1DC: Odyssey 7Q and the difference in picture

N

northwesternpat

New member
Shooting with just the 1DC not in Log the image from the back of the camera looks very different that what the O7Q shows. Is there any special tricks or workflow in order to tie these closer together? Right now the 1DC leaves HDMI to an AJA box and SDI to the O7Q. We haven't tried HDMI->HDMI yet and unsure if that would me a different.

The camera back lcd leans towards alot more warmer and reds while the O7Q is quite the opposite. Just looking for some 1DC shooters and maybe some tips and tricks that helped them. Downloading the manuals again to re-read. :)
 
You should do a comparison of the recorded files and let us know the results (internal MOV files and 7Q DNxHD/DPX etc.). The back LCD never shows accurate colour.
 
marvinhello said:
You should do a comparison of the recorded files ... The back LCD never shows accurate colour.
Click to expand...


+1, 'Word' and all that.

I implore you to get your edit system sorted using the PLUGE test so that you know that the contrast and saturation are set correctly, and observe your comparisons on it. Otherwise you'll be correcting issues that may not actually be there.

Just to get started, here's two articles on that:

http://www.glennchan.info/broadcast-monitors/monitor-calibration/monitor-calibration.htm

http://www.negativespaces.com/blog/...calibration-smpte-arib-and-white-balance.html

This whole 'blue gun' thing can be done by purchasing a sheet of lighting gel using 'Congo Blue' - I bought a quarter sheet from eBay. Don't use Colour Temperature Blue (CTB) - 'not blue enough'. Alan Roberts, ex-BBC and now EBU supremo recommends Congo Blue.

I'd also try and track down a well shot, well graded bit of 'known good' video that includes skin tones, a wide dynamic range, and strictly obeys the recommended tonality guidelines at the highest quality you can get. You need to 'normalise' your eyes to this on your monitor BEFORE you start looking at the other test clips (assuming that we're not doing Log exposure).

This all sounds a bit 'Chinese Tea Ceremony' but unless you know what 'good' looks like, one is - shall we say - 'evacuating in the wind'.

Hopefully you'll find that the Odyssey Q7's monitor is pretty darn close. Then you have to reverse engineer what your built-in LCD monitor looks like when the properly monitored shot is good. Sometimes that means it should look a couple of notches darker and cooler (certainly so on the C100).

I certainly wouldn't be sanguine making considered lighting or colour decisions based on the rear LCD panel. It's there for framing, focus and info.
 
Note HDMI to HDMI on the 7Q has some weird crushing effect. Which when you attach a real scope to it even shows as crushing blacks below zero and stretching highlights above 110. That maintains along the signal path. So when leaving the 7Q via SDI to a Panasonic Production Monitor and out to a waveform box. Both the production monitor waveform and the dedicated waveform box show that it's actually happening.

However, convert the HDMI from camera to HD-SDI and then go to the 7Q and levels are all back to normal.

So for now, still using the AJA box. Confirmed the same thing on the 7 with the Mark III. These were with the August firmware. All 8 7Q's we have have this same effect with the 1DC, C300 and Mark III.
 
Last edited:
The 7Q will be put to rest. I got a response from them and they are working on the HDMI Legalize video problem. Should be corrected in the coming future.

I will say I just shot 4K and proxied out to the 7Q. I noticed a slight green push from the Canon HDMI out. In fact in further testing all 4 1DC's have the same push. It was verified not just by looking at it but actually verifying the signal via our Leader LV 5330 Multi-SDI Test Monitor as well as the 7Q recorded the footage with that same slight push towards green.

I have googled the "Green Push" and came up empty. I find it hard to believe all 4 kits we have are just limited to our 1DC's or Leader Boxes. :)
 
I am pretty sure the 5D mk III HDMI signal also showed a slight green tint compared to the internally recorded footage.
 
Ninja Blade has an option called "5D Mark III input color correction, which also applies to 1D C, it basically brings the color space to Rec709 compliant.
 
I can't speak for the Odyssey as I've only ever used Atomos products but I noted a while back that any externally recorded footage had a slight green tint, less room in the highlights and at time clipped, sometimes more noise and also over 1/2 stop brighter. I contacted Atomos and they explained the uncompressed image from the sensor will always look different to internal as the camera processes the image extensively through compression. You would think then it would look better uncompressed!
I quiet like c-log externally recorded as it comes out like a flat usable grade and have used it barely graded (removed green and little contrast) a few times, looks nice :)
 
Sorry not best examples but these were the original frame grabs I sent to Atomos back in Feb 2013, first is in-camera second is Ninja. Note the clouds and roll off to clip point, noise under ceiling, green hue and also the very slight image crop. However as I mention clouds do take into account the 1/2 stop difference between the two.
 

Attachments

  • image.jpg
    image.jpg
    66.3 KB · Views: 0
  • image.jpg
    image.jpg
    67.7 KB · Views: 0
Last edited:
Wow that's a big difference in the highlights there. It's almost stealing a stop of dynamic range if you account for the cleaner shadows too.

Is this internal 4K MJPEG vs. uncompressed HDMI ouput in 4K mode to 1080p ninja? Or internal full frame 1080p vs full frame 1080p hdmi output to ninja? The output when in 4K mode is the best, much sharper and APS-H crop so it's a pixel-to-pixel readout, in s35 it's good too but not as good.
 
Last edited:
José Alkon said:
I can't speak for the Odyssey as I've only ever used Atomos products but I noted a while back that any externally recorded footage had a slight green tint, less room in the highlights and at time clipped, sometimes more noise and also over 1/2 stop brighter. I contacted Atomos and they explained the uncompressed image from the sensor will always look different to internal as the camera processes the image extensively through compression. You would think then it would look better uncompressed!
I quiet like c-log externally recorded as it comes out like a flat usable grade and have used it barely graded (removed green and little contrast) a few times, looks nice :)
Click to expand...

I can say the 7Q looks great on the C300.. But the 1DC and Mark III both have a slight green push coming out their HDMI into the monitor. It's also recorded as such which is a bummer. Pretty sure it's a canon thing and not a monitor issue. I asked Hurlbut recently about this and he's never noticed it on the Ninja Blade. Any chance that Atomos is auto correcting for the shift?

To us it's a real issue. Regardless if it's a flanders, leaderbox waveform/vectorscope, panasonic production monitors or the convergent design 7 or 7Q. It's a noticable green tint to the picture from the HDMI on those two devices. It's odd that nobody has really noticed it and I suppose if you shoot in C-log you may not notice it during shooting. But I would encourage someone to test in a picture profile (non-log) and see if they see these green HDMI push.

It doesn't exist on the C300 and perhaps because the HD-SDI doesn't have that fault as the HDMI does.
 
Last edited:
You must log in or register to reply here.
Back
Top