https://github.com/x42/harvid

http ardour video daemon
https://github.com/x42/harvid

ardour ffmpeg video

Last synced: 4 months ago
JSON representation

http ardour video daemon

• Host: GitHub
• URL: https://github.com/x42/harvid
• Owner: x42
• License: gpl-2.0
• Created: 2013-01-30T23:24:27.000Z (about 12 years ago)
• Default Branch: master
• Last Pushed: 2022-12-31T21:05:07.000Z (about 2 years ago)
• Last Synced: 2023-04-12T07:00:04.354Z (almost 2 years ago)
• Topics: ardour, ffmpeg, video
• Language: C
• Homepage: https://x42.github.io/harvid/
• Size: 184 MB
• Stars: 16
• Watchers: 4
• Forks: 5
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • Changelog: ChangeLog
  • License: COPYING

Awesome Lists containing this project

README

        
harvid -- HTTP Ardour Video Daemon

==================================

Harvid decodes still images from movie files and serves them via HTTP.

Its intended use-case is to efficiently provide frame-accurate data and

act as second level cache for rendering the video-timeline in

[Ardour](http://ardour.org).

Download

--------

Apart from the source-code and packages from your linux-distributor, binaries

are available for OSX, Windows and Linux at http://x42.github.com/harvid/ .

Usage

-----

Harvid is a standalone HTTP server, all interaction takes place via HTTP.

After launching it, simply point a web-browser at http://localhost:1554/

The OSX bundle and window installer come with a shortcut link to launch

the server. On Linux or with the OSX package, harvid is usually started

from a terminal by simply typing `harvid`<enter>.

Harvid can also be run directly from the source folder without installing

it. Get its build-dependencies (see below), run

	make

	./src/harvid

When used from ardour, ardour will automatically start the server when

you open a video. Ardour searches $PATH or asks your for where it can find

harvid. The easiest way is to simply run:

	sudo make install

Harvid can be launched as system-service (daemonized, chroot, chuid, syslog),

and listen on specific interfaces only in case you do not want to expose

access to your movie-collection. However, is no per request access control.

For available options see `harvid --help` or the included man page which

is also available online at http://x42.github.com/harvid/harvid.1.html

Build-dependencies

------------------

[ffmpeg](http://ffmpeg.org/) is used to decode the movie. The source

code should be compatible and compile with [libav](https://libav.org/).

For encoding images,

[libpng](http://www.libpng.org/pub/png/libpng.html)

and [libjpeg](http://libjpeg.sourceforge.net/) are required.

Internals

---------

Harvid is highly concurrent makes use of all available CPUs. It will

spawn multiple decoder processes, keep them available for a reasonable

time and also cache the video-decoder's output for recurring requests.

The cache-size is variable only limited by available memory.

All images are served from the cache, so even if you are not planning

to use the built-in frame cache, the cache-size defines the minimum

number of concurrent connections.

Interface

---------

The HTTP request interface is documented on the homepage of the server

itself: http://localhost:1554/

The default request-handler will respond to `/?file=PATH&frame=NUMBER`

requests. Optionally `&w=NUM` and `&h=NUM` can be used to alter the geometry

and `&format=FMT` to request specific pixel-formats and/or encodings.

`/index[/PATH]` allows to get a list of available files - either as tree or

as flat-list with the ?flatindex=1 as recursive list of the server's docroot.

`/info?file=PATH` returns information about the video-file.

Furthermore there are built-in request handlers for status-information,

server-version and configuration as well as admin-tasks such as flushing

the cache or closing decoders.

The `&format=FMT` also applies for information requests with

HTML, JSON, CSV and plain text as available formatting options.