In the past, recording toddlers involved mounting a wireless mic onto the toddler using a custom vest or trying to capture their voices with a mic on an adult or fixed in the environment. Portable minidisc recorders are now small and reliable enough that they can be mounted directly on the toddler, creating a portable recording setup that is entirely self-contained (no wireless link). Moreover, a servicable recording harness can be assembled quickly, at low cost, from commonly available components.
So far, I've only been able to find one reference to someone mounting minidiscs directly on toddlers, at Wisconsin.
The following design is a prototype, robust enough for experimentation and casual recording under supervised indoor conditions. It can be used under other conditions (e.g. outdoors, mealtimes, active play) but with increased risk of destroying the recorder. It has been tested with one reasonably normal toddler: other toddlers might devise some new way to disrupt its operation. Details of the design may need some tweaking and it may need to hardened for more demanding use.
Here are some pictures of the prototype in use. Our attractive model is about 34 inches tall and about 25 pounds, i.e. slightly tall and average weight for his 18 months.
|
|
|
|
To build this, you will need the following off-the-shelf items, plus a sewing kit and electrical tape.
This prototype was assembled in January 2004. The exact items specified here will quickly become unavailable. However, you should be able to find similar items that will work.
The toddler harness has a 2 inch wide chest strap which fastens with velcro in the back. Straps run over the shoulders with length adjustments in the back. We will be using just the harness, without the removable leash. It is designed to be attractive, sturdy, adjustable, easy to clean, and impossible for a toddler to remove.
The Case Logic case is a 4.5 by 4.5 inch neoprene case, closed with a zipper. It is intended to protect the minidisc from modest bumps and limited encounters with water (e.g. while jogging). It is only barely large enough for our purposes: a slightly larger case would be desirable. It has a wide strap sewn across the back, allowing the case to be slid onto a belt up about 2 inches wide.
The microphone is a small omnidirectional mono mic with a 1/8 inch plug, which receives its power from the minidisc recorder ("plug in power"). The microphone was simply a basic default option that I could obtain quickly . Substitute whatever kind of mic you prefer. It was ordered with the shortest available cable length, which ended up making it about 3 feet long. 12-18 inches of cable would have been sufficient.
This minidisc recorder is able to record successfully from this microphone without an additional battery pack or preamp. This may be partly because it has two "microphone sensitivity" settings, so that it can be adjusted to pick up low amplitude signals properly. My older minidisc recorder (Sony MZ-N707) seemed to have trouble producing good recordings directly from the microphone.
Assembling the harness from these components took me about 20 minutes. For initial experimentation (especially to adjust microphone position), sewing can be replaced by strategic use of electrical tape.
Step 1: Snip a small hole in the hinge side of the neoprene case, just large enough to slide the microphone cable through. Be conservative: neoprene is stretchy.
Step 2: Slide the neoprene case onto the free end of the harness's chest strap. Adjust its position so that the case will sit as close to the middle of the toddler's back as possible, without interfering with the velcro closure. The hinge side should face towards the part of the strap that goes around the front of the toddler. Sew the case's belt strap to the harness chest strap.
Step 3: Remove the tie clip and the windscreen from the microphone. (The windscreen is a choking hazard.) Sew the microphone in your favorite position on the front of the shoulder strap nearest to the neoprene case.
I left the microphone uncovered because I could not figure out a way to make a waterproof cover. A less-than-waterproof cover seemed worse than useless, because it would impede cleaning the microphone when it did get hit with liquids or semi-liquid foods.
Step 4: Sew the microphone cable along the shoulder and chest strap, to the neoprene case. Push the rest of the microphone cable into the case and cover the hole in the case with electrical tape.
Step 5: Put the minidisc recorder into the case. Attach the microphone cable to it, using the right angle adaptor. The adaptor allows the microphone's plug to run along the side of the recorder. Without the adaptor, the microphone plug sticks out perpendicular to the recorder, which makes it more likely to be damaged or pulled out and prevents the whole assembly from fitting inside the neoprene case.
Step 6: Coil up any surplus microphone cable inside the case and secure as necessary. (This particular model of case has an interior pocket into which the surplus cable can be tucked.)
Step 7: Zip up the case. Secure the zipper with a small safety pin perpendicularly across the zipper teeth.
With this particular model of toddler harness and neoprene case, the case ends up being mounted upside down. Therefore, unzipping the zipper even a modest amount would cause the recorder to fall out the bottom. So it's critical to secure the zipper.
In mono mode, the minidisc recorder can record up to 160 minutes of high quality speech onto a single minidisc. That's probably more than you felt like transcribing. Unfortunately, for political rather than technical reasons, the recorder's data cannot be uploaded directly to a computer. Rather, you need to pipe it out the headphone/line output and redigitize with a computer.
Mounting the minidisc on the todder's back offers several advantages. It is almost impossible for the toddler to fiddle with it. It is out of the way of common activities such as scrambling onto chairs and toileting. And it is one of the least likely places to be bumped, fallen onto, or get hit with food or water.
The neoprene case cushions the toddler enough to make it reasonably tolerable for diaper changes. It will probably protect him reasonably well from bruising should he fall onto the recorder. Not perfectly, but remember that these kids regularly bruise themselves running straight into furniture and tripping over toys, so the "background level" is pretty high.
The microphone is exposed and vulnerable to damage, especially from liquids. I see no way around this and I assume that it will need to be replaced every so often.
Children may object to the harness if their family has previously used it as a leash. This is probably a minor issue, because only a minority of families use leashes. But plan in some extra time convincing some toddlers (e.g. mine) that this isn't really a leash.
This rig has four major limitations. First, you cannot monitor recording quality. This is somewhat obviated by the minidisc's ability to automatically adjust recording level and the fact that it has a "hold" feature to lock the buttons during recording. However, you're up the creek if recording wasn't properly turned on or the microphone jack gets pulled out. A secondary backup recording by other means might be wise if lost recording would create major problems.
Second, it doesn't protect the minidisc recorder (or in some cases the child) well enough for very demanding conditions. These include water play, snow play, messy feeding situations (e.g. learning to use an open cup), bathtime, rough play (e.g. with friends or siblings), prewalkers and unstable walkers likely to roll around on the floor and/or fall on their backs, etc.
Third, the microphone is unlikely to work properly if the child is fastened into a carrier (e.g. a backpack style carrier). Similarly, it probably won't work inside a jacket and, thus, the harness must be taken off and replaced outside the jacket.
Finally, the minidisc recorder seems not to work in cars, perhaps because it is unable to tolerate the vibration.
Thanks to Brian MacWhinney for encouraging me to try building this. He is not, however, responsible for any of its flaws.
Margaret Fleck, Hewlett-Packard Labs, Palo Alto, mfleck@uiuc.edu