Expert report review
Enhanced audio evidence: what does a reviewer check in the other side's enhancement?
Enhanced audio evidence is tested against the examiner's own record. SWGDE's 2025 best practices for audio enhancement call for work on a copy, a final comparison with the unprocessed original and a log of every process, setting and software version detailed enough to repeat the result. A reviewer rebuilds the enhancement from that log, listens for speech the filters removed or sounds they created, and turns each gap into a discovery request, a deposition question or a Rule 702 point.

What does audio enhancement do, and what can it not do?
Enhancement filters a recording to improve signal quality and intelligibility. SWGDE's best practices state that it may or may not improve the signals of interest and that greater intelligibility may not be achievable.
The Scientific Working Group on Digital Evidence (SWGDE) approved version 2.0 of its Best Practices for Enhancement of Digital Audio, document 20-A-001, on June 27, 2025, and formatted it on August 21, 2025. The first section defines the work: "Audio enhancement is the processing and filtering of audio recordings to improve the signal quality and intelligibility." The same section sets the test a reviewer applies to the other side's version: "Enhancement results must be repeatable and reproducible to the extent that they are perceptually consistent."
The document states its limits in plain words. Section 5 reads: "The goal of increasing intelligibility or signal quality may not be achievable." It adds that perceived quality and intelligibility "can be evaluated differently by different listeners." The challenges table in Appendix A lists no mitigation for a drop-out, a stretch where the signal is missing; the examiner documents its location and duration. Filters can lower hum, hiss, wind noise or reverberation. Words that were never captured stay missing.
Research on courtroom recordings points the same way. Helen Fraser, writing in the Griffith Journal of Law & Human Dignity in 2020, found "no research-based evidence supporting the capability of any general engineering techniques" to objectively improve the intelligibility of indistinct audio (p. 90). She treated listenability and intelligibility as two separate dimensions (p. 91). On her account, a version that is easier to listen to does not by that fact let a listener understand more of the words.
Did the examiner enhance a working copy and keep the original unchanged?
The reviewer looks for a working copy, a hash of the original and a record of what happened if the original never arrived. SWGDE says all examinations should be carried out on a working copy.
Section 9 of SWGDE 20-A-001 states: "All examinations should be carried out on a working copy of the evidentiary materials." SWGDE's Best Practices for Forensic Audio, document 08-a-001, version 2.5 of June 9, 2022, describes how that copy is made: "To preserve the integrity of a digital original, make a bit stream duplicate to use as a working copy." The exactness of a digital copy "can be tested and confirmed through the use of a hash function," and the same document asks for the hash comparison to be documented.
When the examiner never had the original, the enhancement guide asks for a record of that as well. Section 8 reads: "If the original audio recording is not provided, document your efforts to obtain the original evidence." The same section has the examiner contact the submitter to clarify task-relevant issues and "document communications relevant to the request and examination." Those communications tell the reviewer what the examiner was asked to find.
Format conversions leave their own traces. Document 08-a-001 warns that "re-encoding and resampling should be avoided due to the possible introduction of audible artifacts." If the report names a compressed export as the source file, the reviewer marks which sounds may come from the conversion before asking what the filters did.
Did the examiner assess the recording before applying filters?
The guide lists four assessments before processing: critical listening, waveform analysis, spectrographic analysis and spectral analysis. The examiner also documents the regions of interest, the parts of the recording that hold the target signal.
Section 9 names the assessments: aural review or "critical listening," "waveform analysis for time-domain issues," "spectrographic analysis of the frequency-domain with respect to time" and "spectral analysis of frequency content." Section 7 asks the examiner to "locate the region of interest (ROI) within the recording that contains the target signal," and Section 9 to "document the regions of interest for enhancement."
Without these steps the reviewer cannot tell what problem each filter was meant to solve. Appendix A pairs each challenge with mitigation strategies: sample interpolation for clipping, a notch filter, comb filter or spectral editing for stationary tones, de-reverberation for reverberation, time-stretching that keeps the pitch for fast speech. The assessment tells the reader which of those challenges the examiner found and at which time codes.
Language enters at this stage. Section 9 warns: "There is a risk of negatively affecting linguistically relevant information when processing speech in an unfamiliar language." It advises: "Consider consulting with a fluent speaker." When the conversation is in a language the examiner does not speak, the reviewer looks for that consultation in the report and for the name of the person consulted.
Does the report list every process, setting and software version?
It should. Section 10.1 of the SWGDE guide asks for the processes, settings and applications with their versions and the time segments processed, in enough detail to repeat or reproduce the final result.
The sentence in Section 10.1 reads: "Document the processes, settings, and applications used, with their versions, the time segments processed, and other relevant information in sufficient detail to repeat or reproduce the final results." The next sentence names the usual forms of that record: "Application project files, history logs, and screen shots may serve as documentation of their content."
Fraser and eight co-authors, writing in the Journal of the Audio Engineering Society in November 2024 (vol. 72, no. 11, pp. 748-753), described the gap that court practice leaves: "Practitioners are required only to report the techniques they used" (p. 749). A list of techniques without settings and segments cannot be repeated. A reviewer who receives only the line "noise reduction and equalization were applied" has no way to rebuild the version the jury will hear.
In federal civil cases the log belongs to what must be disclosed. Rule 26(a)(2)(B)(ii) requires the report to contain "the facts or data considered by the witness in forming them," and the 2010 committee note says the phrase is to be "interpreted broadly to require disclosure of any material considered by the expert, from whatever source, that contains factual ingredients." The same note states that the protection for drafts and attorney communications does not exempt "the expert's testing of material involved in litigation" from discovery.
Were the software tools tested, and was the processing order planned?
SWGDE notes that software may implement proprietary or unpublished algorithms and says their effectiveness should be tested before use. The examiner also decides an order of processing, because some filters work better before or after others.
Section 10 reads: "Software applications may implement proprietary or unpublished algorithms. Their effectiveness should be tested before use." A report that names a plug-in and a preset without saying how the examiner tested it on known material leaves that step open. Rule 702(c) asks whether the testimony is "the product of reliable principles and methods," and since December 1, 2023 the proponent must show the court that each requirement of the rule is met more likely than not.
Order is part of the method. Section 10 tells the examiner to "decide an order to apply the selected mitigation strategies" and explains that "certain mitigation processes may be more effective when applied before or after others." Version 1.2 of the same guide, dated September 17, 2020, gave a suggested sequence. It began with distortions such as de-click and de-clip, moved through source separation, continuous and varying noises, de-reverberation and the balance of sibilants and plosives, and ended with gain correction.
The reviewer reads the log in sequence and asks why each step came where it did. Where the log is missing, the same question goes into the deposition outline, together with a request for the project file.
Could the processing have removed speech or created sounds that were never recorded?
Yes. SWGDE warns against over-processing and asks examiners to review the filter residue, the signal components being removed, so that parts of the target speech are not removed with the noise.
Section 10.1 reads: "Avoid over-processing which may introduce new challenges. Review filter residue (the signal components being removed) to avoid removing target signal components." Section 7 asks the examiner to plan the workflow "without creating additional challenges."
Robert C. Maher of Montana State University described the mechanics in IEEE Signal Processing Magazine in March 2009 (pp. 84-94). When the noise estimate in spectral subtraction is wrong, "the noise reduction is incomplete and prone to undesirable audio artifacts," and the leftover energy can sound like "birdie noise" or "musical noise" (p. 90). Spectral filtering can also remove fricatives and plosives, consonants he calls critical to speech intelligibility (p. 90). Because such enhancement is performed on monophonic data as a blind, single-ended process (p. 88), "the determination of noise versus signal cannot be perfect" (p. 90).
Maher also wrote that a court may need to be convinced that enhancement "could not have resulted in a change to the meaning or interpretation of the recorded dialog" (p. 85). The reviewer listens to the residue file and to each disputed passage in every version, and records the time codes where the enhanced version and the original differ in what can be heard.
Did the examiner compare the result with the unprocessed original?
The guide calls for a final review that compares the result with the unprocessed original, after the examiner's hearing has rested. During the work the examiner compares output with input and versions with each other.
Section 10.1 states: "Compare the process output to the input" and "Compare different versions or iterations of processed audio." Section 10.2 adds: "Perform a final review comparing the result to the unprocessed original" and "The auditory system should be rested before review." The same section warns that "listener (ear) fatigue occurs over extended periods of use."
SWGDE's authentication guide, document 15-a-001, version 1.3 of September 20, 2018, starts from the same place. It recommends a "critical aural review of the submitted recording, without applying any processing." A listener who hears only the enhanced version has nothing to compare it with.
Section 10.3 of the enhancement guide states that "production of multiple enhanced results may be helpful to the submitter" and that "it may be useful to accompany the output product with a copy of the submitted audio in a universally-playable format." The reviewer asks whether the other side produced every version the examiner made or only the one it plans to play.
Can an enhanced version lead listeners to hear words that are not on the recording?
In Fraser's experiment, listeners of an enhanced version reported hearing more words, but all reported different words and none came close to the true content. A 2024 paper adds that enhancement can make an inaccurate transcript more credible.
Fraser described the experiment in her 2020 article (p. 93). Listeners were randomly assigned to the original or the enhanced version of an indistinct recording whose true content the researchers knew. Those who heard the enhanced version were more likely to report hearing words. They reported different words from one another, and none came close to what was said.
The 2024 paper by Fraser and her co-authors explains why a transcript changes the picture. Transcripts prime listeners, and "this priming effect can be just as powerful if the transcript is inaccurate" (p. 749). Enhancement can then have "the paradoxical effect of enhancing the credibility of an inaccurate transcript" (p. 750). The authors also wrote that "most jurisdictions now routinely admit enhanced versions of forensic audio in court" (p. 748).
The reviewer keeps two questions apart: what the processing did to the signal, and who decided what the words are. If the opposing examiner listened with a transcript prepared by a party, that transcript is among the facts or data considered. Attorney communications that "identify facts or data that the party's attorney provided and that the expert considered" are discoverable under Rule 26(b)(4)(C)(ii), and so are assumptions the attorney provided under Rule 26(b)(4)(C)(iii).
Was the enhanced file used for a voice comparison or an authenticity test?
If it was, the reviewer applies SWGDE's limits: speech meant for speaker recognition may receive only processing that does not change recognition results, and authenticity testing starts from listening without any processing.
Section 3 of 20-A-001 states that if the target signal is speech intended for a speaker recognition process, "only those mitigation strategies that do not impact speaker recognition results should be employed." Equalization changes the spectral balance of a voice, and time-stretching changes its rate. The reviewer checks whether the file the voice examiner compared was the raw recording or a processed version, and which processing it carried.
Document 08-a-001 (section 8.5) states that "a subject matter expert should be consulted as speaker comparison and identification go beyond the technical aspects" of audio work, which it places outside SWGDE's scope. An enhancement examiner who goes on to say whose voice is on the recording is giving an opinion in a second field, and Rule 702 asks about qualification and method for each opinion.
Authentication runs on the file as received. SWGDE 15-a-001 states: "Transcoded or re-encoded versions of an original recording are not original recordings." Once a file has been transcoded, analysis of the original file format is no longer possible, and the guide adds that no single analysis is to be relied upon solely. A finding about edits made on an enhanced export lacks the file-structure tests the guide lists.
What should counsel request so the reviewer can repeat the enhancement?
The native recording, the working copy and its hash, every enhanced version, the project files and history logs, the software names and versions, and any transcript or notes the examiner had while listening.
Rule 34(a)(1)(A) reaches "any designated documents or electronically stored information," and Rule 34(b)(1)(C) lets the requesting party specify "the form or forms in which electronically stored information is to be produced." For audio the form decides what can be checked: the native file keeps its metadata and structure, while a re-export may lose them. The written response is due "within 30 days after being served" under Rule 34(b)(2)(A).
A deposition notice can carry the same request. Under Rule 30(b)(2), materials designated in a subpoena duces tecum "must be listed in the notice or in an attachment," and a notice to a party deponent may include a Rule 34 request. Rule 26(b)(4)(A) permits the deposition of a testifying expert only after the report is provided, so the reviewer's list of files goes out before that date.
With the project file and the native recording, the reviewer can run the same chain on the same segments and compare the result with the version produced. If the outputs do not match by ear and on the spectrogram, the difference becomes a question for the examiner and a paragraph in the rebuttal.
How does the review of an enhancement reach the court?
Through a rebuttal report due within 30 days after the other side's disclosure, questions at deposition and trial, and a Rule 702 motion. Whether the enhanced version is played, and how it is described, is for the court.
Rule 26(a)(2)(D)(ii) gives 30 days after the other party's disclosure for evidence "intended solely to contradict or rebut evidence on the same subject matter," unless a stipulation or court order sets other dates. A rebuttal on an enhancement answers the same subject: what the processing did, whether it can be repeated, and what it cannot show.
Rule 702 as amended on December 1, 2023 admits expert testimony only if the proponent establishes that "it is more likely than not" that the testimony "is based on sufficient facts or data" and "is the product of reliable principles and methods." The 2023 committee note adds that "each expert opinion must stay within the bounds of what can be concluded from a reliable application of the expert's basis and methodology." The review lists each place where the opposing opinion goes past what the processing record supports.
Rule 705 lets an expert state an opinion without first testifying to the underlying facts or data, "but the expert may be required to disclose those facts or data on cross-examination." The 2010 note to Rule 26 confirms that counsel may question experts about "alternative analyses, testing methods, or approaches to the issues on which they are testifying, whether or not the expert considered them." Rule 403 lets the court exclude relevant evidence if its probative value is substantially outweighed by a danger such as unfair prejudice or misleading the jury.
What does IAE do when the other side's enhanced recording is disputed?
An IAE examiner reviews the opposing audio report for a US case: which file was received, which tests were run and documented, and whether the conclusion stays within the method. Up to 5 business days.
The IAE review of the opposing forensic audio report is for a party facing the other side's examiner on whether a recording was edited, whose voice is on it or what was said. The reviewer checks which file the examiner received, which tests were run and documented, whether the voice exemplars were comparable and whether the conclusion stays within the method. In IAE's own enhancement work, noise reduction is done on a working copy with every filter and setting logged, and a transcript marks speakers, time codes and unintelligible passages against the original file.
The document can take the form of a review of the opposing report, a rebuttal report or a statement of grounds for a Rule 702 motion. The decision on any motion belongs to the attorney and the court.
The client sends the opposing report and the case materials and describes what to check. The secretariat confirms receipt the same day, and within one business day IAE confirms the format, the cost, the deadline and the specialist from its registry. The client approves the candidate and the CV before work starts. The signed PDF is issued in up to 5 business days after the format and cost are confirmed, and a 48-hour rush is available for tight deadlines.
Specialists are members of the public IAE registry and passed the council's attestation under the published Criteria Matrix; the Member ID or QR code in the document opens the specialist's record. The specialist confirms in the document that he or she has no interest in the outcome, and materials are not passed to third parties without the client's consent. A review is evidence the court weighs with the other materials; neither IAE nor the reviewer guarantees the outcome, and IAE is not a law firm.
Sources
- SWGDE, Best Practices for Enhancement of Digital Audio, 20-A-001, version 2.0 (approved June 27, 2025; formatted August 21, 2025)
- SWGDE, Best Practices for the Enhancement of Digital Audio, 20-a-001, version 1.2 (September 17, 2020)
- SWGDE, Best Practices for Forensic Audio, 08-a-001, version 2.5 (June 9, 2022)
- SWGDE, Best Practices for Digital Audio Authentication, 15-a-001, version 1.3 (September 20, 2018)
- Fraser H., Aubanel V., Maher R. C. et al., Forensic Speech Enhancement: Toward Reliable Handling of Poor-Quality Speech Recordings Used as Evidence in Criminal Trials, J. Audio Eng. Soc. 72(11), 748-753 (2024)
- Fraser H., Enhancing Forensic Audio: What Works, What Doesn't, and Why, Griffith Journal of Law & Human Dignity 8(1), 85 (2020)
- Maher R. C., Audio Forensic Examination: Authenticity, Enhancement, and Interpretation, IEEE Signal Processing Magazine, March 2009, pp. 84-94
- Federal Rules of Civil Procedure, Rule 26, with committee notes (LII, Cornell Law School)
- Federal Rules of Civil Procedure, Rule 30, Depositions by Oral Examination (LII)
- Federal Rules of Civil Procedure, Rule 34, Producing Documents, Electronically Stored Information, and Tangible Things (LII)
- Federal Rules of Evidence, Rule 702, Testimony by Expert Witnesses, with committee notes (LII)
- Federal Rules of Evidence, Rule 705, Disclosing the Facts or Data Underlying an Expert's Opinion (LII)
- Federal Rules of Evidence, Rule 403 (LII)
- Federal Rules of Evidence, Rule 702, amendment history (Apr. 24, 2023, eff. Dec. 1, 2023), Office of the Law Revision Counsel
Read also
All articles
ShotSpotter evidence: is it admissible, and how does a reviewer test the other side's acoustic gunshot report?
An acoustic gunshot report tells the jury how many shots were fired, when and where, on the basis of sound picked up by sensors. Courts have admitted
The opposing audio report rests on a compressed file from a messaging app or a phone: what does a reviewer check?
Most recordings in disputes today reach the examiner as compressed files: voice messages, call recordings, memos made on a phone. Compression
Was the opposing audio examiner told what to find, and how does a reviewer check for contextual bias?
An audio examiner who knows that the suspect confessed, or that other evidence points to him, is working with information that the examination doesExpert report review
Ready to discuss your case?
Send the materials: the secretariat will confirm the format, cost, deadline and the specialist within one business day.

