Audio Normalizer addresses a familiar production problem: a recording can peak safely yet sound much quieter than the next episode, while an aggressively raised file can become tiring or distorted. This Windows utility combines loudness measurement, processing and reporting. Its official description targets podcasters, audiobook producers, video editors and people organising music libraries. This First look is based on that description; we have not verified its measurements or listened to processed files.
The important distinction is between a loudness target and a promise of better sound. The app can help make delivery more consistent, but the appropriate result depends on the recording, the destination and how much processing you accept. A clean, well-balanced source still matters.
Based on official store information. We have not tested this app hands-on.
What it does
The listing describes integrated LUFS measurement using ITU-R BS.1770-4, true-peak measurement with oversampling and loudness-range analysis. Named presets provide starting points for podcasts, streaming, audiobooks and broadcast. You can also specify a target and peak ceiling yourself or derive a target from reference files. Before applying a preset, compare its values with the actual delivery instructions you received.
A practical workflow is to retain the original, analyse a working copy, choose a destination target and process that copy. Inspect the resulting curve and play difficult sections. The built-in player can jump along the loudness timeline, loop sections and compare before and after in sync at matched loudness. That comparison is particularly useful: a louder version often seems better simply because it is louder.
For dialogue, the developer lists a leveler, speech-only loudness measurement and noise-reduction steps. Batch processing, whole folders and watched folders extend the same approach to repeated work. Album mode and ReplayGain tagging serve a different purpose: managing listening consistency in a library. They should not be confused with producing a newly mastered audio file.
Reports can be exported as PDF, CSV or RTF, with loudness curves in the PDF. The description also names measurement information in Broadcast Wave chunks and ReplayGain tags for supported music formats. This creates a possible record of what was measured, but a generated report is not independent certification that a client or platform will accept the delivery.
Before you install
True-peak limiting and voice leveling are audible interventions when they work hard. Listen for pumping, altered consonants, noise rising between phrases and changes in musical dynamics. A curve or compliant number cannot tell you whether a dramatic pause still feels right. The listed markers showing heavy processing can help locate places to inspect rather than replacing the listening pass.
Format behavior also needs care. The developer describes lossless handling for some formats and amplification without re-encoding for MP3 and Opus unless limiting is required. That is a qualified claim, not a statement that every operation on every input is lossless. Check the output codec, metadata, artwork, chapter markers and video sync on representative files before processing a large collection.
A watched folder adds convenience but can also process an unintended export if your folder discipline is poor. Use a dedicated input location, separate output names and preserved originals. For video, the developer says the picture stays untouched while audio is normalised; we have not checked container compatibility or synchronisation.
Price and access
The linked US Microsoft Store listing asks $5.99, described as a one-time purchase without a subscription. It specifies Windows 10 or later and x64 or ARM64 support. Local storefront pricing can differ. The developer says files remain on the PC and describes no network access, advertising or data collection. These are developer statements rather than the result of our own privacy audit.
Our take
The appealing combination is analysis, an adjustable target, matched comparison and a retained report. That is a stronger basis for a delivery decision than raising volume until the waveform looks large. It could suit a creator who regularly prepares a season of episodes or a set of narrated chapters and wants a repeatable local process.
Our recommendation is conditional on your material. Start with one quiet passage, one loud passage and one representative full file. Confirm the output plays correctly in the intended player and meets the relevant specification. We have not assessed the accuracy of the standards implementation, speed, noise reduction or export preservation. Treat this as a source-based introduction and make listening, not the preset name, the final decision.
For a music library, first decide whether you want playback tags or permanently changed files. Those are different commitments. Keep a small tagged sample and verify that your preferred player actually honours the tags before adding them throughout an archive.
For spoken content, compare the same sentence before and after processing at equal playback loudness. Pay attention to breaths and room tone as well as the words. A voice can meet a numeric target while sounding less natural after excessive leveling.
If a client requests a report, establish which measurements and units they expect. A PDF containing many values is useful only when both parties understand the target and the exact file to which the report applies.
Sources
App icon from the official store listing.
