Volumes/MyData/myResponse.txt) which you could parse separately and then toss away, rather than keeping all of the raw response data in the CatDV database.)
Please find the current list of languages supported by Amazon Translate here:
Amazon Transcribe – Supported Languages
Please find the current list of languages supported by Amazon Transcribe here:
John Vaudin for the generous development support
Rolf Howarth for creating the CatDV platform
Dave Clack for the opportunity to develop this application.
my wife Katherina for her incredible patience during many long nights of coding.
Updated October 6, 2020