Wednesday, February 15, 2017

Comments on field recording gear

This article references gear that is no longer available, with anachronistic prices. Nonetheless the same principles and conclusions are valid in July 2024.

This article will provide technical details and discussion for my Valentine's Day microphone comparison. First I will outline some requirements for field recording and discuss one useful specification. Then I'll discuss the specifics of the gear, typical pricing, and conclude with some remarks about relative value.

Hopefully this article will suit beginners as well as slightly more seasoned recordists.

Requirements

For field recording we want a sensitive microphone that has a low self-noise. That's because we are often recording ambiance and other signals in a low pressure regime. The requirements for live music or studio recording are quite different. For example, in those contexts it might be more important for a mic to withstand high sound pressure levels (SPL). That's not usually a priority for field recording.

When it comes to a recorder, portability is naturally the first constraint. A device must be usable in the field, so the form factor and interface are key. Battery power is a must. The correct microphone connectors and phantom power are important. Then, we want the best pre-amps possible. But how good is enough?

Happily there's an answer. The RANE Tech Note series has an excellent article, Selecting Mic Preamps that you should read for details. In particular, Table 3 is very useful. This allows us to determine the Equivalent Input Noise (EIN) of a microphone given two pieces of data that we can usually find on a mic's spec sheet: the self-noise in dB-SPL, and the sensitivity rating in mV/Pa. EIN numbers are negative, and the lower the better. (In other words, the higher the absolute number the better. Some people get confused by this.)

The total noise we hear is the combination of the microphone and the recorder's pre-amps, though this is not calculated by linear addition. The goal is to choose a recorder that adds insignificant noise to the microphone. The article mentions that a recorder with a 10 dB better EIN rating will increase the noise of the microphone by only 0.4 dB. That sets a useful benchmark.

Today we have many excellent choices when it comes to larger recorders, those that are designed to be carried over the shoulder or in a bag. But we are still a bit limited in terms of the hand-held form factor.

My own bias is towards relatively inexpensive gear. Though I am a trained professional audio engineer, I do not work as a researcher and artist. Hence I don't have a steady revenue stream that might fund expensive gear. Luckily, such is not necessary with a bit of ingenuity and research.

The opposite perspective is provided by those who wish to spend essentially nothing and still get excellent results. Be aware that there is a minimum outlay required for a decent recording. 

Hand-held recording setup

EM172 and EM272 capsules are popular with hobbyists looking to build small electret microphones. They are dirt cheap and low noise, plus they come complete with a FET. There's little to do except wire them up and put them in an appropriate housing. A cottage industry has sprung up around this capsule. As a result, you can buy nice products from Audiotalaia in Spain, LOM in Slovakia, and Micbooster in England. These mics are small enough to go anywhere. I carry a pair of clips to attach them to fences, wires, branches, etc.

EM172 mics have a sensitivity of 40 mV/Pa and self-noise 14 dB. This works out to an EIN of -106 dBu. The mics have a nominal operating voltage of 5 V but seem not to mind operating at the bottom of their range (specified as 3-10 V).

The Olympus LS-11 is a small (139 cubic cm and 165g) hand-held recorder about which I have written plenty... just search this blog. It provides 3V plug-in power (PIP). That puts it in the top category of portable recorders. The two AA batteries last something like 22 hours. The pre-amps have EIN of -122 dBu. It's tiny and the interface is easy to work with. Basically, it's a perfect little machine, which is no doubt why Olympus discontinued it!

Shoulder recorder

Sometimes I require more flexibility, and then I turn to the Zoom F8. Basically, this is equivalent to a Sound Devices 700 series, but with 8 input channels at a fraction of the price.  The Zoom F8 is 1346 cubic cm in volume, which makes it a tad bigger than an SD 702, but smaller than a Fostex FR-2LE, Edirol R-44, etc. It is slightly lighter than the SD 702 at 960g. The EIN of -127 dBu makes it eminently suitable for any mic you are likely to use in the field.

Phantom microphones

The Zoom F8 supplies a full 48 V phantom power, so we can use most of the professional microphones on the market. The choice is vast, but less so once we consider only low-noise and low-profile condensers. I tested two different sets.

The popular DPA 4060 is a tiny omni (lavalier) with Micro Dot output connector. This requires an adapter to XLR that is immensely larger and heavier than the microphones themselves. They have a sensitivity of 20 mV/Pa and 23 dB noise, for an EIN of -105 dBu. That's not terribly quiet, but the mics do have an appealing sound that has made them the de facto standard for field recording. Perhaps it's obvious from the test tracks?

The AT 3022 are omni small-diameter condensers that were released to market over-engineered. The spec sheet says their self-noise is 16 dB, but it was discovered that they measure at only 8 dB. When Audio-Technica got wind of this fact, they discontinued this model and replaced it with the AT 4022, higher-priced and rated at 13 dB noise. Back in the day I bought the very last AT 3022 units sold in North America. With a sensitivity of 19.9 mV/Pa, the EIN is an excellent -118 dBu. Few microphones at any price do better.

The AT 3032 are designed for studio use, but as they are compact, they work well in the field. I have a simple cross-bar and portable "desktop" tripod. This enables me to place them close to the ground.

Suitability of microphone-recorder combinations

Now we can apply knowledge we obtained from the Rane Tech note, by comparing the EIN of the recorder with microphone in each case.

The Olympus LS-11 has an EIN of -122 dBu which is more than 10 dB quieter than the EM172 at -106 dBu. This is a suitable recorder that will not add any audible noise to the mics.

The Zoom F8 has an EIN of -127 dBu which is more than 10 dB quieter than the DPA 4060 at -105 dBu. Again, all the noise we hear will be from the microphones.

However the AT 3022 are rated at -118 dBu, which leaves only 9dB difference. So, with these microphones the recorder might add about half a dB of pre-amp noise to the resulting signal. Not enough to worry about.

In any case, the test recordings I provided are unlikely to reveal the noise floor. The ambiance of the suburban landscape was enough to mask any microphone noise.

Hopefully this summary proved useful. Certainly it is not sufficient to compare specifications. That's why I started with a listening test. But specs can tell us a lot of useful information once we learn how to read them.

I should note that all the noise figures quoted above are A-weighted, since that's how the manufacturers do things. I should address this issue in a future post, because really it's a big problem.

Price comparison

The EM172 can be obtained as fully built mics for €70 to €100, or handy folk can make them themselves. The LS-11 was €245 brand new. I found my most recent unit for half that price, on the used market. The LS-10 and LS-5 are cheaper and almost identical. The total cost of this setup hovers around €200.

How good is the sound for that bargain basement price? Well, recordings I make have been played in concert and used by other artists. I would say that the sound is "very good" but not "excellent". Well, maybe you can tell from the sound comparison!

The Zoom F8 costs €1030, but you will want to invest in a heavy-duty LiPo battery and Hirose connector. The total comes to about €1200.

The DPA 4060 can be bought in a stereo kit with all mounts and adapters, under the product name "DPA d:screet SMK-SC4060" for €930.

You can buy the AT 4022 for €380 each. Cables, metal tripod, and stereo bar add €100. Total: €860.

If we average out those prices, we can say that a Zoom F8 kit with stereo mics (either set) comes to €2100.

Conclusion

How can we compare the hand-held recorder plus tiny microphones to the larger Zoom F8?

In terms of size, the F8 is ten times (!) the volume and about seven times as heavy (once batteries are taken into account) as the LS-11. That doesn't include the bulk of cabling, tripod, mount, etc. My full kit fills a rucksack despite being quite minimal compared to other recordists.

By contrast, the LS-11 plus mics fit into a pocket. I can carry them everywhere, all the time.

Comparing price, the F8 setup is €2100, ten times the cost of the LS-11 configuration at €200.

So, listen back to those recordings. Which one sounds ten times better? A rhetorical question, of course. But this illustrates a practical principle: diminishing return for your investment.

RELATED POSTS

No comments:

Post a Comment