A major regression slipped into the 0.9.4 release plus some other good fixes rolled in just afer that release prompting this new release. Please discontinue using version 0.9.4 and upgrade.(Mon, 02 Mar 2015 07:18:00 +0000)
Ushodaya Enterprises Limited of India, the original sponsor of the MLT and Melted projects has assigned their copyrights to Daniel R. Dennedy, sole proprietor and CTO of Meltytech, LLC. What does that mean? Not much. There are no planned changes to project direction at this time, but it may provide some options in the future. It is just good to have the copyrights under control of active members of the project to make it possible to exercise future options. Many thanks to project co-founder, BGa, for facilitating this transaction.(Tue, 04 Nov 2014 01:32:00 +0000)
Synfig Studio is a good, free, open source, cross-platform 2D animation program. They wanted to add audio support, but they needed more than just a hardware abstraction layer for audio output. They needed more than a multi-format/-codec library. They also wanted something that provides timing and mixing with a succinct API. They chose MLT.
There are a few new filters in MLT git. The "videostab" filter is a new version based vid.stab replacing the old videostab2 filter. Because it uses an externally compiled version of vid.stab, you can use the latest and greatest version of vid.stab, which is not only faster but gives better quality results. In addition, while vidstab is a dual pass filter - first pass generates a .stab file, second pass reads it - there is a new single-pass "deshake" video filter that uses vid.stab's single-pass mode. It will not give as good results as dual pass, but it is handy for realtime/live applications. Thanks go to Jakub Ksiezniak, a new MLT contributor, and Brian Matherly for helping to get it into shape for merging. Also, Brian contributed a new audio filter named "loudness" that uses libebur128 to provide better audio normalization because EBU R128 is based on perceived loudness and newer research than traditional RMS- and peak-based approaches.(Wed, 12 Feb 2014 04:18:00 +0000)
Earlier in the year we announced the introduction of OpenGL-based image processing on the GPU. However, now it is finally available in a release version of the software. Not only that, but we have also added a new, properly integrated property animation API to replace the limited, awkward mlt_geometry API. Here are the features of the new API:
The time of a keyframe can be set using frame number, time clock string, or a timecode string.
A negative time value makes it relative to the end of an object's duration.
Keyframes support 3 forms of interpolation: discrete (no interpolation), linear, and a smooth Catmull-Rom spline.
Supports a new mlt_rect propery type in addition to string (discrete only), integer, and floating point.
API for most apps is just a few new, simple methods on the Properties object.
Integrated with the OpenGL-based effects only at this time to ease the migration - will be extended to other plugins in the next version or two.
Above is a plot of the curve described by the following property value. 0|=50; 50|=100; 100=200; 200~=60; -7:00~=180; -2:00~=100; -1=220 This shows that it is possible to combine a number of things in the same curve. It starts out with discrete keyframes (|), switches to a linear interpolated keyframe (=), and ends with a few smooth spline keyframes (~) set using timecode relative to the end/width (-).
Here are other highlights of the release:
Improved pause behavior when using buffered rendering in mlt_consumer.
Added mlt_color type.
Deprecated mlt_geometry API.
Support for the latest versions of FFmpeg and Libav (but dropping support for 0.5 and 0.6 versions).
Added alpha channel output to avformat consumer.
Added reconnect and exit_on_disconnect properties to avformat producer.
Added qglsl consumer to use opengl with avformat, sdi, and decklink.
For a few months, we have been working on a set of image processing filters and transitions that utilize the OpenGL Shader Language (GLSL) of your video card/chip. This means that not only are they very fast but also cross-platform and widely supported. GLSL is generally faster and better supported than OpenCL and NVIDIA CUDA because those technologies are rather new or vendor-specific and designed for more general purpose computing. The new filters are also high quality by using 16-bit linear floating point per color component. All of this means that you can apply more effects without a huge reduction in frame rate.
This work has now been merged into the master code branch and will be available in the next release. They have already been integrated into the Shotcut daily builds for all 3 major OS platforms! However, a caveat - there is not much to see there yet because Shotcut does not yet have filters. Nevertheless, it has served as a feasibility exercise of the cross-platform integration. Meanwhile, one can open a clip, save as XML, edit the XML to put something like <filter mlt_service="movit.mirror"/> inside of the <producer></producer>, open the XML in Shotcut, and see the result. But first you will need to enable GPU Processing in the Settings menu.
What is currently available?
movit.convert (colorspace conversion)
movit.lift_gamma_gain (color correction)
movit.rect (position and scale)
qglsl (a wrapper to multi consumer that uses Qt to abstract platform-specific OpenGL context)