i like the idea of having the unit attached under the camera for monitoring, but it seems like a really elaborate device just to control the auto gain... now i'm wondering if it's worth it at all or if i should just go with an external unit? the audio is still, just okay... not great! please upload any other samples you might shoot and keep us up to date with your impressions.

thanks, great video.

I will do some more tests this morning.
The AGC is nice, I will try to show you what it looks like in Final cut and how easy it is to use.

One solution that I have been using quite a bit is a program caled FCP auxTC reader.

videotoolshed.com/?page=products&pID=26

This product can be downloaded and demo'd for free from this link
videotoolshed.com/images/demo/fcpAUXtcMAC.zip

I have the sound person send audio timecode (LTC) to the 5D. Once the files are copied to a drive, this application then auto-converts the audio timecode (LTC) files to real quicktime timecode and embeds into the movies. When the movies are dragged into final cut there is suddenly a real timecode that relates to your audio file timecode. The second audio pair could be just a reference audio track with AGC on (who cares its just reference).

Then you bring in the sound person's audio with timecode and its as easy as aligning the 2 identical timecodes. This eliminates the need for slates and all that old fashion junk.

Perhaps I will make a video on this.

I'd like to see a video on "FCP auxTC reader" I tried to get it to work, using the Digital Clapper Board to feed timecode into one channel of the 5D and a mixer for the other channel, but no go. I never could get FCP auxTC reader to read the files. I figured it was the audio distortion added by the 5D. Of course I was only using the demo so it may be crippled. I just keep syncing by hand, really not much of a problem.

Very interesting to me. I have been doing the same thing with my 5D, but using an MP3 player to create a high range tone in my 5D, and running that through my Juiced Link CX 231. It has preamps that will ad gain to the mic or tone side. I am not sure why you ran mic in the same channel as voice, but in my process, I find it works better to run the tone in one channel, and run the voice in the other channel.

Well, that was disappointing. The noise is clearly audible before and after the pilot tone is turned on. I loaded it into Soundforge and found that the noise before the pilot tone went on was 36dB left and 39dB right. After the pilot tone was turned on, the noise was only reduced to 42 dB left and 44 dB right. That's about the same as I've gotten with an iPod shuffle.

42 dB SNR just won't do. That's like 7-bit audio. Also, if you could really turn the 5D2 preamps down another 20 or 30 dB, you'd lose your signal too. To have a workable solution, we need to turn down the 5D2 audio, and we need another 20 to 30 dB of gain from a clean active preamp, like the juicedLink products offer. That will give us back our second channel as well.

With a juicedLink and an iPod, we can match these results today. Recording through the jucedLink into most anything with controllable gain, we can get far better results. Once we get the Magic Lantern firmware fix for the 5D2, the CX231 will be the right match. A passive front end just won't do the trick.

Hopefully, the