Setting gain for mic and line in on Sound Devices 302

[Brooks Harrington]

The XLRM cable that comes with the G2 RX is for Mic level inputs.

The 302 has 60dB gain on the preamps, plus another 15dB gain on the fader.
Use the fader for more gain if you need it.

[Pietro Jona]

john

i would stick to this when using the mixer:
-input: mic, setting the receiver's af out at -6
-output: line, camera pots all the way open (as you said people experimented this and it is ok)
as for reading 0 at mixer's levels and -12 at camera's even after setting the tone at -20, this sound strange to me. are you reading vu only or vu and peak? i would check that although the vacuum is supposed to be a source of constant sound. during normal recording, like dialogs or interviews, the difference in reading vu only and peak is dramatic.
best

pietro

[Steve House]

I might mention that an invaluable tool for doing this sort of testing is a "tone plug" or tone generator. I've got one of these ... http://www.trewaudio.com/store/produ...&cat=48&page=1, http://mixonline.com/mag/audio_gtc_industries_tone and it's far better than a turning on a vacuum (gotta give props for creativity with that one though) or humming when you need a constant level tone. Behringer's cable tester also has a sine wave tone generator built-in.

Would be interesting to plug one of these into the Canon's XLR mic input with the input level control set to "Auto" in order to discover the meter reading that the camera's designers felt represented the optimum recording level, then switch to manual mode and line level, replace the tone plug with the 302 sending 0VU tone on its meters, and see where you need to set the camera's input pots to produce the same meter reading.

Note that a VU meter and a peak meter read the same when they're fed a steady sine wave tone at the same average level. The differences come out when you feed a complex waveform such as speech and the degree to which they differ is dependent on the mix of frequencies and the dynamics of the waveform.

[A. J. deLange]

Originally Posted by Steve House

Note that a VU meter and a peak meter read the same when they're fed a steady sine wave tone at the same average level.

Could you try your tone plug on that and let us know what you find (unless you already know the answer)? A VU meter is to be calibrated to read 0 when connected across a 600 ohm load driven by a 0.7746 volt rms sin wave at 1 kHz. An average of 1 mW is delivered to the load under these conditions. But the peak voltage of such a sin wave is sqrt(2)*0.7746 = 1.09545V i.e. 3 dB higher than the rms voltage. If a peak meter really responds to the loudest sample it will read 3 dB higher. If it's "peak" measurement is really an average over a short period (a cycle or so) then it will read the same as the VU meter.

[Steve House]

Originally Posted by A. J. deLange

Could you try your tone plug on that and let us know what you find (unless you already know the answer)? A VU meter is to be calibrated to read 0 when connected across a 600 ohm load driven by a 0.7746 volt rms sin wave at 1 kHz. An average of 1 mW is delivered to the load under these conditions. But the peak voltage of such a sin wave is sqrt(2)*0.7746 = 1.09545V i.e. 3 dB higher than the rms voltage. If a peak meter really responds to the loudest sample it will read 3 dB higher. If it's "peak" measurement is really an average over a short period (a cycle or so) then it will read the same as the VU meter.

Yep, been there, done that and repeated the experiment as I write this. With my SD442 mixer, fed Tone Plug's 1kHz sine wave into input 1. Pan channel equally to L & R outputs. Fader to mid-point (0). Meter in VU mode. Adjusted coarse gain to produce steady state meter reading of 0VU. Toggled meter ballistics between all 4 options (VU, Peak, VU+Peak, Peak Hold) using front panel switch without touching any other settings. No change in meter indication in any of the 4 modes. However repeating the procedure replacing the tone source with a couple of regular microphones (A/T 3031 cardioid, Schoeps CMC641 hyper) and using my normal speaking voice, placing the meter in VU/Peak mode showed peak indications at about + 6dB when the VU section of the meter was hovering around 0dBVU

[A. J. deLange]

At first I thought that quite interesting then I realized that the next LED above 0 on the SD "meters" is +4 while the peak in a sin is only 3 dB above its rms (a sinewave is not very peaky at -1.5 on the peakiness scale - not dB). OTHOH voice is quite peaky (sample I looked at recently was at +12 on the peakiness scale). It just makes sense that the time period for a peak in the digital world would be 1 sample. No change would imply a window of at least half a millisecond (24 samples at 48K)

Thanks anyway!

[Steve House]

Originally Posted by A. J. deLange

At first I thought that quite interesting then I realized that the next LED above 0 on the SD "meters" is +4 while the peak in a sin is only 3 dB above its rms (a sinewave is not very peaky at -1.5 on the peakiness scale - not dB). OTHOH voice is quite peaky (sample I looked at recently was at +12 on the peakiness scale). It just makes sense that the time period for a peak in the digital world would be 1 sample. No change would imply a window of at least half a millisecond (24 samples at 48K)

Thanks anyway!

On the 442 meters the LEDs represent 2dB steps from -14 up through +20 so the first LED above 0VU represents +2dB. But be that as it may, all I can do is report my observations. When carefully setting the channel gain and then the fader on a sine tone so they're set at the point when the 0 LED is illuminated and the +2 LED is just extinguished, there's no change in meter reading when switching the meter from VU to PPM ballistics - the +2LED remains dark. What else can I say? <grin>

[A. J. deLange]

My comment about 4 dB steps was drawn from the manual for the 302. Given 2 dB steps I go back to "quite interesting". For PPM the integration time constant is 5 ms so the observation is consistent. But for peak? There must also be a short time constant implied for that mode.