Using Zoom H4N or Sony TC-D5 ProII as preamp for Canon D5MII

[Greg Miller]

You have to be very careful with 3.5mm <> 3.5mm cables. Some are shielded, and some are not!

Local retailers, e.g. Radio Scrap, have at times sold unshielded cables that were apparently designed for headphones, where shielding is not necessary. Of course if you use one of these for a mic circuit, you're leaving yourself open to all sorts of noise pickup. Other times you might get a cable there that actually is shielded. It's a crap shoot.

Sadly, someone -- manufacturer, sales dept., package designer, retail outlet -- doesn't think shielding is an important factor, and so the packages are very rarely marked! If you get cables from a Lowest Common Demonimator retail outlet, you need to test the cable very thoroughly to be sure it's shielded, before using it for an actual gig.

Even with a shielded cable, if the circuit is unbalanced, you're only partly safe. The shield protects the internal signal wires from electrostatic noise, but the shield itself can still pick up electromagnetic noise, which is then coupled into the signal circuit.

Of course your best bet is to purchase a cable that is known to be shielded, from a dealer who is known to be reputable and who caters to people in the industry... not just to kids with iPoos.

[Pedanes Bol]

I don't know it is shielded or not, but this was the cable I used. This came with a cheap AT wireless microphone system I purchased a few years ago. I think I'll go ahead and buy the Sescom cable which is shielded.

[Greg Miller]

If that cable came with a wireless mic system, most likely it is properly shielded. The culprits are usually intended as rather long "headphone extension" cables, where shielding is optional

After re-reading a lot of this thread, and listening to your audio file, I agree with Jay Massengill. I think you are hearing noise from the varying circuit resistance caused by a dirty connection somewhere. Whenever any one of the connections (between mixer and camera) moves even slightly, dirty contacts can generate that kind of noise. Jay's cleaning instructions are good, and very important!

Also, the problem is greatly exacerbated if the camera supplies "plug in power" over the mic input jack. The voltage is typically between +5 and +9 volts. Even a tiny bit of changing resistance can produce a few millivolts of change in the DC voltage. Now a few millivolts (let's say 5/1000 volts) seems like a tiny amount. But that's very much comparable to the audio voltage coming out of a microphone. So the camera preamp boosts it nicely and it becomes very audible.

If possible, disable the "plug in power" if you're feeding the camera from a mixer's output.

If not possible, then either modify the mixer by adding a series capacitor ahead of the output jack, or else have someone build you an isolation box/cable with such a capacitor inline. The capacitor isolates the DC voltage to the camera end of the system, so changing resistance in the cable/connections will have much less of an effect. (The capacitor itself should cost less than a dollar; it's the connectors and labor that will run up the price.)

But even if you turn off "plug in power" or use a capacitor, it is still important to get all the connections perfectly clean!

[Bill Davis]

Originally Posted by Jon Fairhurst

Bill, the Canon 5D2 firmware has had manual gain control for some time now. Magic Lantern offers additional quality as you can control both the analog and digital gain independently. Canon has the analog at full blast and limits the user to controlling the digital gain only. Using ML, one gets roughly a 9dB S/N improvement.

AGC is still available. (I generally record with AGC ON when I want a sync track for an external recording.) But manual is available out of the box as well.

Fair enough. I haven't tried to use the camera audio on my 5d in the entire time I've owned it (nearly 2 years now) it since it was so awful originally.

I am interested, however in just how good the mic circuit in a DSLR could be. (my experience is that the standard video manufacturer adding an audio circuit to a video device is typically more adding a "feature" than designing for quality - but I could be wrong.

You indicate that even adopting the hassle of booting into Magic Lantern on every shoot to avoid the Canon EPROM settings, you can get an 9db improvement. But 9db improvement over what? Is there any reliable source you've seen that's done a properly weighted audio test of that system to reveal the S/N, noise floor and dynamic range measurements for a 5d?

I seem to remember that Ty Ford did a basic test and was NOT impressed. And that someone from the audio world (Perhaps Jay Rose?) did one as well as was even less impressed.

Before I'd ever use any camera for the single most critical information carrying signal I'm recording (my audio track!) I'd want to be sure I wasn't shooting myself in the foot with something that's baking in deficiencies in my recording.

I'd love to be able to change my thinking on this because on-camera audio is certainly more convenient than double system. But I'm leery largely because if somebody had found a way to make 5d on-camera sound even reliably close to the quality of double system, I kinda think that would have been BIG news in the industry.

Have you seen any reliable tests? And if so, where could I find them to review?

[Jon Fairhurst]

Years ago, I performed this test:

5. Canon 5D Mark II Audio Exposed - Noise on Vimeo

Since then, juicedLink has released new product with higher gain and Magic Lantern now allows you to turn off the low-cut filter.

Are the results as good as double system? No. It's a 16-bit recording, rather than 24-bits. And the results are crisp but slightly harsh. I think the A/D's anti-aliasing filter is poor - not unlike DSLR video.

My general rule is to shoot double system when I have a crew or when I need the best results. When shooting solo, I generally record into the camera. Why? Two reasons: 1) recording into the camera is simpler and less prone to mistakes, and 2) when shooting solo, I'm probably doing an interview in the wild where the sound quality into the DSLR is more than good enough for the application.

I use the following methods:
* For lightweight solo work in the wild, I go with the VideoMic Pro into the camera. The gain really helps. But it's camera mounted, for better and worse.

* For solo interviews, I use the juicedLink and lavaliers or a mic on a stand. The exception is at tradeshows. I use a wide lens and get close with the VideoMic Pro using a monopod. It keeps people from walking in front of the camera and by not asking to mount a lav, I give the subject one less reason to say "no".

* For a sync track, I use the internal mic with AGC enabled. It's simple, reliable, and good enough for syncing an external source.

The best solution always depends on the problem at hand. :)

[Bill Davis]

So, essentially the bottom line is that if you record audio to camera directly on a DSLR, you've got to be in a circumstance where you feel the convenience trumps audio quality to at least some degree.

Obviously, there are circumstances where the trade off is sensible. And if you're willing to jump through moderate hoops - you can get closer to optimal results than the stock system plus software provides.

Which I think is a very fair conclusion.

It leaves the judgement as to what is "acceptable" quality practices up to the individual shooter.

Which is OK by me.

[Jon Fairhurst]

I think that's exactly right, Bill.

BTW, the tradeoff hasn't been signal to noise, for me. The noise floor is down near the limit of 16-bit recording. (Of course this depends on the sound pressure levels, mic sensitivity, etc. If the signal is faint and you boost the gain in post, noise increases.) In fact, the S/N with a juicedLink preamp and Magic Lantern beats the results for a Zoom H4n or Tascam DR-100 (the original version, anyway.) No, the tradeoff in quality has to do with the slight harshness from a DSLR recording.

Externally, we use the Fostex FR-2LE, which has low noise, friendly operation, and a smooth, pleasing sound. The downside with the Fostex is that it's large and unwieldy for a solo shooter. It also has a plastic construction and doesn't feel robust for harsh environments. I don't want it flopping around while hanging on a shoulder as I'm running around trying to frame the image. When I'm indoors and can set it on a table, I'll use it solo when sound quality really matters. Other than that, it's best used by a dedicated audio person on a crew.

Is the harshness from a DSLR bad? That's hard to say. Given a controlled test, an audio pro won't like it. Most people won't notice it. Use it well in a real-world shoot with a good mic and it's going to be fine. But when sound quality really matters, go external. And use a crew.