Convert DSS to AIFF — Free Online Tool
Convert DSS dictation recordings to AIFF by decoding the proprietary OKI ADPCM audio into uncompressed 16-bit big-endian PCM — the native lossless format used on macOS. This is ideal for archiving voice recordings from Olympus, Philips, or Grundig digital dictaphones into a format compatible with professional audio editing tools.
to
FFmpeg Command
Copy this command to run the same conversion locally with FFmpeg on your desktop. Download FFmpeg
Drop your DSS file here
or click to browse
Free — no uploads, no signups. Your files never leave your browser.
Settings
Note: Browser-based encoding uses approximate quality targets. For precise CRF compression, copy the FFmpeg command above and run it on your desktop.
Estimated output:
Conversion Complete!
DownloadHow It Works
DSS files store audio using the adpcm_ima_oki codec, a low-bitrate adaptive delta PCM compression scheme specifically designed for speech intelligibility on digital dictation hardware. During conversion, FFmpeg decodes this proprietary ADPCM stream into raw PCM samples and then encodes them as pcm_s16be (16-bit signed big-endian PCM) wrapped in the AIFF container. Because ADPCM is a lossy format, no quality is recovered during this process — the AIFF output faithfully represents the audio as decoded from the DSS file, without the additional compression artifacts of re-encoding. The resulting file is fully uncompressed and significantly larger than the source DSS, but is now universally readable by macOS audio applications, DAWs, and transcription software.
What Each Flag Does
| Flag | What it does |
|---|---|
ffmpeg
|
Invokes the FFmpeg program. All conversion logic — demuxing the DSS container, decoding the adpcm_ima_oki stream, and encoding to AIFF — is handled by FFmpeg's built-in format and codec libraries. |
-i input.dss
|
Specifies the input DSS file. FFmpeg detects the DSS container and selects its adpcm_ima_oki audio decoder automatically based on the format header. |
-c:a pcm_s16be
|
Instructs FFmpeg to encode the decoded audio as 16-bit signed big-endian PCM, which is the standard uncompressed audio codec for AIFF files on macOS. This replaces the lossy ADPCM compression with a lossless PCM representation of the same decoded samples. |
output.aiff
|
Specifies the output filename and triggers FFmpeg to use the AIFF muxer based on the .aiff extension. The AIFF container wraps the pcm_s16be audio stream in a format natively understood by macOS and Apple audio applications. |
Common Use Cases
- Archiving recordings from Olympus DS or WS series voice recorders into a lossless AIFF format for long-term storage on macOS systems
- Importing dictation audio into macOS-based transcription software such as Express Scribe or Apple Logic Pro, which may not natively support DSS files
- Delivering interview or deposition recordings to a legal transcription service that requires uncompressed audio for accuracy
- Preprocessing speech recordings for use with machine learning or voice recognition pipelines that expect standard PCM audio
- Converting a medical or corporate dictation archive from DSS to an open, non-proprietary format before decommissioning legacy dictation hardware
- Editing voice memos recorded on a Philips SpeechMike in a professional audio editor like Adobe Audition or GarageBand, which handle AIFF natively
Frequently Asked Questions
No — DSS uses lossy adpcm_ima_oki compression, meaning some audio detail was permanently discarded when the recording was originally made on the dictation device. Converting to AIFF produces an uncompressed representation of the already-lossy DSS audio; it does not recover frequencies or clarity that were lost during recording. The AIFF output will sound identical to the DSS source, but the file will be much larger because it is no longer compressed.
DSS uses low-bitrate ADPCM compression optimized for speech, which achieves very small file sizes at the cost of audio fidelity. AIFF with pcm_s16be stores every audio sample as a raw 16-bit value with no compression at all. For a typical dictation recording sampled at 8 kHz or 11 kHz, you can expect the AIFF to be roughly 10 to 20 times larger than the original DSS file. This is expected and normal for lossless PCM audio.
Yes — AIFF with pcm_s16be is Apple's own native audio format and is fully supported by GarageBand, Logic Pro, Final Cut Pro, and virtually every other macOS audio application. It will import without any transcoding step. Note that the source audio was recorded as narrow-band speech, so it will not sound like a studio recording regardless of the container format.
DSS files embed proprietary metadata fields specific to dictation workflows — such as author ID, worktype, and priority level — that are not part of the AIFF specification. FFmpeg does not map these DSS-specific fields to AIFF metadata, so that information will be lost in the conversion. If preserving this metadata is important, record or export it separately from your dictation management software before converting.
Replace pcm_s16be in the command with pcm_s24be for 24-bit output: ffmpeg -i input.dss -c:a pcm_s24be output.aiff. AIFF also supports pcm_s32be and floating-point formats like pcm_f32be. However, because the DSS source is a low-quality speech codec, increasing the bit depth beyond 16-bit provides no audible benefit and only increases file size.
Yes — on macOS or Linux you can run a shell loop: for f in *.dss; do ffmpeg -i "$f" -c:a pcm_s16be "${f%.dss}.aiff"; done. On Windows Command Prompt, use: for %f in (*.dss) do ffmpeg -i "%f" -c:a pcm_s16be "%~nf.aiff". This applies the same conversion to every DSS file in the current directory and outputs a matching AIFF file for each one.
Technical Notes
DSS (Digital Speech Standard) is a closed, proprietary format developed jointly by Olympus, Philips, and Grundig for handheld digital dictation devices. Its audio is encoded with the adpcm_ima_oki variant of IMA ADPCM, which is tuned for speech at narrow sample rates (typically 8,000 Hz or 11,025 Hz) and very low bitrates. FFmpeg's DSS demuxer support is read-only and handles the format reasonably well, though some extended DSS+ features or encrypted files from certain enterprise dictation systems may not demux correctly. The target codec, pcm_s16be, is 16-bit signed big-endian PCM — the canonical audio encoding inside AIFF containers as defined by Apple's original specification. Because the source is narrow-band speech, the output AIFF will have a low sample rate and may need resampling (using -ar 44100 added to the command) if the destination application expects CD-quality audio. No subtitle, chapter, or multi-track audio data exists in DSS files, so none is lost in the conversion. The AIFF format does not support the proprietary dictation metadata fields present in DSS headers.