Fostex DC-R302. You can't beat it for the price hands down. 24 bit, records to sd, can output to camera at the same time (easier syncing and still clean output), good battery life, readout screen with all the info including levels, compact, standard shoe mounting, 3 xlr inputs, can pan mics to differnent channels, and perfectly clean preamps.

Picking up a little debate that Trevmar and I were having about gain settings an the Zoom H1... I'll put it here because it applies to digital audio recorders generally and we don't already have a topic dedicated to the Zoom H1. Be constructive. Support your statements.

What's the goal in recording audio? It's to record your subject with acceptably low noise, I contend. And that's your only goal: everything else is secondary. I use 'noise' in the most general sense to mean any difference between the signal you're trying to capture and what ends up in your recording. I don't need to go into much detail about what noise is and how to measure it. It includes all kinds of distortion as well as white noise and quantization noise. Suffice it to say that there are different kinds of noise, and how much noise you can accept is personal and subjective. As for measuring noise, generally you are concerned about the noise level relative to the level of the signal (aka signal-to-noise ratio), since you can always adjust the volume of your recording after the fact, changing the noise and the signal levels by the same factor.

Rules for setting the gain in a digital audio recorder, in order of priority: These account for the most significant and unacceptable sources of noise in digital audio recording.

1. Set the gain low enough that you don't have clipping (leave adequate headroom). That's because clipping sounds nasty.
2. Set the gain high enough that you record the signal with adequate resolution (leave adequate "footroom"). That's because too low of resolution means too much quantization noise.
3. Set the gain to keep the recorder's internal noise acceptably low.

How does dynamic range relate? If the signal has more dynamic range than the recorder, you won't be able to satisfy all of the above rules. Having more dynamic range can help. Also, having more dynamic range can increase the range of gain settings that satisfy all of the rules, meaning that you wouldn't need to be as careful about setting the gain. But increasing the dynamic range won't always help. If the gain is already at the maximum and you are using your most sensitive mic, having more headroom will be of no help at all. Maximizing dynamic range is not a goal. It's helpful only to the extent that it helps you record your subject with acceptably low noise.

wipeman made some measurements of the Zoom H1's gain and internal noise for different gain settings, and calculated the equivalent input noise. http://zoomforum.us/viewtopic.php?f=15&t=15488

This information will guide some of our choices. Gain settings lower than 20 (or 16?) use digital attenuation, which means there is basically never any reason to choose a gain setting lower than 20. Gain settings lower than 20 do not reduce the likelihood of clipping, and they always result in greater quantization noise. Gain settings of 37-100 have the lowest equivalent input noise; higher than 1-36. A gain setting of 37 maximizes the dynamic range.

So what do you do with this information? You can minimize the recorder's internal noise by choosing a microphone with sensitivity such that a gain setting of 37 is the right gain setting for whatever subject you're trying to record. In effect, you match the dynamic range of the signal from the mic for whatever you're recording to the dynamic range of the H1 with a gain setting of 37. In practice this usually means choosing a mic with high sensitivity.

But should you always choose a gain setting of 37? Absolutely not, and it's easy to prove with a couple of counter examples. If you set the gain to 37 and you record a loud signal that has unacceptable clipping in the recorder, and a lower gain setting in the 20-36 range would prevent the clipping, you are absolutely better off with a lower gain setting. If you set the gain to 37 and you record a quiet signal that is just a little bit louder than the internal noise, and your signal is recorded with an effective resolution of only a few bits resulting in unacceptable quantization noise, you are are absolutely better off to use a higher gain setting. A higher gain setting won't reduce the internal noise, but it will reduce the quantization noise.

A gain setting of 37 is not magical. You should have a slight preference for a gain setting of 37 over settings of 20-36, if 37 does not result in clipping. But for quiet signals, higher gain settings are better.

@5thwall Sounds like it.

By the way, do you tend to have time to do post-work on the audio or do you have to use it all as shot? If you have time for post, just use the same e-mail I gave during our testing to get in touch and I will mention a few tricks that make a big difference in the sound quality (though you may already be using most of them, there is one I am sure you have not tried yet :).

@5thwall Makes sense. Best of luck getting enough work to be able to step up, though. :)

I'm still using the DR-680. It was nice to see that it fared well even in the Beachtek assessment, but for me the quality question was answered long ago. I've had some pro mixers listen to recordings I've done on the 680 and rave about it. I've also done a few recordings where I had to cut in sound from an H1 with sound recorded on the DR-680 and the noise floor difference is incredible. The 680 has a very low noise floor.

@5thwall In that case, my advice would be to buy a mid-level recorder ($100-300 more than a Zoom H4n) and sell the H4n to reduce the cost. If you find the right one, it will cut your noise level noticeably. I seem to remember there some that would cut the perceived noise level in half or reduce it even further.

I am not fully up to date on the recorders released this year, but there are some great tests done by Martin Doppelbauer, including ones that use the Zoom h4n against other recorders. Search for "zoom h4n noise floor" and his site should be one of the first results.

Remember that a 10dB difference "sounds" like doubling or cutting it in half (depending on the direction) even though it mathmatically doubles or halves at a lower dB value (6dB if I remember correctly, but I am very, very sleep deprived right now so you could look it up for better accuracy :)

@5thwall What is your signal-to-noise/dynamic range like on the Zoom vs going to direct to camera at the moment? What would you like it to be and how much do you want to spend?

The lowest signal to noise ratio for a given setup is just as influenced by the A/D converter and the microphone self-noise level as by the pre-amp itself.

The lowest possible noise floor on the GH2 is somewhere around -66 to -72 dBFS (that is with nothing plugged into the jack except an empty adapter) and that is depending on certain variables. If you microphone has a self-noise level far below that, you will not get the most out of it by going straight into the GH2 or into the GH2 via any analog pre-amp, but generally that is mainly the case for certain large capsule mics and not many small capsule mics go too far below it (some do, of course).

If you want maximum sound quality in a portable setup, look to combine a high quality digital system mic with a recorder that supports a compatible digital input (often SPDIF coaxial).

For reference, with a Neumann KM184D feeding a digital recorder with SPDIF coax, I was usually getting 77dBFS of dynamic range (closer to 90dB if I used a 70Hz low-cut filter) on real-world 48KHz recordings (as opposed to 66-72dB of dynamic range before you add in additional pre-amp noise or mic self-noise). ANd that was in my fairly noisy house.

Note that 16-bit audio has a dynamic range of 96dB (even that is usually reduced by dithering) while 24-bit has a dynamic range of 144dB. But a lot of gear does not come close to making the most of either.

@DrDave

Please insert empty line between paragraphs. I fixed it for now, but next time check it.

Audio Recorder Shootout

A Comparison and review of signal to noise tests of: Zoom H4n, Tascam DT-100 MKII, Sound Devices 702, Beachtek DXA-SLR PRO, Edirol R44, Tascam DR-680, juicedLink Riggy Micro / Assist

@shian what do you think of the DR-40 Its looks looks like it will do the job- Its cheap too. I like the dual recoding mode thing...Been thinking about buying one ...

http://www.amazon.com/DR-40-4-Track-Portable-Digital-Recorder/dp/B005NACC6M/ref=sr_1_1?ie=UTF8&qid=1342592028&sr=8-1&keywords=tascam+dr-40

@brianluce He's known as Baba Desi (The 'Wizard' of Belgrave) aka Desmond Bergen. He's an eighty three year old local identity and somewhat of a 'character'. I decided to make a short documentary about him as he generates quite a bit of interest.

He's reputed to have some kind of mystical healing powers. I can't say myself whether that is the case or not. ;>) I'll just spend some time following him around (fly-on-the-wall) and shoot him over several weeks. What the final result will be is anyone's guess.

Where's Louis Theroux when you need him? ;>)

A Fostex FR2 LE here also but with a Sound Devices MixPre-D for the front end. The pre's in the FR2 LE are good but the limiters, metering and headphone amp in the MixPre-D are far better. I use the MixPre-D's mic level out into the GH2 which provides a good signal for syncing later in post. Plus the MixPre-D can also simultaneously record direct to a laptop via USB.

Mics are a Rode NTG-3, Rode Pin Mic and two Oktava MK-012's. The Oktava's sound good but are extremely sensitive to handling noise and the slightest air movement. After much trial and error I was beginning to have doubts about using the Oktava's in anything other than very controlled environments.

After scouring many forums for a solution the consensus seemed to be that a Rycote Baby ball gag and Invision shock-mount was the most effective option so that's what I finally went with and it has made a big difference.

See attached BTS photo's of a recent shoot using a MK-012 fitted with a hyper-cardioid capsule...

A pair of AKG 414, and a Marantz 661 for piano performance, classical music concerts or other demanding situations. Absolutely no complaint sound quality wise.

Also have a pair of Pulsar II, and a Tascam DR-100, for field uses as both the mic and the recorder are smaller. The pre-amp of the DR-100 is a tiny bit noisy in very quiet situaions, but minor post production work can usually clean it up.