What this solves

Most Dahua audio complaints fall into three buckets: audio that was never switched on, audio that is technically present but unintelligible, and audio that howls back at the operator through a speaker. This guide walks through all three using the audio configuration described in the Dahua Network Camera Web 5.0 Operation Manual, so you can move from "no sound" to "clear sound" without guessing at settings.

The second half of the problem is the alarm tone. Many integrators assume the same file that plays on an alarm is the same thing as the live audio stream. It is not. The manual treats alarm audio as its own configuration page with its own file formats and its own recording limitations, and mixing the two up wastes a lot of commissioning time.

By the end you should know which parameters to touch, which to leave at default, and which symptoms point at the room rather than the camera.

How the workflow works

Dahua splits audio into two independent paths. The first is the live audio path: an input source is sampled, encoded, and carried inside the video stream you select. The second is the alarm tone path: a stored file is played when a trigger fires. They are configured on separate pages and they fail in different ways.

On the live path, the manual describes a sequence of decisions. You enable audio on Main Stream or Sub Stream, select a channel if the camera has multiple channels, then set compression, sampling frequency, input type, noise filter, and the microphone and speaker volumes. Every one of those choices affects intelligibility, and several of them interact.

The alarm path is simpler: you add a file, either by recording it on the camera or by uploading one, then select it for use. The manual notes that recording is only supported by select models and produces .pcm files, while uploads accept .pcm, .wav2, .mp3, or .aac. That asymmetry is the single most common surprise during handover.

Prerequisites and site checks

Before opening the web interface, confirm the camera actually has an audio input. The manual distinguishes between LineIn, which requires an external audio device, and Mic, which does not require an external device. If the model has no built-in microphone, selecting Mic will not conjure one — check the datasheet for the specific model before you promise audio on a bid.

Second, confirm the network path can carry the stream you intend to use. Audio rides inside the selected stream, so a heavily constrained Sub Stream will carry lower-quality audio alongside lower-quality video. If the recorder or client is pulling Sub Stream, that is the audio you are hearing.

Third, do a physical survey of the room. Note where any speaker sits relative to the camera microphone, whether the camera is mounted on a shared wall with a door that slams, and whether HVAC vents blow across the mic port. These are the conditions that make noise filtering necessary and, in bad cases, insufficient.

Finally, verify firmware and browser compatibility against the camera's own documentation. Menu layout and available codecs vary, and where the manual does not confirm a specific behaviour for your model, treat it as depending on model/firmware. For mixed-vendor estates, confirm the audio codec expectations of the receiving platform against the ONVIF profile in use.

Step-by-step configuration

The manual's procedure begins at Camera > Audio (Operation Manual page 43). From there, click next to Enable in Main Stream or Sub Stream, and for a camera with multiple channels, select the channel number. The manual frames this as activating or deactivating the audio acquisition function based on the requirements of the scene — in practice, enable it only on the stream your recorder actually records.

Next, configure the audio parameters. The table below reproduces the parameter set from Table 6-10 (Operation Manual page 44) with practical starting points. The manual explicitly recommends the default encode mode, and that advice is worth following unless you have a specific interoperability reason to change it.

Click Apply to commit. Then move to the alarm tone page at Camera > Audio > Alarm Tone (Operation Manual page 45) if the site needs triggered playback. Click Add, then either select Record, enter an audio name and click Record, or select Upload, click Browse to choose the file and click Upload. Select the file you need, and use the related operations to edit the name, delete, play, or download it.

Keep the format constraint in view: the camera supports recording audio files in .pcm format only, and recording is only supported by select models. Uploads accept .pcm, .wav2, .mp3, or .aac. If your model cannot record, prepare the tone offline and upload it.

Tuning and verification

Commission audio the same way you commission video: with a person standing where the audio matters. Walk the coverage area, speak at normal conversational level, and listen at the client. Adjust microphone volume in small increments rather than jumping to maximum, because a clipped input sounds worse than a quiet one and cannot be repaired downstream.

For two-way audio, test with both ends live. Feedback is a loop: the far-end speaker is picked up by the near-end microphone and re-amplified. The fix is almost always physical or level-based, not codec-based. The table below maps the symptoms you will actually encounter to their likely cause and the first check to run.

Record the final settings — encode mode, sampling frequency, input type, filter state, and both volumes — in the commissioning sheet. When a camera is replaced or a firmware update resets behaviour, that sheet is what lets you restore a known-good configuration in minutes. Where the manual does not document a specific behaviour for your model, note it as depending on model/firmware rather than assuming.

Buying notes and deployment fit

Audio capability is a specification line, not a given. Before specifying, confirm whether the model provides a built-in microphone, a LineIn input, or both, and whether it supports on-camera alarm tone recording. The manual's note that recording is only supported by select models is the detail most often missed at the quotation stage, and it changes what you need to prepare before the site visit.

Match the camera to the scene. Indoor interview or service-counter positions benefit from a model with a built-in mic and a modest sampling frequency; perimeter and plant-room positions need the noise filter and realistic expectations about intelligibility at distance. Browse the network cameras range to compare audio-equipped options, and check the WizMind line where analytics and audio are both required.

On the recording side, confirm the network recorders in the design can store the audio stream you enable and that the client software can play it back. Audio that exists only in the live view is a common and avoidable gap. If you are uncertain which combination fits a specific site, the fastest route is to ask before ordering rather than after.

FAQ

Where do I enable audio on a Dahua camera? Go to Camera > Audio, click next to Enable in Main Stream or Sub Stream, and for a camera with multiple channels select the channel number (Operation Manual page 43).

Which audio encode mode should I choose? The manual lists PCM, G.711A, G.711Mu, G.726, AAC, and G.723, and recommends the default value. The configured mode applies to both audio and intercom, so change it only for a specific interoperability reason.

What sampling frequency is best? The manual offers 8000, 16000, 32000, 48000, and 64000, and explains that higher sampling restores the signal more accurately. Start at a moderate value for speech and raise it only if bandwidth and storage allow.

Why does my two-way audio howl? That is acoustic feedback: the speaker is being picked up by the microphone and re-amplified. Lower Speaker Volume, separate the devices physically, and re-test before changing codec settings.

Can the camera record its own alarm tone? It depends on the model. The manual states the camera supports recording audio files in .pcm format only, and that recording is only supported by select models. Uploads accept .pcm, .wav2, .mp3, or .aac.

What is the difference between LineIn and Mic input types? LineIn requires an external audio device; Mic does not require an external device. Select the type that matches the hardware actually installed, or you will get silence or an unusable level.

Sources

• Dahua Network Camera Web 5.0 Operation Manual V1.3.1 — pages 43, 44, and 45 (audio parameters, Table 6-10, and alarm tone configuration).
• Dahua official site: https://www.dahuasecurity.com/
• ONVIF: https://www.onvif.org/