V‑Spark 3.2

This section provides release notes for V‑Spark 3.2, describing new features and capabilities that were introduced in the 3.2 version of the product. This document also provides information about Known Issues in 3.2 with V‑Spark 3.2 that could affect your use of the product.

V‑Spark 3.2 Release Notes

V‑Spark Version 3.2 provides new language models, significant performance, usability, and stability improvements, and API enhancements to simplify transcript and audio retrieval.

  1. Added new request API to retrieve Text and JSON transcripts and MP3 audio from V‑Spark outside of the traditional callback interface. The request API provides 4 verbs to query job status and retrieve results by RequestID: status , summary , details , and result . See "Using the V‑Spark 3.5.0 API" for more information.

  2. Added Support for Non-English Languages

    1. These language models support the following features (requires Automatic Speech Recognition (ASR) greater than or equal to 5.4.1):

      1. File loading

      2. Playback and Display

      3. Ad-hoc Search

    2. The following features are not supported in these language models for 3.2:

      1. Applications

      2. Application Templates

      3. Reporting

  3. Internet licensing supported in V‑Spark. Internet licensing simplifies license management, making it easy to remotely add, upgrade, or update licenses.

  4. Improved responsiveness of application graphs, summary tables, and settings pages

  5. Improved timezone handling in the user interface, databases, and JSON transcripts. Date and time now stored in UTC in JSON transcripts.

  6. Expanded callback support to include SFTP and Amazon S3, providing the ability to automatically send Text and JSON transcripts and MP3 audio to a URL

  7. Additional ASR tags can now be configured in the V‑Spark UI

  8. Redesigned date range pickers for improved performance and usability

  9. Icon added to Settings pages to enable direct navigation to Dashboard view of applications and folders

  10. Simplified folder hierarchy is retained in downloaded zip files, preventing filename collisions

  11. Increased usability of Activity log: Filter settings are stored as query parameters to enable sharing URLs with other users, and "API" type added to filter log entries related to API calls

  12. System administrators now have the ability to restrict users from downloading MP3, JSON, and text transcriptions from the File Details page.

  13. Support for Red Hat Enterprise Linux (RHEL) 6

  14. Approximate search mode is no longer supported

  15. Fixes

    1. Corrected and improved metadata handling in Voci JSON output. The intermediate <metadata> section was removed from the <client_data> section to flatten the JSON and simplify client data processing. See the Voci JSON Output Format Guide for more information.

    2. Application queries using ~n are now scored correctly even if a turn contains one of the phrase words multiple times

Known Issues in 3.2

The items that are described in this section may affect your use of V‑Spark, and are recognized by Voci as items that we expect to address in a future release.

  1. Dashboard Displays

    1. Stat summaries for dates in the past are not updated when files with historical timestamps older than the current day are loaded

    2. Files uploaded to a folder that is linked to a disabled application or that have not been reprocessed will still appear in the Files View on the application dashboard

  2. Unlinking an application from a large folder (greater than 100,000 files) may cause errors where the folder will be unselectable in file editing mode, and where the app is unlinked, but scores may still be visible in certain views

  3. Search

    1. Time-based search

      1. Searches are turn-based and may return results for phrases that are outside of the requested timeframe, if they are within the same speaker turn

      2. Ad-hoc Search snippets in the Files View may display results outside the specified time interval

    2. The (-) operator for excluding a search tag in ad-hoc search is not supported for file names

    3. Exceptions to standard File Details Highlighting (in order of severity)

      1. Searches using (~n) may highlight phrases with the specified number of words in between each of the search terms, rather than between all of the search terms. For example, "get help phone ~4" may highlight phrases with 4 words between "get" and "help" and another 4 words between "help" and "phone". Searches using (~t) and (~n) will only be highlighted if the words matched appear in the order specified in the search tag.

      2. Highlighting may be incomplete for matches marked with emotion

      3. Highlighting may be incomplete for matches with intervening silence tags greater than 2 seconds

      4. Matches for "@*" (such as when searching for email addresses) will only be highlighted until the first "." in the email address. Regular expression searches may be used to circumvent this issue.

      5. Excluded application search phrases may be highlighted on the File Details page. Note that these exceptions only affect highlighting, and that all search results using the above characters will still be accurate.

  4. Other

    1. The Trending View may take some time to fully load, especially for longer date ranges

    2. Users playing longer audio files via Firefox may encounter playback/syncing issues when jumping to different sections of the transcript. This is an external issue caused by the Firefox browser. Users who encounter this issue can simply wait for the browser to finish buffering the file before resuming their search.

  5. The transcribe API S3 option can currently only be used in cloud mode