1.214 (10/16/2020)

Features

• Updated S3 access to use virtual host-style access where feasible based on bucket name to support future S3 path style access deprecation. Buckets with dots in their name or buckets in us-east-1 with non-DNS-compliant names are still accessed via path style. More info on Amazon’s S3 Path depreciation plan: https://aws.amazon.com/blogs/aws/amazon-s3-path-deprecation-plan-the-rest-of-the-story/
• Added support for specifying custom fonts when imprinting subtitles using the webkit render engine
• Added support for fil (Filipino) language code for audio tracks
• Added support for hiding and showing individual datasets on the Analyzer Results tab by clicking on the labels in the graph legend
• Removed redundant Y axes from Analyzer Results tab when multiple datasets use the same Y axis
• Added ability to edit existing computing groups even if the AWS IAM API is unavailable (e.g. due to outage)

Improvements

• Improved messaging in web console when creating a Google Cloud Platform computing group fails because the credentials don't have sufficient permissions
• Web Console security improvements
• For Google Cloud Platform credentials, the Credentials Vault now displays the GCP service account email in the Key ID column
• Added option to set a default language for audio tracks that are missing a language tag or have a non-ISO639-compliant tag

Fixes

• Resolved issue where error workflow connections may not be triggered if they have multiple elements in the from array
• Resolved transcoding performance issue when using JPEG2000 sources
• Resolved issue where start pending records for on-demand machines could be left behind, causing delayed machine launches for later queued jobs
• Resolved error when using binary_sequence sources
• Resolved issue where a 2-channel track may be detected as 5 channels
• Resolved issue where some audio sources were erroneously detected as DTS format
• Resolved credentials error when using the package_task to create DASH outputs on Google Cloud Storage
• Resolved issue where channel_compare analysis would omit results if correlation_threshold was set to 1
• Resolved issue where the wrong instance pricing was displayed on the Machine Activity page for on-demand instances
• Resolved issue where applying a trim in the source_pipeline when stitching multiple components would trim each component instead of the fully composed presentation