Structured records and telemetry
Include actual JSON, CSV, logs, configuration and numeric records. Preserve original formatting where exact file bytes matter. Measure small-message overhead and the chunk sizes your pipeline uses.
Compression guide
What is Saar lossless compression?
Saar is a proprietary, byte-exact lossless compression engine for text, structured data, media, telemetry, firmware and archives. A successful decode returns the original bytes. Savings depend on the input and configuration.
Saar is the product brand. HEOSSI (PTE.) LTD. develops and publishes it. This guide describes the product, its evidence and how to decide whether to evaluate it.
What byte-exact lossless recovery means
Keep the original file or byte stream as the acceptance reference. Compress it, decode the result, and compare every recovered byte with the reference. Headers, metadata, padding and ordering count when they are part of the original bytes.
Saar accepts a candidate only after recovery verification. When compression does not help, a stored representation preserves the bytes instead. Integrity checks detect corruption; they are not encryption or proof of sender identity.
Read the formats, limits and integration contracts or review integrity and security boundaries.
How to compare Saar with Zstandard, Brotli, LZ4 and gzip
Include your current codec as the baseline. If you test Zstandard (zstd), Brotli, LZ4, gzip or other candidates, give each the same original input and record its exact version, compression settings and execution environment. A codec name alone is not a reproducible comparison.
- Freeze representative inputs, including small records, large files and the inputs that compress poorly. Keep a record of provenance and hashes.
- Measure the complete emitted object or stream, including framing and integrity overhead. Do not compare one codec's payload with another codec's complete output.
- Measure encoding and decoding separately, along with memory and the latency relevant to your application. Repeat timed measurements and record variability.
- Decode with the intended reader and verify exact original-byte recovery. Test corrupted and truncated inputs and enforce resource limits.
- Assess decoder availability, format compatibility, streaming or random-access needs, operational support and licensing alongside size and speed.
A smaller output on one sample does not establish better throughput, lower cost or leadership across a workload class. Saar's local development machine is not a universal hardware baseline.
Inspect the recorded comparisons and read the methodology and limitations. Use the recorded results to select evaluation candidates, then decide with your own measurements.
Which workloads should an evaluation include?
Media and firmware
Separate original-file preservation from decoded pixel or sample preservation. Include already-compressed assets and actual firmware images so that incompressible inputs remain part of the result.
Archives and stored objects
Measure complete files and realistic collections, then test restore time and reader compatibility. Include metadata and framing in storage calculations and keep an independently checked recovery record.
No lossless compressor can guarantee smaller output for every possible input. Saar does not claim a universal win or compression of encrypted data beyond its entropy limit.
Questions about Saar
What is Saar compression?
Saar is a proprietary lossless data compression engine. It accepts original bytes and produces a compressed object or checked stream. A successful decode reproduces the original bytes exactly. It is offered through controlled technical evaluation; public package distribution is not enabled yet.
Is Saar the same as SAR radar or MRI compression?
No. Saar is the name of this lossless data compression product. SAR can refer to synthetic aperture radar or specific absorption rate in other fields. Those terms do not identify the Saar engine described here.
Is Saar better than Zstandard, Brotli, LZ4 or gzip?
There is no universal winner. Saar publishes sample-level comparisons with named codec settings, environment and limitations. A decision requires measuring your representative files, complete encoded sizes, encode and decode time, memory, integration requirements and exact recovery.
Does lossless media compression preserve the original file?
Preserving decoded pixels or audio samples is different from preserving an entire original file, including metadata and container bytes. Saar's recovery contract is the original byte sequence. Verify the same requirement for every codec or conversion pipeline in your comparison.
Can I download Saar or install a public package?
Public package distribution is not enabled yet. Request a controlled technical evaluation. Scope, access, licensing, data handling and support must be agreed in writing before use.
Evaluate on evidence
Saar is available through controlled technical evaluation. Public package distribution is pending. Agree scope, data handling, access and commercial terms before starting.