Vesper

This release installs correctly on Apple silicon Macs, fixing issue 217. For other platforms this release is functionally equivalent to 0.4.13.

This release mainly improves support for cloud archives, as part of an effort to improve support for collaborative monitoring.

In particular, this release:

* Supports running the Vesper server using Docker Compose, either locally or using AWS ECS. See the Docker Compose files in the root Vesper project directory for examples.

* Supports clip audio files stored in an AWS S3 bucket. A future Vesper version will also support cloud storage of recording audio files.

* Adds a `health-check` Django view for efficient health checks, e.g. by an AWS load balancer.

* Allows (and encourages) configuration via environment variables rather than changes to the `vesper.django.project.settings` module. See the `vesper.django.project.settings` module for details.

* Allows exclusion of TensorFlow detector and classifier extensions from the Vesper server via the new `VESPER_INCLUDE_TENSORFLOW_PROCESSORS` environment variable and Django setting.

* Introduces *detector provider* plugins as a first step towards moving most or all detectors and classifiers out of the core Vesper server and into plugins.

* Makes the clip metadata CSV file delimiter configurable via the `delimiter` setting of the `Clip Table Format` preset type. The setting is optional and defaults to a comma. The `Raven Selections` preset in the Vesper archive template shows how to use the new setting to export a Raven selection table from Vesper.

* Modifies clip, recording, and recording file end time measurement values. The values are now one sample period later than they were in Vesper 0.4.12, since the end time is now considered to be the time of the first sample *after* a clip, recording, or recording file rather than the time of its last sample.

* Adds a Django view (`vesper.django.old_bird.CreateLrgvClipsView`) that creates clips for the Old Bird LRGV monitoring project.

* Uses the new Python Standard Library `zoneinfo` module for time zone computations instead of the older, third-party `pytz` module.

* Fixes issue 211 concerning missing annotation creating job and creating processor fields.

[](https://zenodo.org/badge/latestdoi/19112486)

This release fixes a broken documentation link in v0.4.11, released earlier today.

Relative to v0.4.10, this release:

* Adds the `Export clip counts by classification to CSV file` and `Export clip counts by tag to CSV file` commands. These commands export clip counts for an entire archive to a CSV file. You can open the exported CSV file in your favorite spreadsheet program to look at clip count pivot tables, charts, etc.

* Adds tag support to the `Export clip metadata to CSV file` command. You can include a column in the CSV file for any tag, and specify the strings to use in the column to indicate the tagged/untagged status of a clip. See the `Basic` `Clip Table Format` preset in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for an example.

* Adds support for clip padding to the `Export clips to audio files` command. The new `Clip Audio File Export Settings` preset type allows specification of the time interval to be exported. The interval can be specified in terms of either total duration or left and right clip padding.

* Adds support for clip padding to the `Export clips to HDF files` command. The new `Clip HDF5 File Export Settings` preset type allows specification of the time interval to be exported. The interval can be specified in terms of either total duration or left and right clip padding.

* Adds `Tag clips` and `Untag clips` commands. These commands tag and untag sets of clips, including randomly selected subsets of specified sets. You can use these commands to mark sets of clips for review or export, for example.

* Improves detection schedule compilation efficiency.

* Adds the `Create random clips` command. This command creates a specified number of clips at random in a set of recordings. This is useful for sampling recordings, for example for review or to obtain noise clips for detector or classifier training.

* Improves the `Transfer clip classifications` command. The command no longer requires configuration information for a specific detector in order to work with its clips. This allows the command to work out of the box with all detectors, both existing and future.

* Adds a _page auto-advance_ feature to clip albums. Page auto-advance displays the pages of a clip album one after the other automatically, so you don't have to navigate from one page to the next manually. You start and stop page auto-advance and control the rate of advance via the keyboard. See the any of the `Clip Album Commands` presets in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for key binding examples.

* Adds a _clip auto-advance_ feature to clip albums. Clip auto-advance selects the clips of a clip album one after the other automatically, playing each clip as it is selected. The feature also navigates automatically from one clip album page to the next as needed to display the currently selected clip. You start and stop clip auto-advance and control the pause between consecutive clip playbacks via the keyboard. See the any of the `Clip Album Commands` presets in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for key binding examples.

* Streamlines clip album database queries, improving responsiveness.

* Includes several UI improvements. These include (finally) a favicon, thanks to RichardLitt!

* Updates Bootstrap from version 3.3.6 to version 5.1.1.

See the [Vesper documentation](https://vesper.readthedocs.io/en/latest/index.html) for [installation instructions](https://vesper.readthedocs.io/en/latest/installation.html).

[](https://doi.org/10.5281/zenodo.6525072)

This release:

* Adds the `Export clip counts by classification to CSV file` and `Export clip counts by tag to CSV file` commands. These commands export clip counts for an entire archive to a CSV file. You can open the exported CSV file in your favorite spreadsheet program to look at clip count pivot tables, charts, etc.

* Adds tag support to the `Export clip metadata to CSV file` command. You can include a column in the CSV file for any tag, and specify the strings to use in the column to indicate the tagged/untagged status of a clip. See the `Basic` `Clip Table Format` preset in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for an example.

* Adds support for clip padding to the `Export clips to audio files` command. The new `Clip Audio File Export Settings` preset type allows specification of the time interval to be exported. The interval can be specified in terms of either total duration or left and right clip padding.

* Adds support for clip padding to the `Export clips to HDF files` command. The new `Clip HDF5 File Export Settings` preset type allows specification of the time interval to be exported. The interval can be specified in terms of either total duration or left and right clip padding.

* Adds `Tag clips` and `Untag clips` commands. These commands tag and untag sets of clips, including randomly selected subsets of specified sets. You can use these commands to mark sets of clips for review or export, for example.

* Improves detection schedule compilation efficiency.

* Adds the `Create random clips` command. This command creates a specified number of clips at random in a set of recordings. This is useful for sampling recordings, for example for review or to obtain noise clips for detector or classifier training.

* Improves the `Transfer clip classifications` command. The command no longer requires configuration information for a specific detector in order to work with its clips. This allows the command to work out of the box with all detectors, both existing and future.

* Adds a _page auto-advance_ feature to clip albums. Page auto-advance displays the pages of a clip album one after the other automatically, so you don't have to navigate from one page to the next manually. You start and stop page auto-advance and control the rate of advance via the keyboard. See the any of the `Clip Album Commands` presets in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for key binding examples.

* Adds a _clip auto-advance_ feature to clip albums. Clip auto-advance selects the clips of a clip album one after the other automatically, playing each clip as it is selected. The feature also navigates automatically from one clip album page to the next as needed to display the currently selected clip. You start and stop clip auto-advance and control the pause between consecutive clip playbacks via the keyboard. See the any of the `Clip Album Commands` presets in the [Vesper archive template](https://www.dropbox.com/s/ylscsoyttnu1qxk/Archive%20Template.zip?dl=0) for key binding examples.

* Streamlines clip album database queries, improving responsiveness.

* Includes several UI improvements. These include (finally) a favicon, thanks to RichardLitt!

* Updates Bootstrap from version 3.3.6 to version 5.1.1.

See the [Vesper documentation](https://vesper.readthedocs.io/en/latest/index.html) for [installation instructions](https://vesper.readthedocs.io/en/latest/installation.html).

[](https://doi.org/10.5281/zenodo.6524961)

This release:

* Adds support for clip tags, allowing users to mark arbitrary sets of clips for display and processing.

* Adds clip count tooltips to the dates of a clip calendar.

* Adds page tooltips to the rug plot of a night clip album.

* Includes the clip album page number and preset names in the clip album URL.

* Improves interaction between clip albums and the browser history, for example so that each viewed clip album page has its own history entry.

* Adds new versions 1.1 of the MPG Ranch NFC detectors and coarse classifier. The new versions use TensorFlow 2.x instead of TensorFlow 1.x.

* Improves cross site request forgery (CSRF) protection.

* Provides various small UI improvements and bug fixes.

See the [Vesper documentation](https://vesper.readthedocs.io/en/latest/index.html) for [installation instructions](https://vesper.readthedocs.io/en/latest/installation.html).

[](https://doi.org/10.5281/zenodo.5274229)

This release:

* Adds support for recent versions of [BirdVoxDetect](https://github.com/BirdVox/birdvoxdetect). Note that BirdVoxDetect is no longer included with Vesper. Instead, it must be installed separately in its own Conda environment, along with the [vesper-birdvox](https://github.com/HaroldMills/vesper-birdvox) package. There are three advantages to this:

* Vesper can more readily support new versions of BirdVoxDetect as they appear.
* The potential for dependency conflicts between Vesper and BirdVoxDetect is eliminated.
* A single Vesper server can be used with multiple versions of BirdVoxDetect.

See the [Vesper FAQ](https://vesper.readthedocs.io/en/latest/faq.html#how-do-i-use-birdvoxdetect-with-vesper) for how to use BirdVoxDetect with Vesper.

* Adds clip table format presets for use with the `File->Export clip metadata to CSV file` command. See the presets in the [Vesper tutorial](https://vesper.readthedocs.io/en/latest/tutorial.html) for examples. The default table format, the `Basic` preset of the tutorial, is more generic than the old default, which was specifically for MPG Ranch use. The old default is still available as the `MPG Ranch Original` preset of the tutorial.

* Adds several new clip measurements and value formatters for use with the `File->Export clip metadata to CSV file` command. Some of the new measurements and formatters are used by the new default table format.

* Uses [Skyfield](https://rhodesmill.org/skyfield/) rather than [PyEphem](https://rhodesmill.org/pyephem/) to compute solar event (e.g. sunrise and sunset) times.

* Includes classifications in exported clip audio file names.

* Improves navigation among night clip albums, for example with `Go to next date` and `Go to previous date` pulldown items that navigate to adjacent dates while preserving preset changes.

* Provides other UI improvements and bug fixes.

See the [Vesper documentation](https://vesper.readthedocs.io/en/latest/index.html) for [installation instructions](https://vesper.readthedocs.io/en/latest/installation.html).

[](https://doi.org/10.5281/zenodo.4677147)