veejay 1.6.0 is released!

[game-stop]

Release 1.6.0: Architecture Refactor, GTK3 Overhaul, & Dynamic Playback

It’s been a minute since the last release, but 1.6.0 is finally out in the wild. This one brings massive architectural refactoring, heavily optimizing the core media processing pipelines, and delivering a complete overhaul of the desktop user interface.

The client UI is entirely rebuilt utilizing GTK3, custom CSS stylesheets, and Glade interface definitions to modernize the workspace. Under the hood, the engine now leverages C11 atomic operations and OpenMP parallelization to accelerate real-time YUV video effect filters and memory copy operations, backed by newly vectorized SIMD routines.

Core Additions and Feature Integrations

• Dynamic Framerates & Audio Trickplay: VeeJay can now play at completely dynamic framerates. We’ve wired up audio trickplay so when you alter the playback speed, the engine pitches the audio down (or up) just like a physical cassette player or tape deck.
• Expanded Codec Support: We’re no longer strictly bound to MJPEG. The engine now supports rendering H.264 and other non-intraframe codecs directly, alongside improved V4L2 capture interface and video loopback device handling.
• Hardware Acceleration (CUDA / VA-API / Vulkan): For the hardware junkies, we've got direct hardware encoding and decoding of MJPEG via the nvjpeg backend (if the NVIDIA CUDA Toolkit is present). We’ve also tied in VA-API and Vulkan support through the FFmpeg backend for serious GPU offloading.
• Massive FX Library & Ecosystem: We've packed in countless new video effects. We're also introducing VeeJay Director for multi-instance venue routing and projector mapping, alongside VeeJay Eidolon, an experimental Auto-VJ that dynamically mutates FX chains based on your real-time frame budget and live feedback.
• NDI Support: Full Network Device Interface integration for frame-accurate, low-latency streaming right across your local network (requires the NDI SDK).
• JACK Audio & Sync: Total audio integration via low-latency JACK ringbuffers, coupled with a brand-new audio beat detection system to drive your visual parameters directly from the tempo.
• Advanced Sequencing & Control: A new MIDI patch bay, an 8-track Pattern Editor for frame-accurate VIMS automation, and a massive 4-bank, 120-slot 4 Grid Sequence Editor for immediate sample triggering.

Binary Release Targets

Packaging this beast took some work, so we’ve overhauled our Linux binary delivery to support a comprehensive architecture matrix. Check out RELEASE_TARGETS.md for the full deployment details.

• The automated GitHub workflow now builds 10 standard suites across Debian, Fedora, and Arch, alongside 10 explicitly enabled -nvjpeg companion suites.
• We know how crucial local optimization is, so we're shipping compiler-validated targets like amd64-x86-64-v3 that aggressively use AVX2 across the board.
• The --with-arch-target=generic flag is still the redistributable default; on x86-64, it defines baseline paths while seamlessly relying on runtime dispatch for isolated AVX2 routines.