When building a home monitoring "nanny audio cam" system, the main component seems to be microphones/sound pickups.

The qualities that I think I need for such microphones for this system seem to be:

  • Sensitivity to somewhat loud human voice.

    They should pick up human - including children - conversing in normal tones.

    I do NOT need super-sensitivity to pick up whispers/quiet conversations (the main purpose of the system is dealing with some kinds of arguments/tantrums/altercations involving the children, meaning that anything less loud than a quiet conversation is irrelevant). Having said that, I don't need something special that will exclude quiet conversations either, just don't care about them.

  • Ability to pick up voice/noise in a fairly wide area - one per room.

    Ideally, I'd like a single mike in a small bedroom or a large living room (25x30 ft?) each.

    If that's impossible, a separate concern is interference between same-room pickups.

  • Clarity.

    Very important. If there is shouting, I need to be able to tell what the shouting is about in detail, as if I was there, as opposed to merely the fact that voices are raised with unintelligible speech.

  • Look and feel.

    I definitely do NOT need the microphones to be hidden. Looking nice/ergonomic is a slight plus but nowhere near a major factor.

  • Connectivity

    Don't care about details, but ideally would be something that can be connected to a central recording hardware (PC-based). Does NOT have to be wireless, unless distances are somehow a concern for sound transmission quality over a given type of connector.

    All things being equal, a pickup with IP connectivity (Ethernet port) would be the best, but NOT at the cost of more important features.

  • Cost:

    I'm unsure of what my options here are, but preferably below $100/piece and ideally below $50/piece. However, I'm more concerned with functionality than price, if these price points can never fulfill my needs.

So, my questions are:

  1. What are the possible kinds of microphones that fulfill my needs that I should consider?

  2. What technical specifications should I look for in such microphones?

  3. What are the other factors I failed to consider in the list above given the stated purpose of the system (see background section below for details).

  4. What are the extra features I should consider in the microphones for such a system?

  5. Are my cost price points above even remotely reasonable for the features/factors I desire?


I'm looking into a project of building an audio home monitoring system. Most of existing security/monitoring systems I'm aware of seem to be video centered, and either lack audio, or have very poor audio quality.

The system would be for "nanny cam" type of situation monitoring - e.g. ensuring real-time or post-factum that no bad behavior occurs verbally in the house when other people with lower level of trust are there - e.g. nanny/babysitter and children.

The main purpose of the system is dealing with some kinds of arguments/tantrums/altercations involving the children, either for me to be able to call in real-time to intervene, or to investigate exactly what happened later.

I am planning to ask a series of questions related to building such a system, and on Meta it was indicated that such questions are on-topic

  • (not really on-topic): You may want to look into legalities, in some places all parties need to consent to audio recordings. Not a deal-killer as of course you can get them to consent as part of hiring them, but something you'll want to be aware of.
    – derobert
    Jan 30, 2013 at 20:45
  • @derobert - ack. As I kinda-alluded to, this is not meant to be a hidden/secret system by any means.
    – DVK
    Jan 30, 2013 at 20:48

3 Answers 3


Honestly, my answer to this question, in which you infer that you plan to rely primarily on audio, is "do it any other way than you are planning".

Here's why; I work for an alarm monitoring company. In our line of work, where every signal from an alarm panel could be snot-nosed employees playing around or a matter of life and death, it's video, not audio, that determines whether we call the cops. If you have live video, you can see a problem; you can see the masked man with a gun, or the belligerent drunk accosting people outside the front door. Or, you can see that there isn't a problem; what sounded like a violent argument or gunfire on the audio feed, if you have one, was the TV in the breakroom turned up just a little too loud on the Spike channel.

If you have audio, you hear... raised voices, maybe. More likely you'll hear a lot of environmental noise; condenser mics, which are pretty much a must for "ambient" micing (further then 2 feet), will pick up a lot of noise, even in a relatively quiet setting like your home. If the TV's on, or the microwave is running, that's all you'll hear and it will drown out even loud conversation. You'll also get a lot of false positives from someone yelling or shooting a gun on TV.

Another reason is the lack of "graceful degradation". If the customer's internet connection is slow or they're using most of it when we load up their video feeds, we'll get a lower framerate. Annoying, but we can still accurately assess what's happening with as few as 1 frame per second per camera. Most audio codecs on the other hand, especially live streams, simply have to stream continuously; more than a second of "buffer" is unacceptable because the sounds won't be anywhere close to the images, and less than that is next to useless if the Internet connection is even slightly herky-jerky. For this reason, we don't bother with VoIP solutions for audio, unless the customer insists on it and has a reliable, high-upload-bandwidth Internet pipeline available.

That's why monitoring systems are so video-centric; cinema-quality dialogue recording is a technical impossibility outside of a sound stage, and what you hear on audio isn't necessarily happening in the room at the time. We install audio systems costing thousands of dollars at our client locations, and have exactly these same problems. You won't solve it for $100/mic hooked to your computer.

So why do we install audio monitoring at all? One word; interactivity. If everything seems to be OK, we can call in over the intercom feature of our audio systems, and verify audibly with the customer that everything actually is alright. The customer, and anyone else in the store at the time, now knows we're watching them. We can, however, perform the same verification with a phone call, and such "voice-downs" always, always follow a visual scan of the situation on-site via the cameras; calling or voicing down to a site when you don't know what's going on can get people killed.

Now, as a supplement to video, a condenser microphone is acceptable. Most video camera systems have at least one audio input. I would recommend an omni boundary mic; the omni capsule will mean there's no "proximity effect" where you gain or lose bassiness depending on the person's distance form the microphone (which can greatly affect intelligibility). Boundary mics are used for a wide variety of applications, including being excellent for videoconferencing as they'll pick up anyone around the table who talks.

If this mic can be placed a short distance (within about 10 feet) from an input to a system that can send you the audio (probably your computer), then an unbalanced line between them should be fine. Most computers expect an unbalanced "computer mic" or a USB mic, like this CAD U7 computer boundary mic; these use unbalanced circuits for signal. A powered USB hub will act as a rudimentary repeater, but you'll still be limited to about 20 feet from the computer tower. You could also invest in a "professional" boundary mic with XLR connections, like this Shure MX393-O, and hook it up to your computer's line-in with something like this: Pyle-Pro 2-Channel Mixer - It says 2-channel, and technically it is, but basically it is little more than a microphone preamp with phantom power and rudimentary gain controls, and will serve as a converter between the mic's XLR connector and your computer's line-in or USB audio input. The mic is $200, the amp is $75. If you want more mics in more places, you can invest in a rackmount mixer like this ART MX821 that can handle up to 8 microphones and output one combined mix of them into your computer (line-in only; no USB function with this particular unit).

Understand that all anyone has to do is turn off or unplug any part of this "net nanny" system and it's dead. The fact that this is not intended to be a secret system means you'll have to consider what to do when, not if, this system gets intentionally powered off or disconnected by someone. We get around it by making as much of our AV systems physically inaccessible as we can; you'd need a ladder, a screwdriver, and more time than it's worth to get to a mic or camera, while you'd need a key, or a crowbar and more time than it's worth, to get into the cage containing the recording computer. Despite this, we still have to put some pretty beefy UPS units on all the core components of the surveillance system and network infrastructure, in order to keep the device sending us data in a power failure long enough to determine whether it's a service failure or someone trying to keep us from seeing what's about to happen next.


KeithS is correct on most likely going with a "condenser" microphone.. As he mentioned it may pick up additional "noise" and such as this type microphone is designed to basically pickup any sounds with-in a pretty wide field... But, they have gotten pretty good and "setup correctly" (for your situation) can be perfect for discerning "noise" from the audio you are looking for. Especially if you are setting it up for particular things such as "loud outbursts" or "altercations.."

I would specifically recommend possibly a stereo condenser microphone and even better would be one (for your situation) with a USB connection... Or (since hiding the device is not an issue, possibly an advantage (?)) something like Samson - SAMTR Meteor Mic (which can be found at Best Buy) http://www.bestbuy.com/site/Samson+-+SAMTR+Meteor+Mic+With+USB+Cable+and+Pouch/1950242.p?id=1218303699398&skuId=1950242

(As far as the stereo condenser... this would give you additional options and clarity... also may be the answer in determining different sides of a discussion and help determine who did/said what...

The Reason I recommend USB is because, if you are developing this as a system.. I imagine a computer will be involved and most likely some software (which I imagine would be custom and thus the "actual product" in the package...). This would give you the best bus/interface to "add to/connect" to you monitoring software and has voice recognition capabilities... Which is where I see this project going and the point... Even opens possibilities to maybe add features like "text-transcriptions" of any outbursts, etc. The options would be endless.. even reading/setting EQ, Intensity/Velocity, and Db's to accomplish what you are trying to achieve... And, as mentioned with a stereo mic, things such as help in determining different sides of a discussion, this may be the answer...

I hope this helps...

As far as my opinion on this idea... I like it, but have no comment.. As whether it's worth trying, was not the question...


From a Home Security/Automation point of view the ELK M1 (or NESS M1) allows you to place their microphones throughout your home and connect them back to the M1 Alarm Panel in which you can listen into via the internet or by dialing over the phone.

Not only can you listen in you can also enable two-way communication in which you can listen in but also heard.

Here are some links:

Microphone/Speaker - http://www.elkproducts.com/product-catalog/elk-m1twsf-flush-mount-two-way-speaker-microphone M1 Panel - http://www.elkproducts.com/product-catalog/m1-gold-cross-platform-control

  • While it meets the needs of the OP, I think my warning stands; letting someone in the location know that they are being monitored, without knowing who that person might be or what they've already done, gets people killed. An intruder in the home can be silent, and if audio is all you have then you have no way to verify there even is an intruder. He cannot, however, be invisible to a video system.
    – KeithS
    Jan 24, 2014 at 17:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.