The Pixel Farm Pfclean 3.0r4
The Pixel Farm Pfclean 3.0r4 | 43.30 MB
PFClean operates flexibly in an ever changing and complex DI environment, working as a fully automatic, hybrid system or manual system. The user is able to define frames and the type of anomalies to be fixed in a non destructive environment. Image sequences can then be processed immediately or committed to the Pixel Farm’s tightly integrated cross platform batch processing system.
All fixes, whether applied automatically or manually, are executed via a per frame meta data map. Fixes, right down to individual paint strokes, are stored as meta data for later modification if required. This provides total control over any image modification and effectively provides the user with an unlimited number of non-sequential undos. The data is maintained throughout the clean up process with additionl annotation tools allowing further information to be added to inform processes further down the DI chain.
Production Director at Ollin Studios, Charlie Iturriago explains, “The PFClean team receive the movie rolls of 1000ft to the computer and remove scratches and dust, cleaning the film frame by frame, in a fast and automated process. Then the final frames go to output servers for deliveries and archival. So, the entire movie is QC’d with PFClean in 2k for a completely clean delivery.”
Comprehensive DI and RestorationToolset
PFClean has all the tools required to address the workflow from scanners, archival footage and digital film cameras into post production and finishing including:
Automatic dirt, Dust and Scratch detection and fix on any combination of color channels
A real-time float Paint system including clone and dust-busting brushes
Image stabilization including frame rebuilding.
Rotoscope tools to assist paint and motion analysis.
De-flicker.
Global de-noise/sharpen/blur.
Noise/Grain adjustment.
Full float pipeline with HDR support.
Full support for defect maps from IR scanning data.
The PFClean Film Restoration Pack* adds contains a number of additional tools including:
Tear repair
Frame repair
Automatic RGB channel alignment
Colour balance
Colour grade transfer
Colour correction
De-warp
PFClean * Film Restoration Pack
The additional tools provided by the purchasable restoration pack are design to extend PFClean’s core functionality to address problems found in very damaged film and with achieve restoration projects. The tools have been design to fit seamlessly into the PFClean environment and operate using the benefits of the PFClean non destructive workflow. For a modest sum you can add these state of the art tools to PFClean with a simple addition to your license.
*The PFClean Film Restoration Pack is available as an option for £3000, €4500, $5500.
PFClean Video IO Module
An option on the turn key system, the Video IO modules provides the means to ingest both SD and HD video direct from a deck onto fast internal storage from withn the PFClean UI. PFClean is able to work with both progressive and field based footage with full format and aspect ration conversion for lay off.
*The PFClean Video IO module is available as an option for £5000, €7500, $1000.
Built for speed as well as comfort
PFClean is the first application to offer a fully hardware accelerated toolset in this market. Using totally new massively parallised algorithms, processes have been ported to use the GPU rather than the CPU. Using this ’stream computing’ methodology PFClean offers unrivalled performance in a cost effective turn key system. The fully 64-bit application architecture is built around highly optimised and efficient motion analysis algorithms. The software is fully scalable across multiple CPU’s allowing processing power to grow with user or project requirements. The addition of a batch or distributed processing system and a meta data architecture greatly simplifies system integration. The entire application is built on The Pixel Farm’s own playback architecture allowing real-time 2K playback on any hardware with suitable disk IO.
The GUI can be used to review clips and assess what parameters the batch processing system will require to complete a job. A batch script can then be left to run as files are released from the scanner. As frames become available to the GUI, fully automating the clean up process. If required, the batch applied fixes can be adjusted, modified or rejected in near real-time using the meta data tags created for each fix. The addition of extra information such as the IR scan from the scanner where available, will greatly increase the possibility of fixing the errors with little or no user interaction.
Fast Non-linear Workflow
A non-linear workflow allows the user to go back at any stage and edit the sequences without the need for multiple renders. By making use of meta data every action, automatic or manual, is recorded and made available for editing in a fully non destructive environment. The application interface and toolset is optimised to provide an efficient as well as effective working environment, making PFClean by far the fastest way to get images from the scanner into the DI pipeline.
DI Manager at Deluxe London, Marcus Alexander explains, “the quality of PFClean’s automated dust busting workwas great for ‘Atonement’, I had to only get a grand total of 8frames manually re-fixed for the entire feature!”
Tools for Post
PFClean is also an ideal tool for wire/rig removal and other common paint and effect tasks within the post-production environment.
Key Application Features:
Multi-threaded motion analysis, including paint-based motion vector editing.
Manual, Automatic and hybrid error detection and removal (dust and scratches).
Automatic and manual ‘repair’ modes.
DPX and Keykode film handling.
Optional Video IO.
Proxy export to editing systems, with DPX and Keykode data embedded.
Non-linear work flow.
Non destructive image editing.
3D LUT support and open API.
Multi resolution mastering and format conversion.
Fully scriptable.
Fully optimised for Mac Intel platform.
Full 64-bit implementation**
** 64 bit Hardware and OS required.
Download The Pixel Farm Pfclean 3.0r4
0 comments:
Post a Comment