Any chance of an 'audio only' version of Magic Lantern?
 

Tramm...is there any chance of a version of Magic Lantern that only overides the audio?
I don't use the H4N but an Edirol R44 4 channel location recorder. Like much semi-pro gear the line-outs on it are on phonos at a nominal -20dB.
Setting the Magic Lantern firmware to have an analogue level of '0dB' allows the 5D2 to record any two of the R44's outputs cleanly via the ext mic input, which gives perfect 'backup and sync guide audio' on the camera's audio track.
However I also want to see an SD video feed on my small 2.5" Zigview monitor, which works well...but when the external monitor is plugged in, currently ML doesn't allow me to switch back to the camera LCD or viewfinder to check focus or help framing.
I don't use the camera's AV op for audio monitoring but listen to the R44.
Consequently I'm currently back with the Canon firmware and padding the audio input from the R44 to the camera and accepting the AGC bumps.
David
PostFade

I've created a feature proposal for hot plug support for A/V jack and HDMI. It won't be in 0.1.6, but likely in a future release.

This isn't too difficult to re-code the hotplug_task() so that we can select which output we want, but it will require some further work to deal with the aspect ratio and resolution changes between the different monitors. Ideally we'll also have the full HDMI issue issue solved at the same time.

I am interested in the 'Audio Only' ML as well.

I think audio control and monitoring is by far the most important need followed by HDMI 1080 output for focusing. Realistically, a monitor won't be attached to the camera most of the time due to portability issues, but it would still be very nice. Audio is definitely something needed on every video clip. A bug-free audio only ML would be a fantastic improvement over the stock 5Dii and would make this a true video camera in my book. European users most certainly want 24/25p so I would want to see that just so this camera becomes the tool that levels the playing field as it should be. Features like zebra bars are fantastic but I would put audio and reliability as my most important desires for ML.

In the config file in 0.1.5 , you can turn everything off, and set your audio only up there in the config file. You can shut down most other enhancements.

In 1.6, you will likely have access to dual channel settings, and be able to adjust digital gain different on each channel. It will also like provide some adjustment to the sound out put coming out of the AV jack out.

Thanks Tramm for the info on likely changes.
Chris you said:
<<In the config file in 0.1.5 , you can turn everything off, and set your audio only up there in the config file. You can shut down most other enhancements.>>

I set up my audio analogue gain to be '0' and whether to have zebras that way, but alas with 1.5 I can't change the loss of camera LCD and viewfinder once I boot with my external SD viewfinder plugged in.
Therefore currently if I want the ext finder I have to have it on always.

DaveT

From what I am reading, the config file is a fantastic feature. I guess ML already has what I want so I guess it is about time to start experimenting.

Yes, this is true for 0.1.5. 0.1.6 is slated to have some changes to that on board. The version should allow AV monitoring, as well as some volume control of the output jack. Also slated to be on board is option to activate camera mic if you are going to use it for a pilot track in double system.

Dave, I don't know if you have worked with camera sound and monitor out in with the standard firmware, but I have noted that there is an interference cause by plugging in monitors on the SD AV output.

It carries into ML, and appears to be a camera design issue. I am not sure if some kind of inline filter would help. I ran into this problem in April in shooting a short-- then without ML. Jon indicates its an easy notch filter to remove it. But it doesn't occur with HDMI out, so this is why I want to see Tramm's efforts on the HDMI out succeed. There he is dealing with multiple issues.

This is true, but it's very low. If you have a hot enough signal to run 0dB gain in the camera, you won't be able to detect it. With 17 dB of gain in the camera, it's down at -88 dB.

No. The video chip creates a 15,750 Hz horizontal sync signal, as well as a 7,875 Hz bump. It's at a very low level, but can crosstalk into the audio chip, since the video line is adjacent to the audio monitor lines. High gain in the camera or in post will amplify it. Run a low gain and it disappears.

If you must run a high downstream gain, add a narrow notch filter at 7,875 Hz in post, and it should disappear. Since the audio clock and video clock in the camera are likely on the same timebase, the frequency should be exactly 7,875 Hz. You can run a notch at 15,750 too, if you'd like.

In general, good EQ practice avoids narrow boosts, but allows narrow notches. Better practice is to run a hot signal and low downstream gain. :)

I was hoping Jon would chime in here and explain this better.

And Marcus-- I like the config file idea too. If you want a standard setup to use continuously, you simply use the config file and set up the configurable variable in note pad or other text editor. Tramm has a simple explanation of each variable in the config file too.