Dark Energy
Particle Physics Unleashed
Dark Energy is the motion picture industry’s go-to toolset for Grain/Noise Reduction and Texture Management. And, as a premiere spatial processing platform, Dark Energy produces impeccable Deinterlaced Video and cranks out theatrical-quality Up-Res Conversions.
Its uncanny ability to remove only the noise, but not the detail, means that Dark Energy is also a great solution for high-end Pre-Encode Optimizations. With Dark Energy, you get better looking streams with fewer bits.
Restore. Optimize. Texture.
Dark Energy Professional has been described as incredible, amazing, and unequaled for image texture management by the most discerning eyes in the world.
Starting with an intuitive, real‐time feedback GUI, users can immediately see the results of their adjustments without waiting for a render, without “hoping” the adjustments will be what they want.
Dark Energy Accelerator is Cinnafilm’s batch/automated rendering tool that allows users to:
- Background render already completed projects while starting work on a new project
- Batch process complex folder structures based on pre‐defined templates
- Combining complete control of image texture and incredible rendering speeds, Dark Energy Professional is the ultimate software for demanding users
- Shatter real-time performance with 20-24 FPS for each GPU Accelerator calls
Features
- High-quality image scaling (DE handles any size)
- Automatic dustbusting, 80% hands-free without artifacts
- Premium denoise/degrain without image degradation
- Texture simulation to emulate any film stock or texture – No more film-outs with the best looking film grain on the market
- Homogenize the texture from many sources (film, various cameras, etc.)
- Fully scalable render speeds – no limits
- The fastest, most precise image texture management software in the world for cinematic releases and archive restoration
- Temporal Noise Reduction
- Spatial noise reduction with mix-back function
- Image sharpening
- Logarithmic transfer space
- Gamma adjustment
Performance
Dark Energy Professional has always been the fastest, most accurate degrain technology in the universe. With the 5.0 release and processing on Nvidia’s latest Tesla GPUs, we have expanded that lead to unbelievable lengths.
While other platforms struggle with slow processing, Dark Energy Professional renders as fast as your workflow demands. Dark Energy Professional is 100% GPU-based, providing users 20-24 FPS of rendering for each Tesla GPU installed in your properly equipped, Windows workstation (for 2K images).
Performance for Dark Energy is heavily dependent on two items: GPU speed and storage speed. Cinnafilm only supports GPUs that are Kepler Class or newer, and we prefer our clients use GPUs that are Tesla Compute Cluster (TCC) capable. Storage must be fast enough to simultaneously read/write uncompressed image files at the chosen bit depth and frame rate. Typically for UHD/4K, that means over 1.2 gigabytes of simultaneous read/write throughput to the drive for each GPU in use. HD/2K need 600 megabytes of simultaneous read/write throughput..
A typical Dark Energy Professional workflow is as follows:
Denoise the image, sharpen the image, resize/scale the image and add image texture (film grain). With this typical workflow, the performance of Dark Energy Professional V is as follows:
Render Speed – FPS, Single GPU
| Resolution | Kepler K40 | Maxwell M60 | Pascal GP100 |
|---|---|---|---|
| 2k, 10bit DPX | 12 | 18 | 24 |
| 4k, 10bit DPX | 6 | 8 | 16 |
Versions
Dark Energy is available as standalone software and a plug-in for some of today’s most popular transcoding platforms (Telestream Vantage, Imagine SelenioFlex, Harmonic ProMedia Carbon, HS-ART Diamant, Root6 Content Agent).
Dark Energy for After Effects was discontinued in 2015 and we continue to support our current customers (directions here: Dark Energy for After Effects).
If you want to get the latest version of Dark Energy as well as access to the full suite of Cinnafilm solutions in the cloud, be sure check out PixelStrings by Cinnafilm.