PROFESSIONAL SERVICES
Engineering services for broadcast, OTT and post-production, from signal analysis to delivery
Beyond our products, we support broadcasters, streaming platforms, post-production houses and technology vendors with hands-on engineering. Every service builds on the same expertise: separating noise from real image content, measuring quality honestly and running it efficiently on GPUs.
Video Quality Audits and Encoder Shootouts
Independent encoder comparisons and bitrate ladder optimization with VMAF, PSNR and QBVE. Averages hide failures, so we also count frames below quality floors and measure the stability of your streams. Measuring quality metrics needs a lot of compute power, which we provide as well. In the design phase of a new workflow you know encoding speed and quality in advance, and in a vendor shootout you get reliable numbers.
- VMAF, PSNR and QBVE analyses
- Encoder and vendor shootouts
- Bitrate ladder optimization
- End-to-end tests with test sequences for resolution and dynamic range
Encoding Pipeline Design
Per-shot encoding driven by IRIS.ANALYST, built on FFmpeg and industry-standard tools, without proprietary encoders. Output for HLS and DASH in H.264, HEVC and AV1, including AV1 film grain synthesis with our own grain tables. Fully automated with Kubernetes, Docker or VMs.
- Per-shot and adaptive encoding
- HLS and DASH delivery
- AV1 with film grain synthesis
- Automation on Kubernetes, Docker and VMs
Pre-Postproduction, Mastering and Restoration
Automated denoising before the grade: BRAW, RAW and any FFmpeg-readable format are processed overnight into a lossless FFV1 master (12 or 16 bit) plus a ProRes proxy for editing. Colorists start in the morning with a clean signal, even for 8K and 12K HDR. The same technology cleans archive material and film scans for restoration.
- IRIS.RAW and BRAW2IRIS
- FFV1 masters and ProRes proxies
- Overnight processing on-premises or in the cloud
- Archive and film restoration
Software-Defined Video: Media eXchange Layer (MXL) and Dynamic Media Facility (DMF)
We integrate video processing as MXL Media Functions in containers and help plan software-defined live production: GPU requirements, memory exchange, redundancy with Kubernetes and a hybrid cloud model that keeps the base load local and scales peaks into the cloud.
- MXL Media Functions in Docker and Kubernetes
- GPU and memory architecture
- Hybrid cloud and cost planning
- Requirements for the MXL standard, see IRIS, MXL & DMF
GPU, CUDA and AI Development
Years of CUDA experience in real-time image processing: algorithm development, image processing libraries, deep learning for video analytics and AI-based media functions for live broadcast pipelines.
- Project management and implementation
- CUDA and image processing libraries
- Algorithm development
- Deep learning for video analytics and clean AI training data
FFmpeg Training and Consultancy
FFmpeg is the open-source engine behind most of the largest cloud encoding farms in the world and an alternative to commercial transcoding products. We offer hands-on training and consultancy.
- Quality metrics: VMAF, PSNR and QBVE
- Optimizing your bitrate ladder and per-shot encoding
- FFmpeg automation on Linux
- Speeding up adaptive bitrate encoding
- Analyzing files with MediaInfo, Bitrate Viewer and Apple's AVQT
- FFV1 as intermediate format
Camera, Scanner and Film Analysis
Vendor-independent 35mm film test strips and test chart generation, plus the analysis of cameras, film scanners and scanned material. Sensor performance changes over time: noise levels rise, dead or hot pixels appear. The earlier you detect quality issues, the less time and money you lose.
- 35mm film test strips and test chart generation
- Dead and hot pixel detection
- Noise floor and image quality analyses
- Workflow analyses
Broadcast and Film Projects
From requirements to rollout for digital TV and film: we support you with project management, implementation, system analyses and RFQs.
- Project management and implementation
- System analyses and RFQs
- Test chart generation and image quality analyses
Tell us about your project: info@waveletbeam.com