Home
Software
Guru
Dynamic Score
Making decisions on automated dynamic range compression requires a scoring system that takes three different measurements into consideration.
Audio encoding artifacts
How audio encoders use psychoacoustic masking to reduce file sizes, and why this process creates audible artifacts in compressed formats.
Loudness and normalization
Understanding what is loudness, how it's measured and why standards exist.
Bit depth in digital audio
Understanding bit depth, quantization and why float sample rates are needed.
Audio dithering
How randomization helps to alleviate the effects of quantization.
See all Guru articles
Research Blogs Contact

A27 technical specification

Architecture

A27 runs on a single server. Liquidsoap runs in a Docker container; supplementary systems run as Cloud Jobs on the host (if in GCP), or cron jobs for non-GCP deployments. HLS segments are served directly from the same server. No separate infrastructure to coordinate.

Deployment time: under ten minutes.

GCP-hosted deployments available for high-profile customers requiring geographic redundancy or serverless scaling.

Engine

Built on Liquidsoap 2.4.3 for audio routing, processing, stream generation and source switching. Prometheus collects audio-specific metrics. Full logging enabled.

Hosted on Intel infrastructure, located in EU for EU clients, locations in USA and Asia available.

Content acquisition

Native RSS feed ingestion with health monitoring. System checks feed status, detects stale content, alerts on failures. New episodes appear automatically in rotation.

Supports public and private RSS feeds. Manual file uploads available. Handles any podcast hosting platform as long as the audio files are accessible over the Internet.

Feed health metrics exported to Prometheus.

Audio processing

Professional broadcast chain using Stereotool and custom DSP. Multiband compression, limiting, EQ. Loudness normalization to EBU R128 standards.

Source management and failover

Multiple input types: RSS feeds, live SRT streams, HTTP, file-based playlists. Priority-based fallback chains. When a source fails, system switches to backup in under 200 ms.

Live source contributions via SRT — works over cellular, satellite, or unstable networks. Automatic jitter buffering. Designed for remote broadcasts over unstable networks.

Distribution

HLS is the default delivery mechanism across all tiers. Segments are served from the host server and distributed through Bunny CDN’s network of 114 PoPs worldwide.

All tiers include a basic web player.

AI news generation

Ingest news sources (wire services, RSS feeds, custom inputs). Generates broadcast-ready news copy using language models. Text-to-speech conversion using voice synthesis.

Automated scheduling — newscasts appear in programming at specified times. Manual triggering available.

Output integrated directly into A27 playlists, and to your AOD platform.

Scope of service does not include paid wire service API’s. Use your own key for paid wire services.

Monitoring and observability

Prometheus metrics for LUFS, RMS, source rotation, uptime, and dead air. Grafana dashboards available. Metrics can be exported to your own monitoring stack.

Silence detection with configurable alerting. If stream goes silent for X seconds, you get notified. Silence detection triggers failover before alerting.

Container health monitoring with automatic restart on failure.

Integration capabilities

Controlled via Telnet or HTTP.

Pairs with OB1 for remote contribution. Pairs with Scribbles for AI-assisted journalism — covers the full workflow from recorded interview and written notes to a broadcast-ready spoken news article, published directly to A27 playlists with one click.

OB1 in studio

OB1 can be used in studios. It connects users to the A27 instance, allowing slot control and live unit on-air. It’s an on-premises installation. Ask for more information.

What’s not included

Automated playout only. Host-driven shows require separate playout software. Any software will work as long as it can output audio in any common protocol.

We don’t handle licensing/royalty reporting, but export a list of played songs. Podcasting equipment (microphones, mixers, interfaces) not included. Available on request.

Security and access

We handle server maintenance, security updates, and firewall configuration. You deploy content, we maintain infrastructure.

SSH access available on request in Enterprise tier. Full root access with managed support contract.

Server hardening and automatic security updates included. SOC2 compliance available in Enterprise tier.

Technology stack summary

Infrastructure: ISO 27001 certified, redundant EU datacenters available in Enterprise tier. Protocols: HLS, SRT, (Icecast/HTTP for legacy connections) Compliance: SOC2 available (Enterprise tier) Engine: Liquidsoap >2.4.3, FFmpeg, Stereotool (license included) Runtime: Docker (Liquidsoap container), Cloud Jobs (supplementary systems)

FAQ

Can we self-host?

No. Managed infrastructure only.

What happens if your infrastructure fails?

Geographic redundancy available in Enterprise tier and GCP deployments. Standard tiers run single-datacenter with automatic container restart. SLA guarantees available at Enterprise tier.

Can we integrate with our existing systems?

API available in Professional tier and above. Enterprise tier includes custom integration support. A27 can handle any common input format and protocol over the Internet.

What’s your stack’s reliability track record?

Production deployments maintain 99.5%+ uptime. Downtime primarily from upstream CDN issues and scheduled maintenance.

Do you use Docker?

Yes. Liquidsoap runs in a Docker container. Supplementary systems run as jobs on the host. This is what makes sub-ten-minute deployments possible.

What is GCP support?

High-profile customers can run A27 on Google Cloud Platform, enabling serverless scaling and geographic redundancy beyond what single-server deployments offer. GCP also allows easily serving different content for different geographical areas.