BebuLamar

I don't think it would help but did you try to put a resistor in series with the output of the mixer?

terskelton

That's what I started with: a resistive pad. It uses several resistors to create the proper load, source resistance and required loss. That's what has not worked, although it should.

TriX Loc: Raleigh, NC

How about just plugging an amp into the output of the pad and listening? If the signal is clean, then you know it’s the camera, assuming BOTH channels are showing up on the camera’s VU meter (it’s easy when using adapters to leave one channel open or shorted,) Then you’d want to set the camera gain manually for BOTH channels to close to 0 VU on peaks (leaving a bit of headroom). No way a resistive pad, whether terminated correctly or not, can cause this issue unless the signal level is being attenuated too much going into the camera (the VU meters should tell the story). One last thought: check Magic Lantern and see if they have a port to the T3 - it will add many features that are needed for video and audio recording. BTW, is your mixer providing phantom power on the audio line? An isolator (which is a transformer as you mentioned) or a large (large enough to provide low impedance at low frequencies) non-polarized cap will isolate the DC component. If you don’t have a large non-polarized cap available, you can use 2 electrolytics back-to-back. Please pardon me if I’m telling you things you already know...

terskelton

TriX Thanks. The mixer does have phantom power on the inputs but not the output of course and that's the connection feeding the camera. The camera however DOES have "mic power" present on the mic input. It's only about 1.5 to 2 volts dc but a flow of electrons does create noise. There's no way to turn it off though. As I said, I can try a cap there before the pad or transformer. As far a resistive pad making noise, exactly how a mic input is terminated will affect the noise of the mic pre-amp. So...no the resistor is not making noise but in effect making the following amp make the noise. In any case, your email was very helpful by it's reference to Magic Lantern. Basically I found the information there that says the mic input circuit on the EOS T3i is very noisy and has a limited usable range if you want to stay above that noise floor. IF we installed that program there is also some reference to DIGITAL gain as well as "analog" mic gain. Juggling those two might improve the audio in general. However, I'm not sure that benefit is worth the risk of installing this complex after-market un-authorized program. Their audio comments have given me ideas about how to approach this and for that thank you for the reference!

TriX Loc: Raleigh, NC

You are welcome. I wouldn’t be afraid of Magic Lantern it’s available for your camera. I’ve researched the “urban rumors” of ML “bricking” a camera, and I cannot find any credible instances. Believe me, I looked before installing, and I’ve used it for years with zero issues. You simply load it onto an SD card, and the cameras loads it at start-up and adds a host of useful functions, especially for video and audio recording. You involve the ML menus by pressing the Del button on the camera. You can uninstall it at any time by simply formatting the card.

terskelton

TriX. If you are familiar with Magic Lantern, perhaps you could tell me if there is one gain adjustment for audio that is digital and another gain adjustment for audio that is analog and the effective difference? I saw a reference to that but not explained. Thanks.

TriX Loc: Raleigh, NC

I’ve never used the audio functions on any of my Canon bodies running ML, so I can’t comment from experience, but I will review the ML documentation and give you my opinion shortly.

Cheers

Bobspez Loc: Southern NJ, USA

The best is probably to use a stand alone recorder and sync with the camera audio in the editor. Then you can mute the camera audio in the editor once you are sync'd with the external recorder tracks.

terskelton

Hi TriX,
Have you had the chance to look at the possible audio settings in the MAGIC LANTERN program? Thanks. Terry Skelton.

TriX Loc: Raleigh, NC

Here’s all ML has to say on the subject in the docs, but I actually think it’s pretty straightforward. I think that analog audio gain control changes the gain of the preamp between the mic/line input and the A/D while digital gain multiplies the digital output of the A/D by a constant. This is very similar to the way setting ISO works on a digital camera - changes in analog gain of the amp between the sensor and the A/D are used up to a certain point for ISO changes (typically up to around ISO 1000) while multiplying the digital output by a constant is used to adjust ISO beyond that point.

I would adjust the analog gain to get the VU meter to 0 VU on peaks and then make adjustments to digital gain to get the digitized output level you need to drive whatever will attach to the output. Make sense?

terskelton

TriX. You said: "I would adjust the analog gain to get the VU meter to 0 VU on peaks and then make adjustments to digital gain to get the digitized output level you need to drive whatever will attach to the output. Make sense?"
Actually it does not (to me). And I don't have any experience with digital gain affecting ISO settings. The manual section you excerpted doesn't refer to the digital gain as having anything to do with the output from the camera. It also says to run the analog input as low as possible if using an external device to drive it. That could be OK but in a typical analog situation doing that costs you headroom in the preamp stage. An analog gain control is almost always after some circuitry that can be over-driven. My trials have shown that their advice (run the analog setting low and drive it hot) to be the way to get away from the apparently internally generated noise. It just doesn't make sense. The trick to keeping noise low and headroom (available) high is proper gain staging. But their explanation doesn't tell me how and their approach isn't "standard." In typical equipment you run gain adjustments around mid-range to stay in the ballpark for good noise vs. headroom. However, I'm going to consult with a foremost digital audio expert friend and see if he can shed some light on my lack of understanding. I do appreciate your intelligent assistance on this and it has provided valuable insight that's driving me back to more testing. Thanks a lot.

terskelton

TriX, let me also mention that their drawing of an XLR adapter for the camera makes no sense if you are using a stereo mic or mixer. I am seeing more digital devices (laptops, digital cameras) that are re-configuring their audio connector (usually a single connector these days) as required. So one mini connector can be a line input OR a mic input OR a stereo OR a mono OR a mic and earphone. But their drawing would be for mono. And sometimes the device removes the dc "mic power" voltage if you tell it you need a line input. But I gather the ML doesn't mention that.

TriX Loc: Raleigh, NC

Are you using any type of compressor or limiter, and what type of audio content are you recording? Speech? Music? Wide or narrow dynamic range?

terskelton

In this particular case we're recording drums (like rock or jazz drums) but it could be anything. No compression at this point. We do that in post if needed. So when I say you can hear the hiss over the drums you know it's loud. Those on-line users who have used external mics with this camera seem to be feeding it a low line level signal or even a line level signal, which would normally be too hot for a mic input. That's what we're doing so far but I was hoping to find out why that's necessary and what (if anything) is going on when problems occur then this is fed from a restively buffered source (ie. a resistive pad).

TriX Loc: Raleigh, NC

Digital gain must refer to the output from the A/D since the data is analog before the A/D.

Regardless of whether we’re discussing your mic/preamp/signal processing/camera gain, the ultimate goal is to just set the MSB (most significant bit) of the A/D on the peaks (or maybe just 1 bit below MSB to allow some headroom) to use the maximum dynamic range of the A/D. If you don’t do that, there will be unacceptable noise.

You need to do testing with a known source and a measurement method. Find a suitable audio test application for your PC (a laptop is preferable because when on battery, there’s no hum except from pickup of the cables). Use the generator output as the input to your signal chain and look at the FFT of the output of the camera - then you can see the noise floor as you try different combinations of preamp and camera gain. You can also see distortion products as you overdrive the preamp, camera or the A/D. I think that as soon as you have good tools to do the testing, the answer will become clear very quickly, even if the answer is that the camera is an unacceptable recording device and you have to go with an external recorder and sync with your video in post. If you decide to take it to that level, I’ll be glad to post the name and version of the audio test SW I use.