Reverse Audio does not translate speech; it reverses every decoded channel sample locally and downloads the result as a PCM16 WAV.
The search phrase “translate reverse audio” can describe two different jobs. To hear an existing recording from its end toward its beginning, Reverse Audio performs the reversal. To understand words in another language or create dubbed narration, you need a separate transcription and translation workflow.
This guide covers the local reversal process, the distinction between translation and playback reversal, and the checks to apply before using the finished file in a video, podcast, sound-design session, or editor.
Language translation and sample reversal are different
Language translation changes words or meaning from one language into another. Sample reversal changes the order of sounds that have already been decoded. The second operation does not identify a language, rewrite speech, create subtitles, or generate a dubbed voice.
Reverse Audio handles only the reversal stage. For a literal reverse-then-translate workflow, create and download the PCM16 WAV here, submit it to a speech-to-text service, review the transcript, and send that transcript to a translation service. If you need intelligible translated narration with a backward effect, translate or dub the original first, export it, and then reverse that finished recording here. It can support reverse-speech effects, transition ideas, sample inspection, and hearing the end of a recording from a different perspective. If the goal is translated narration, translate or dub the recording first and reverse that finished audio only when the backward effect is also wanted.
- Translation: converts spoken content between languages.
- Reversal: moves decoded samples into the opposite time order.
- Reverse Audio: performs local reversal and PCM16 WAV export without translating the words.
Expecting translation from a reversed file can lead to a wrong conclusion. The unusual playback may come from inverted timing and sound order, not from any change to the spoken language.
What the reversed file retains
Reversal operates on decoded audio rather than on the compressed file itself. An accepted recording is processed completely, with every channel reversed and no samples deliberately skipped or shortened.
| Property | Result after reversal |
|---|---|
| Sample order | Each decoded channel runs from its end toward its beginning. |
| Duration | The decoded duration is preserved, with no intentional truncation. |
| Sample rate | The original decoded rate is retained when it meets the stated limits. |
| Channels | The decoded channel count is retained, and every channel is reversed independently. |
| Output encoding | A new PCM16 WAV is created rather than reusing the source compression. |
| Metadata | Original tags, artwork, chapters, and container fields are not promised. |
A stereo input remains stereo. The left channel is reversed independently, the right channel is reversed independently, and the two are interleaved again in the correct WAV frame order. The operation does not intentionally change pitch, stretch time, remove silence, normalize loudness, or repair clipped peaks.
During output encoding, decoded floating-point samples are clipped to the range from -1 to 1 and mapped to signed 16-bit PCM. That conversion can change numerical precision. A compressed source may also produce a much larger WAV because the result contains uncompressed PCM samples rather than the original compressed stream.
Input formats and decoding limits
The file picker accepts MP3, WAV, M4A, AAC, Ogg, and WebM audio. Selection by MIME type is only the first check: successful decoding depends on the codec and the visitor’s browser and operating system. A familiar container can contain a codec that the current browser cannot read. The actual decode uses the browser’s Web Audio API, as described in MDN’s decodeAudioData documentation.
- Compressed input size: no larger than 50 MB.
- Decoded duration: five minutes or shorter.
- Decoded channels: one to eight channels.
- Decoded sample rate: 8,000–192,000 Hz.
- Total channel samples: no more than 30 million, calculated as frames multiplied by channels.
Each limit must be satisfied. A file that exceeds any one of them receives a specific message and is not partially processed. The channel-sample limit also prevents unexpectedly large decoded arrays from overloading the browser tab, even when the compressed source appears small.
File reading, decoding, reversal, WAV encoding, and preview creation all happen in the current browser tab. The file bytes, decoded samples, filename, and finished result are not uploaded to Lizely, and the workflow does not silently switch to a remote conversion server.
Reverse a recording in the browser
- Choose the source file. Open the Reverse Audio tool and select an MP3, WAV, M4A, AAC, Ogg, or WebM recording no larger than 50 MB. Keep the original file available for comparison.
- Preview the original if needed. Use the browser preview to confirm that you selected the intended recording and to hear its normal beginning and ending before changing the sample order.
- Select Reverse audio. Wait while the browser decodes the file, checks the decoded limits, and reverses every channel locally. A file that cannot be decoded or does not meet a limit is rejected without producing a partial result.
- Preview and inspect the result. Play the reversed recording, including its beginning and end. Confirm the duration, channel count, sample rate, and file size displayed by the tool before downloading.
- Download the PCM16 WAV. Save the file labeled PCM16 WAV, then open that download in the editor or player where it will actually be used.
Check the result where it will be used
A successful browser preview is a useful check, but it does not guarantee compatibility with every destination. Older devices may not support the same WAV channel layout or high sample rate, so verify the downloaded file in the final editing or playback environment.
- Compare the original duration with the duration shown for the reversed file.
- Confirm that the expected channel count and sample rate are displayed.
- Listen at both ends, because the source ending now appears at the result’s beginning.
- Check that the target editor or player opens the downloaded WAV normally.
- Confirm that the output meets your delivery requirements rather than assuming they were carried over from the source.
The output is always a newly encoded PCM16 WAV. It does not preserve the original MP3, M4A, AAC, Ogg, WebM, or compressed WAV encoding, and it does not promise original album art, tags, chapters, loop markers, encoder settings, or container-specific fields. Use an audio editor when the project also requires fades, trimming, a particular codec, metadata preservation, or mastering controls.
Choose a workflow that matches the goal
Start by separating the desired outcome from the operation that produces it. The table below distinguishes backward playback from language work and detailed editing.
| Goal | Suitable workflow | Reason |
|---|---|---|
| Hear an existing recording backward | Use Reverse Audio | It reverses every decoded channel locally and exports a PCM16 WAV. |
| Understand or dub spoken content | Use a dedicated transcription and translation workflow | Sample reversal does not change or interpret the words. |
| Add fades, preserve metadata, or choose another codec | Finish those tasks in an audio editor | The reversed WAV is a fresh file without the source container metadata. |
| Reverse only part of a long recording | Create a shorter excerpt first, then reverse it | The decoded input must be five minutes or shorter and stay within the channel-sample limit. |
For translated speech that also needs a backward effect, complete the translation or dubbing first. Export that audio, then use Reverse Audio for the sample-order change. This sequence preserves the actual translation task and treats reversal as the separate effect it is.
Recover from a rejected or confusing result
A decode error does not necessarily mean the filename is wrong. The current browser may not support the codec inside the selected container. Try a browser or operating system with support for that codec, or use an editor to create a compatible WAV and retry. The tool reports a decode error instead of claiming that every file with a recognized extension will work.
For a file that exceeds the duration, channel, sample-rate, or channel-sample limit, make a shorter or simpler source before trying again. Changing files or rerunning the operation also clears the previous download, preventing an older success from being mistaken for the new result after an error. Keep the source recording intact and work from the newly generated download.