Deadline scheduler
Maps each RTP presentation time to a controlled delivery deadline.
Patent-pending userspace AES67 architecture
High-density, deterministic AES67 for embedded Linux.More audio.
Same hardware.
Measured on a quad-core Arm Cortex-A72: a 1 ms configured receiver target through physical PCM5122/I²S output, plus sustainable density of 14 stereo streams at the same receiver profile.
PHYSICAL OUTPUT + RECEIVER DENSITYMEASURED
RTP/L16→DEADLINE PATH→ALSA MMAP→PCM5122
REAL DAC PATH1 msconfigured receiver target sustained to PCM5122/I²S
PCM-CHECKED DENSITY14stereo streams · 28 channels · same target
COMMON 3 MS+65%33 vs 20 stereo streams
COMMON 5 MS+67%40 vs 24 stereo streams
SAME HWhigher repeatable density than the tested stock direct-mmap RAVENNA/Linux implementation
1 msconfigured target through physical PCM5122/I²S output
14 / 28stereo streams / channels at 1 ms
33 / 66stereo streams / channels at 3 ms
40 / 80stereo streams / channels at 5 ms
Reference platform: Raspberry Pi 4, quad-core Arm Cortex-A72, software PTP/software timestamps. Receiver-software timing measurements; the 1 ms physical-output result is not an analog end-to-end latency claim.
EXPLORE THE PLATFORM
Technical depth when your team needs it.
The overview remains concise. Dedicated pages give OEM engineering, product and commercial teams the evidence relevant to their decisions.
01TECHNOLOGY
Linux AES67 architecture
See how network time, deadline scheduling and device-clock control converge in userspace.
Explore the receive path 02BENCHMARKSCanonical measured proof
Compare physical output, embedded receiver density and x86 resource use without mixing measurement boundaries.
Review the evidence 03OEM INTEGRATIONFrom target SoC to product
Move from a target-platform benchmark through feasibility, integration and production licensing.
Plan an evaluation 04REFERENCE PROJECTStagePi stagebox
See the open-source, off-the-shelf physical output reference without confusing it with the commercial IP.
See the reference path01 / OEM VALUEMeasured efficiency.
Measured efficiency.
Commercial leverage.
For AoIP platform vendors, professional-audio OEMs and semiconductor companies, Bama Box turns a Linux data-path bottleneck into higher channel density, lower-latency product profiles and reduced resource demands—from embedded endpoints and stageboxes to mixers, console engines and broadcast systems. The same userspace receiver architecture scales from Linux SoCs such as NXP i.MX to x86 backstage systems.
COMMERCIAL LEVERAGE
Increase channel densityReach lower-latency profilesCreate headroom on the existing SoCAvoid unnecessary processor escalation
Measured at a strict 1 ms receiver target through physical PCM5122/I²S output, with a maximum runtime ALSA queue of 34 frames (0.71 ms).
INTEGRATION ADVANTAGE
Patent-pending architectureUserspace-only LinuxNo custom kernel module
Integrate proven technology without rebuilding the data path from scratch.
02 / THE TECHNOLOGYMove audio.
Move audio.
Not overhead.
Bama Box implements packet intake, jitter buffering and deadline-controlled scheduling entirely in Linux userspace. It requires no custom kernel module and is portable across Linux-capable SoCs. Platform-specific hardware timestamps, clocks and audio interfaces connect through a defined software API.
CONVENTIONAL LINUX PATH
NETWORKOVERHEADBUFFERSOVERHEADAUDIO STACK
System cost rises quickly as channel count grows.
BAMA BOX USERSPACE OPTIMIZED PATH
OPTIMIZEDPACKET INTAKE→OPTIMIZEDJITTER BUFFER→OPTIMIZEDDEADLINE SCHEDULER→OPTIMIZEDAUDIO DELIVERY
The integrated receive path reduces avoidable processing and data movement, leaving more CPU and memory headroom for the OEM product.
MEASURED CLOCK-DRIFT CONTROLTwo clocks.
Two clocks.
One audio timeline.
Network time and the audio-device clock do not run at exactly the same rate. In the measured HiFiBerry/PCM5122 run, the device consumed approximately one extra 48 kHz PCM frame—20.83 µs—about every ten seconds.
The Bama Box controller tracked that drift and made six declared, single-frame repeats during the 60-second run, keeping the presentation timeline aligned instead of allowing phase error to accumulate.
Time-drift control
Tracks the network and device clocks and applies explicit frame-level correction.
Jitter buffering
Absorbs arrival variation while protecting the configured receiver target.
PTP time mapping
Uses the platform timestamp source through a defined OEM software API.
ALSA mmap delivery
Feeds the physical audio path with bounded runtime queue depth and low overhead.
STRICT 1 MS RECEIVER TIMING TARGET
Deterministic timing as density rises.
RTP timing drives deadline-controlled delivery. OEM timestamps, platform clocks and audio interfaces connect through a dedicated software API.
USERSPACE PATH VALIDATED · HARDWARE INTEGRATION IS PLATFORM-SPECIFICUS UTILITY PATENT APPLICATION
Systems and Methods for Processing Sound
- Network-audio receive and transmit processing
- Deterministic timing control
- Scalable multichannel operation
- OEM integration support
03 / OEM LICENSINGPredictable economics.
Predictable economics.
Most of the value stays with you.
Production terms are structured to remain attractive even on platforms where the measured gain is at the conservative end of the performance range.
At shipment
No recurring subscription. The production royalty applies once when an enabled OEM product ships.
Aligned to capacity
Licensing follows the enabled network-audio channel class, keeping low-channel products economical.
Predictable maximum
A negotiated per-device ceiling prevents higher channel density from creating an open-ended royalty.
Your Linux SoC
Bring your i.MX7, i.MX8 or other target hardware for a confidential receiver comparison before production commitment.
Exact pricing stays private. Volume, product-family, platform and strategic-rights terms are matched to the product and scope.
Benchmark your target platform04 / MEASURED RESULTSEvidence for the
Evidence for the
OEM business case.
Measured proof covers physical embedded output, sustainable Arm receiver density and separate x86 backstage resource evidence. Review the canonical benchmark record.
BAMA BOX · MEASURED ON EMBEDDED ARM
Physical output and sustainable receiver density.
Reproducible quad-core Arm Cortex-A72 reference platform. Final performance is established on the OEM target SoC.
configured receiver target sustained
ALSA mmap/DMA → I²S → PCM5122, with a maximum runtime ALSA queue of 34 frames (0.71 ms).
stereo streams at 1 / 3 / 5 ms
28 / 66 / 80 audio channels, with final-mmap PCM verification at the stated receiver-software boundary.
COMMON 3 MS+65%
33 vs 20 stereo streams against the tested stock direct-mmap RAVENNA/Linux implementation.
COMMON 5 MS+67%
40 vs 24 stereo streams against the tested stock direct-mmap RAVENNA/Linux implementation.
X86 / KVM6× / 11×
Lower CPU load / receiver memory versus FFmpeg in a separate 16-stream backstage benchmark.
Physical-output, Arm-density, GStreamer, RAVENNA/Linux and x86/FFmpeg evidence use distinct measurement campaigns. Full current methodology and qualifications are maintained on the canonical benchmarks page.
Review all benchmark evidence05 / OPEN-SOURCE REFERENCE PLATFORMStagePi.
StagePi.
AES67 on Linux.
StagePi is Bama Box's open-source AES67 stagebox for off-the-shelf Raspberry Pi 4 and HiFiBerry hardware. It provides a transparent physical-audio reference and ecosystem proof, while the proprietary commercial receiver remains a separate OEM proposition.
Explore the StagePi referenceCONFIDENTIAL OEM BENCHMARK