To put words in reverse audio, reverse the decoded samples so the recording plays from its last sample to its first, then download a new PCM16 WAV instead of reusing the compressed source file. Reverse Audio applies that time-order reversal independently to every decoded channel, so the end of a spoken passage appears at the beginning of the result. The reversal preserves the recording’s decoded duration, sample rate, and channel count when those values meet the tool’s limits. It does not stretch the clip into a longer file or intentionally change its pitch. The result is newly encoded audio, not a backward version of the original compressed bitstream.
Choose an MP3, WAV, M4A, AAC, Ogg, or WebM file no larger than 50 MB. Actual decoding still depends on codec support in the current browser and operating system, so a familiar extension alone does not guarantee success. File reading, Web Audio decoding, sample reversal, preview creation, and WAV encoding happen locally in the current tab. The source bytes, decoded samples, filename, and finished result are not uploaded to Lizely. Listen to the original preview before processing, then check both ends of the reversed preview before downloading.
A reversed export is suitable for reverse-speech experiments, transition effects, sound-design work, sample inspection, and hearing the end of a recording from another perspective. It does not identify or reconstruct hidden messages. The tool changes sample order only; it does not transcribe speech or turn reversed audio into text.
Confirm the source before reversal
Input acceptance and successful decoding are separate checks. The file picker accepts common MP3, WAV, M4A, AAC, Ogg, and WebM audio types, but a container can contain a codec that the current browser cannot decode. A file passing the type and size checks may still produce a decode error. Before processing a spoken recording, confirm the following limits:
- Compressed file size: The source must be no larger than 50 MB.
- Decoded duration: The audio must be five minutes or shorter.
- Sample rate: The decoded sample rate must be between 8,000 and 192,000 Hz.
- Channel count: The source must contain one to eight channels.
- Channel-sample budget: The total must not exceed 30 million channel samples.
- Decoded size: Even a small compressed file can exceed a decoded limit.
The channel-sample budget equals the number of audio frames multiplied by the number of channels. This protects the browser tab from unexpectedly large decoded arrays. A file that exceeds any limit receives a specific message and is not partially processed. Changing the source or retrying also clears the previous download first, preventing an older successful result from being mistaken for the new one.
The Web Audio decodeAudioData documentation describes how complete audio data is decoded into an AudioBuffer. That browser-based decoding step is why codec support matters even when the filename ends in a familiar format.
| Property | What the reversed result does |
|---|---|
| Event order | Moves the decoded recording’s events from its end back toward its beginning. |
| Duration | Preserves the decoded duration of accepted audio. |
| Sample rate | Retains a decoded rate from 8,000 through 192,000 Hz. |
| Channels | Keeps one to eight channels and reverses each channel independently. |
| Encoding | Writes a new signed 16-bit PCM WAV rather than preserving the source codec. |
| Metadata | Does not copy tags, artwork, chapters, loop markers, or encoder settings. |
Reverse words and download a PCM16 WAV
Choose the recording. Open Reverse Audio and select a supported audio file up to 50 MB. You can preview the original in the browser first, which helps confirm that the correct recording was chosen before its sample order is changed.
Run the reversal. Select the Reverse audio action and wait while the browser decodes the source and reverses every channel sample locally. For stereo input, the left and right channels are processed independently and then placed back into the correct interleaved WAV frame order. No source samples are deliberately skipped, shortened, or capped inside an accepted file.
Check the result and save it. Preview the reversed audio and verify its duration, channel count, sample rate, and displayed file size. Then download the explicitly labeled PCM16 WAV. Keep the original source because the download is a new encoded file, not the original MP3, M4A, AAC, Ogg, WebM, or compressed WAV stream.
Check both ends and the technical details
The beginning of the result should correspond to the end of the original recording, while the end of the result should return to the source’s beginning. Listen to both sections rather than judging only the most noticeable passage. Confirm the displayed duration, channel count, sample rate, and file size against what you expect from the source.
Successful browser previewing does not guarantee that every older device will support the same WAV channel layout or high sample rate. Open the downloaded file in the editor or player where it will be used. If the project also requires fades, trimming, codec selection, metadata preservation, or mastering controls, complete those steps in an audio editor after creating the reversed WAV.
Hear what changes in reversed speech
Reversing speech reverses the waveform’s time order. The final portion of a spoken passage becomes the first audible portion, and the opening moves to the end. This is different from typing a word backward or rearranging individual letters. The tool does not inspect the phrase, identify phonemes, or manipulate the order of written characters.
The recording contains the same decoded number of samples after processing, so its duration and channel arrangement are retained. Pitch is not shifted intentionally, and time is not stretched. The sound can still be difficult to understand because speech sounds are heard in reverse order. Use the result for an experiment or effect, not as a guaranteed method for revealing a secret message.
Stereo handling follows the decoded channel structure. A left-right 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 frame order. Multi-channel files follow the same channel-by-channel method, within the limit of eight accepted channels.
Creative uses include previewing a transition from its tail, inspecting how a sample sounds in the opposite direction, building a reverse-speech effect, or evaluating the end of a recording from a new perspective. Keep the source available so you can compare the two files and decide whether the reversal produces the effect you need.
Understand the PCM16 WAV export
The output is always a new PCM16 WAV, even when the source is a compressed WAV or another accepted format. PCM16 means that the decoded audio is encoded as signed 16-bit samples. The browser clips each decoded floating-point sample to the numerical range from negative one to positive one before writing it. Negative full scale maps to -32768, while positive full scale maps to 32767. The WAV bytes are written in little-endian order.
This process does not normalize loudness. It does not remove silence, repair clipped audio, change pitch intentionally, or add mastering effects. Reducing floating-point samples to 16-bit values can change numerical precision, so the export should be treated as a fresh audio render rather than a bit-for-bit transformation of the source file.
An uncompressed PCM16 WAV can be much larger than a compressed source, and the result’s final file size depends on the retained duration, sample rate, and channel count. Check the size shown by the tool before moving the file into a project or device with limited storage. Do not expect the original compressed bitrate to be retained.
The fresh WAV does not carry over every kind of source information. Album art, tags, chapters, loop markers, encoder settings, compressed bitrate, and other container-specific fields are not copied. Preserve any metadata separately if the reversed version must retain the same tags, artwork, or project information.
Respond to a decode or compatibility issue
If the browser reports a decode error, the familiar file type may contain a codec unsupported by that browser or operating system. A WAV, M4A, or other accepted filename therefore does not guarantee that decoding will succeed. The tool reports the decoding problem rather than claiming that every file with a recognized extension must work or producing a partial reversal.
Limits are checked against the decoded audio, not only the compressed file’s byte size. A source that is 50 MB or smaller can still fail if it lasts more than five minutes, uses an unsupported sample rate, contains too many channels, or exceeds 30 million channel samples. No segment of a rejected file is processed.
When changing files or running the operation again, rely on the newest result displayed by the tool. The previous download is cleared first after a file change or retry, and obsolete preview or download URLs are released as work is replaced. Closing or leaving the page also ends access to those temporary local URLs.
If you need a different output codec, an exact trim, fades, metadata preservation, or detailed mastering controls, use an appropriate local audio editor after the reversal. For a reliable handoff, listen to the original, save the reversed PCM16 WAV, inspect the beginning and end, confirm the displayed properties, and test the downloaded file in the application that will use it.