Magic Lantern and Canon's 2.0.4 firmware

Kirk Candlish · April 2nd, 2010, 03:35 PM

Quote:

Originally Posted by Tim Polster

I am waiting to buy any sound gear until all the news from NAB is out to see if there is any view of what might be next in the camera world.

A test to see if drift is happening is very easy. Record 12min of music with the 5D internal mic while simultaneously recording with the external device.

Put both audio files on a timeline, sync up from the beginning and play. Take note of when/if the two start to separate. Since you are recording music, I would test this just to make sure.

Yup, been there, done that.

I run ProTools HD in the studio and have checked the location files from the SD302/PDM661 by importing them into a PT session. No drift in 12 mins..

Tim Polster · April 3rd, 2010, 08:09 AM

Interesting. When I have used my laptop DAW for external recording I notice drift in between 5-7 minutes without my digital timepiece in the loop.

Have you tested this after the firmware update? True 30p does not have the drop-frame thing going on which would help it stay with normal recording devices. I would be interested to know if it drifted in 29.97.

I don't know if the new 24p framerate is 24p or 23.976.

Scott Bellefeuille · April 3rd, 2010, 11:02 AM

Quote:

Originally Posted by Bill Binder

Typically the headphones out of a consumer recorder (and for that matter even a good prosumer model) will suck bad (e.g., lots 'o noise). If your recorder has a line out you'd probably be much, MUCH better off than the surely POS headphone outputs. Just saying...

On the Zoom H4N the headphone socket is also a line out socket. Hopefully that means the circuitry is a little better than just a POS headphone amp. I'm not sure how it switches between headphone or line though.

Scott

Evan Donn · April 3rd, 2010, 04:43 PM

Quote:

Originally Posted by Tim Polster

Have you tested this after the firmware update? True 30p does not have the drop-frame thing going on which would help it stay with normal recording devices. I would be interested to know if it drifted in 29.97.

I don't know if the new 24p framerate is 24p or 23.976.

Actually with FCP the true 30 caused a lot of problems initially with drift - 2-3 seconds over a 12 minute clip. If you got your project set up properly for 30p though it wasn't a problem. Haven't tried it yet but I assume the new 29.97 should give you proper results with the default FCP settings.

And the new 24p is in fact 23.976.

Tim Polster · April 3rd, 2010, 07:00 PM

Evan, pardon me if you already know this, but your posts seem to be referring to FCP as the source of drifting. The drifting/sync issue I am speaking of is the speed difference between the in-camera audio and an external recorder.

All recording devices run at their own speed or internal clock. these speeds usually differ which over time shows up as drift, or a mis-allignment of the recordings over time. It is really the exact time the device considers to be a complete second.

If I have time I will run this test with a 5D and a laptop with a portable audio interface to see what I find out.

Evan Donn · April 3rd, 2010, 08:29 PM

Definitely - it's just that the most significant/common source of drift I've seen people reference with this camera comes down to problems with FCP and can be corrected with the proper settings.

Once your settings are sorted out then the drift you're referring to comes into play - and for me it hasn't been a noticeable issue at the 12 minute mark. Maybe it's just a lucky combination of hardware for me, or it could come down to the type of material I'm shooting. I'll run a test like you described with music though to check for drift.

Update: Here's a link to my test - this is a 5D at 24p and Zoom H4n. Video has three segments - the beginning sync point, the beginning of the song repeating at approximately 10:30, and a copy of the 10:30 clip offset manually by 1 frame. The offset version has a very clear doubling of the high hat and flanging on the voice, whereas the original version sounds very much like the sync start version to me - so if there's any drift it's far less than a single frame over 10 minutes and well below my ear's threshold to notice it.

http://smalldog-media.com/misc/5D_Sync_Test.mov

It'd be interesting to me to see how much this varies for other people, maybe I just got lucky with the particular equipment I got.

Steven Fokkinga · April 4th, 2010, 05:05 AM

Quote:

Originally Posted by Scott Bellefeuille

On the Zoom H4N the headphone socket is also a line out socket. Hopefully that means the circuitry is a little better than just a POS headphone amp. I'm not sure how it switches between headphone or line though.
Scott

Is there someone with a zoom h4n and 5d with ML who could test the difference in sound quality and noise between the recorder and the line out to 5d?

Tim Polster · April 4th, 2010, 07:33 AM

Thanks for posting Evan, that is good stuff. Without the offset example it would be tough to know there were two audio sources.

I will test today with my Firepod/laptop setup.

Mitchell Yazdani · April 5th, 2010, 11:36 PM

Has the latest version for 2.04 firmware released yet or still in beta form?

Thanks
Mitchell

Peer Landa · April 6th, 2010, 12:35 AM

Quote:

Originally Posted by Mitchell Yazdani

Has the latest version for 2.04 firmware released yet or still in beta form?

Related to Mitchell's question -- I'm working on a quite important shoot, and hence, have been afraid to upgrade. So my question is, how reliable is the 2.04 version..?

-- peer

Tim Polster · April 6th, 2010, 10:47 AM

Update after Audio drift test:

I tested about 13 minutes of music with a 5DMKII, EX-1 and a Firepod multichannel audio interface attached to a laptop via firewire.

Results:

The 5D and the EX-1 kept sync for the entire time, no issues
The 5D and the EX-1 drifted from the DAW to the point of snare hits sounding like flams around 5-6 minutes. Pretty bad.

This is normal to my experience with cameras and DAWs. I use a digital timepiece that allows a video input signal to determine the DAW's sync.

So the devices that have been already checked must be made with video sync in mind if they can avoid drifting for 12 or so minutes.

Phil Curry · April 7th, 2010, 05:12 PM

Is it possible with firmware or other routes to do a gen lock between two 5DIIs for use on a 3D rig? The 5D may not be the best 3D camera but it would allow a very compact rig.

Vincent Ogloblinsky · April 28th, 2010, 01:39 AM

I just test last night 0.1.8 version
It works well on tethering with usb. I was able to shoot some pictures on the camera and directly transfered on my Mac with LightRoom 3 Beta 2.
The big surprise was on focus feature.
After selecting 'NEAR', and going to the next line, i wasn't able to do calibration sequence with Zoom + button. No reaction.
The same works again well with 0.1.6, i tested again after discovering this issue, with the same lens, same parameters.

Anything wrong about that Trumbell?

Thanks

Regards

April 3rd, 2010, 08:29 PM	#51
Evan Donn Major Player Join Date: Feb 2006 Location: Philadelphia Posts: 795	Definitely - it's just that the most significant/common source of drift I've seen people reference with this camera comes down to problems with FCP and can be corrected with the proper settings. Once your settings are sorted out then the drift you're referring to comes into play - and for me it hasn't been a noticeable issue at the 12 minute mark. Maybe it's just a lucky combination of hardware for me, or it could come down to the type of material I'm shooting. I'll run a test like you described with music though to check for drift. Update: Here's a link to my test - this is a 5D at 24p and Zoom H4n. Video has three segments - the beginning sync point, the beginning of the song repeating at approximately 10:30, and a copy of the 10:30 clip offset manually by 1 frame. The offset version has a very clear doubling of the high hat and flanging on the voice, whereas the original version sounds very much like the sync start version to me - so if there's any drift it's far less than a single frame over 10 minutes and well below my ear's threshold to notice it. http://smalldog-media.com/misc/5D_Sync_Test.mov It'd be interesting to me to see how much this varies for other people, maybe I just got lucky with the particular equipment I got. __________________ My latest short documentary: "Four Pauls: Bring the Hat Back!" Last edited by Evan Donn; April 3rd, 2010 at 09:51 PM.

April 7th, 2010, 05:12 PM	#57
Phil Curry Tourist Join Date: Apr 2010 Location: Austin, TX Posts: 1	Magic lantern or ??? Is it possible with firmware or other routes to do a gen lock between two 5DIIs for use on a 3D rig? The 5D may not be the best 3D camera but it would allow a very compact rig.

April 3rd, 2010, 08:09 AM	#47
Tim Polster Inner Circle Join Date: Dec 2004 Location: Arlington, TX Posts: 2,231	Interesting. When I have used my laptop DAW for external recording I notice drift in between 5-7 minutes without my digital timepiece in the loop. Have you tested this after the firmware update? True 30p does not have the drop-frame thing going on which would help it stay with normal recording devices. I would be interested to know if it drifted in 29.97. I don't know if the new 24p framerate is 24p or 23.976.

April 3rd, 2010, 07:00 PM	#50
Tim Polster Inner Circle Join Date: Dec 2004 Location: Arlington, TX Posts: 2,231	Evan, pardon me if you already know this, but your posts seem to be referring to FCP as the source of drifting. The drifting/sync issue I am speaking of is the speed difference between the in-camera audio and an external recorder. All recording devices run at their own speed or internal clock. these speeds usually differ which over time shows up as drift, or a mis-allignment of the recordings over time. It is really the exact time the device considers to be a complete second. If I have time I will run this test with a 5D and a laptop with a portable audio interface to see what I find out.

April 4th, 2010, 07:33 AM	#53
Tim Polster Inner Circle Join Date: Dec 2004 Location: Arlington, TX Posts: 2,231	Thanks for posting Evan, that is good stuff. Without the offset example it would be tough to know there were two audio sources. I will test today with my Firepod/laptop setup.

April 5th, 2010, 11:36 PM	#54
Mitchell Yazdani Regular Crew Join Date: Sep 2009 Location: Sydney, Australia Posts: 79	Has the latest version for 2.04 firmware released yet or still in beta form? Thanks Mitchell

April 6th, 2010, 10:47 AM	#56
Tim Polster Inner Circle Join Date: Dec 2004 Location: Arlington, TX Posts: 2,231	Update after Audio drift test: I tested about 13 minutes of music with a 5DMKII, EX-1 and a Firepod multichannel audio interface attached to a laptop via firewire. Results: The 5D and the EX-1 kept sync for the entire time, no issues The 5D and the EX-1 drifted from the DAW to the point of snare hits sounding like flams around 5-6 minutes. Pretty bad. This is normal to my experience with cameras and DAWs. I use a digital timepiece that allows a video input signal to determine the DAW's sync. So the devices that have been already checked must be made with video sync in mind if they can avoid drifting for 12 or so minutes.

April 28th, 2010, 01:39 AM	#58
Vincent Ogloblinsky Tourist Join Date: Apr 2010 Location: Rennes (France) Posts: 1	I just test last night 0.1.8 version It works well on tethering with usb. I was able to shoot some pictures on the camera and directly transfered on my Mac with LightRoom 3 Beta 2. The big surprise was on focus feature. After selecting 'NEAR', and going to the next line, i wasn't able to do calibration sequence with Zoom + button. No reaction. The same works again well with 0.1.6, i tested again after discovering this issue, with the same lens, same parameters. Anything wrong about that Trumbell? Thanks Regards