
Sample Data Export writes individual waveform values as numbers. Its sample-count limit can stop output before the end of your selected audio: the documented default is 100 samples, not 100 seconds. Set the limit for the intended analysis and check the actual data file. This tool does not make an ordinary listening copy of the recording.
Key Takeaways
- Count samples using the selected duration and actual sample rate.
- Choose linear amplitude or dB deliberately and record that choice.
- Verify numeric values and coverage before importing the file into another analysis tool.
Define a small analysis request
Suppose a colleague needs the waveform values around a short edit, rather than a WAV to listen to or an FFT frequency plot. Identify the source file and selected interval, and preserve the audio project separately. Numbers without their sample rate and selection context are easy to misinterpret.
For a hypothetical one-second mono selection at 48000 samples per second, full coverage requires 48000 values. A cap of 100 covers only about 0.00208 seconds at that rate. This is arithmetic explaining the setting, not a measured export from a real recording.
Export a deliberate number of samples
Select the intended mono passage and choose Tools > Sample Data Export. Set Limit output to first samples to the required count, within the documented maximum of 1000000. Start with a small inspection file before generating a large table.
Choose Linear when the receiving analysis expects amplitude values, or dB when that representation is required. Use Select a file to choose the destination and the intended TXT, CSV, or HTML extension. A changed extension selects the data format; it does not turn the table into an audio file.
For a TXT inspection, Sample Count provides a numeric index. Include a suitable header and note the source interval. The index starts at the first selected sample; do not mistake it for the original project's absolute timecode.
Reconcile the file with the request
Open the actual file in a text viewer or appropriate spreadsheet import. Separate metadata/header lines from sample records before counting them. Confirm the selected units, expected number of values, first record, and final record. Check that the file is the new export rather than a similarly named earlier test.
Keep a short record of source, selection, sample rate, cap, units, and destination. In the hypothetical one-second case, a 100-record file fails the requested coverage even though the export completed. A 48000-record file still needs its context to explain what the numbers represent.
Keep data analysis distinct from delivery approval
Sample values do not by themselves establish speech meaning, synchronization, or acceptable loudness. Supply a separately identified listening reference when the colleague needs to hear the edit. For that distinction, review the listening proxy versus final master and keep the approved version identifiable.
This is a proposed documentation-based analysis handoff. No sample export or measurement was executed.
Related tasks
- Should I Send a Review Copy or the Final Master of an AI Video to a Client?
- How Do I Keep the Approved AI Image or Video from Getting Lost Among Revisions?
Sources and scope
- Audacity Manual: Sample Data Export: Numeric sample export, default100 and maximum1000000 sample cap, linear/dB units, TXT/CSV/HTML, index and header controls.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance, not a claim of hands-on testing or independent review. Report a correction.
Cover: Pignatta / Pixabay, Content License. The city photograph is an illustrative source asset; it does not show this application or a test result.