Audio is not only important it is critical to any video production. This article highlights five low-budget digital audio recorders for video.

Exclusive Bonus: Download my free swipe file on how to shoot night scenes well (PDF file optimized for mobiles and tablets).

I base this list on the principles I’ve already covered in the Chapter on Audio in the Comprehensive Rigging guide. Audio is like Betty. No matter how pretty Veronica (your glamorous camera) is, and how much you lust after her, Betty is the real deal.

I’ve based my choices on the following criteria:

  • XLR inputs
  • At least two channels of recording with a minimum of 16-bit/48KHz
  • Uncompressed LPCM format, in WAV or BWF
  • Output to camera
  • Some form of support for Timecode, or better
  • Records to SD or CF
  • Budget less than $1,000

Here are my choices:

1 Tascam HD-P2

Tascam HD-P2

Approximate Price: $700

This is the only recorder on this list that fully supports Timecode via a BNC port, and it is my favorite choice among this lot. It has its quirks, and the general consensus is that it works better with a good mixer feeding it.

It records 2 channels in BWF to a maximum of 24-bit/192KHz and has 3 XLR inputs with Phantom power. It has a 1/4″ headphone jack, and not a 3.5mm TRS connector.

2 Tascam DR-60D

Tascam DR-60D

Approximate Price: $350

It records 4 channels in BWF to a maximum of 24-bit/96KHz and has 2 XLR inputs with Phantom power. It has three 3.5mm TRS connectors for monitoring, camera and line output. This is my second favorite from this list.

3 Zoom H4N

Zoom H4N

Approximate Price: $270

Is there any video shooter on the planet who hasn’t heard of the Zoom H4N? This is the go-to standard for DSLR shooters.

It records 4 channels in WAV to a maximum of 24-bit/96KHz and has 2 XLR inputs with Phantom power. It has only one 3.5mm TRS connectors for monitoring or line output.

4 Marantz PMD661 MKII

Marantz PMD661 MKII

Approximate Price: $660

It records 2 channels in WAV to a maximum of 24-bit/96KHz and has 2 XLR inputs with Phantom power. It has a 1/4″ headphone jack, and not a 3.5mm TRS connector.

5 Fostex FR-2 LE

Fostex FR-2 LE

Approximate Price: $600

It records 2 channels in BWF to a maximum of 24-bit/96KHz and has 2 XLR inputs with Phantom power. It has only one 3.5mm TRS connectors for monitoring.

My recommendations? I like the Tascam HD-P2, but if you can’t afford that, then opt for the Tascam DR-60D.

What do you think? Which is your favorite digital audio recorder, and why?

Exclusive Bonus: Download my free swipe file on how to shoot night scenes well (PDF file optimized for mobiles and tablets).

7 replies on “5 Low-budget Digital Audio Recorders for Video Production”

  1. Sareesh Sudhakaran SicofanteYou also missed that I won’t start all cameras at the same time or even have all of them turned on all of the time. I must save storage resources and the bigger SDXC cards are “just” 128GB. (The wide shot of the stage, which is the reference for the rest and it’s actually the only constant recording of the whole show, will be recorded directly into an external recorder.)

    I have used PluralEyes and it works perfectly well, except, like I said, for periods in the recorded material where there’s no “significantly informative” sound being recorded (room noise is useless, for instance, unless someone in the audience coughs a little… ;-)).
    I’ve been checking typical WiFi latencies and it’s mostly erratic, but it depends a lot in the hardware and protocols used. Since I can build a pretty cheap prototype for testing, I’ll let you know how it goes when I try.
    Still: do you have a list of inexpensive, portable time-code generators?

  2. Sicofante I missed the ‘events’ part of your message. That’ll need genlock. Without genlock you can’t guarantee precise timecode across all devices. Even if you somehow manage (how?) to start all cameras at the same time, they will drift apart after a few minutes. I don’t know of any other technology that will guarantee this.

    Once genlocked, all devices will record timecode according to the master device (camera, switcher or clock, etc.).
    You’ll need to test timecode over wi-fi to see if works. Each wi-fi device has a separate delay. Even the same device will have a slightly different delay depending on the network. I have no experience with multi-cam wi-fi, but I wouldn’t try it with today’s technology. Even with SDI there’s a delay, which is why they use genlock to force synchronize all devices and keep a check on them from drifting apart.

    I’m not a fan of PluralEyes. You can still make it work, I guess. But I don’t have any experience with it.

    Thank you for the coffee!

  3. Sareesh Sudhakaran Sicofante Just to clarify: I can’t use any form of slate once the event is started. The events ARE live, the editing of my recordings is not.

  4. Sareesh Sudhakaran SicofanteA slate is OK if the cameras are all turned on all the time, but not all of them will, only the one shooting wide to frame the stage. Some of the rest will be handheld and do specific, short shots (I just can’t afford every camera to be running continuously, if only for storage constraints). I’ll be left with many clips shot at different points in time. Proper time code would put them “magically” in the right spot on the editor timeline. Actually, I’ve used an add-on called PluralEyes (http://www.redgiant.com/products/all/pluraleyes/) that simply uses the audio recorded from each camera to automatically do the same (analyzing the audio signal), but it’s not reliable enough (or at all) when there are silences.

    I don’t quite get why I would need genlock. I never worked in live TV but isn’t that basically needed for syncing real-time video sources to switch between them? (I worked for many years in 3D and never ever used genlock at all. The only people who used it where those making real-time 3D graphics for live TV: sports, weather, virtual sets, etc.) I’m not doing any real-time switching at all. All the material will be edited (probably with Lightworks). I just need every frame recorded to get a timestamp. A timestamp, of course, which is a reference to an absolute “zero” time which is identical for all the cameras.

    Why exactly would a timecode sent over wifi and converted to audio -somewhere in the camera rig- not work? I don’t think wifi latency would be so big to make it fail, or would it? An in-rig wifi to audio converter wouldn’t produce enough latency to make it fail either, I believe? After all, if simple audio works so wonderfully to put every clip in place (when there’s actually some sound to analyze)… As a matter of fact, feeding one of each camera’s audio channels with random audio “broadcast” over short-range FM to every camera (and the audio recorder too), then using Pluraleyes on that material in post would be probably good enough. Not very elegant, though, and PluralEyes requires pre-processing; that’s why I’m looking for a “proper” solution.

    (PS: I’m using a pseudonym for privacy reasons. I’ve paid you a coffee, but the Paypal receipt will have no relation to my pseudonym, obviously. I hope I can help you with some more in the future.)

  5. Sicofante What you’re looking for is ‘genlock’, tone generators, etc. I have some info about these in the Comprehensive Guide. You should not rely on timecode from different devices unless they are genlocked.

    The easier way to do things is using a slate religiously. Actually, genlock is ‘easier’, but that means you’ll always need genlocked audio and video gear on set. What if you don’t? The slate method works on any production.

    The BMPCC cannot genlock. None of the BMD cameras can, even with the SDI input. I’ve written articles on wireless transmission, so you can check that out.
    Genlock is great for live TV productions, with numerous cameras that need to cut seamlessly, or for 3D work. Other than that, my advice to you would be to use a slate, or at least clap your hands or snap your fingers.

  6. Since I discovered this site, it’s like a constant flow of useful information. Thank you very much for making it.
    I want to record audio separately from video. I agree that this is the best way of doing things. I film with many cameras (mostly events), then edit as if it was a TV show, so I absolutely need perfect sync. The solution is obviously timecode. I can’t find “timecode workflows” here, so I’m sure it’s something of an easy matter to solve. Would you be so nice to write about how to generate, use, transmit, record, etc. timecode? Or if you have written already about it, can you point me to where?

    I’ll be probably buying a few BMPCCs and one of the recorders shown in this article. Now I need to know how to generate timecode, wirelessly transmit it to each camera and record it there (I’m thinking of using one audio channel for that, the other one for reference).

Comments are closed.