EOS 1D C firmware upgrade further delayed until end of May

Well, 25p working fine here and no issues.

Also no strange blocking in Canon Log that I've seen in some clips that were released early. But I suspect those were shot at ISOs below 400. Might have to go back and check that...

I simply think that Canon need to coordinate the hardware changes that are unofficial up to this point. No meaning in releasing a half-baked firmware 2-3 weeks early only to have to send those cameras back in for service again.
 
here's hoping the firmware update will include focus peaking and improved 720P sharpness(I know I'm dreaming)
 
here's hoping the firmware update will include focus peaking and improved 720P sharpness(I know I'm dreaming)

=)

There are no new features apart from 25p this time around. But Canon has reached out and asked for feedback on what we want implemented. Some time at a later point, I'd like:


​

  • [*=1]customizable framerates in movie mode (1-25 or what ever) to allow for timelapse without using the shutter mechanism.
    [*=1]Log setup position like on the C100
    [*=1]Wide DR picture profile
    [*=1]punch in zoom during filming, movable area with joystick (but it would be OK if I couldn't move the area, if the position I set before filming was maintained).


​
I personally don't care for peaking, waveform or zebras, but false colors would probably be of some use if well implemented (tuned to Canon Log for instance).
 
Actually there IS one thing I hope Canon fixes.
Apparently when using 'View Assist' it also affects the histogram. I just found this out on watching Phillip Bloom's review a 2nd time. So the histogram no longer accurately depicts the recorded signal. WTF!
Can anyone tell me why you might want the histogram to work this way?
 
Last edited:
Actually there IS one thing I hope Canon fixes.
Apparently when using 'View Assist' it also affects the histogram. I just found this out on watching Phillip Bloom's review a 2nd time. So the histogram no longer accurately depicts the recorded signal. WTF!
Can anyone tell me why you might want the histogram to work this way?

The histogram should be that way, otherwise how can you accurately judge exposure? when VA is on, you expose the midtone at 50%, it will also be shown in the histogram as 50%, and in the recorded Canon Log file, you get midtone at 33%.
 
Its not the 50% tones I'm concerned about. What I'm talking about is easier illustrated in Photoshop. Open a sample Canon Log image, preferably with average contrast, details in shadow and highlight areas. Open the histogram as well. This should show ample headroom in both shadow and highlite areas in the histogram. Now create a curves adjustment layer to simulate view assist mode. Create an s curve to approximate what applying a Standard profile might look like.
Now while viewing the histogram, toggle the adjustment layer off and on. View assist mode is showing that you have less headroom than you actually might have. If you were trying to light a scene to take full advantage of Canon log dynamic range by viewing the histogram in view assist mode, you couldn't!
 
Its not the 50% tones I'm concerned about. What I'm talking about is easier illustrated in Photoshop. Open a sample Canon Log image, preferably with average contrast, details in shadow and highlight areas. Open the histogram as well. This should show ample headroom in both shadow and highlite areas in the histogram. Now create a curves adjustment layer to simulate view assist mode. Create an s curve to approximate what applying a Standard profile might look like.
Now while viewing the histogram, toggle the adjustment layer off and on. View assist mode is showing that you have less headroom than you actually might have. If you were trying to light a scene to take full advantage of Canon log dynamic range by viewing the histogram in view assist mode, you couldn't!

Hehe, I guess you don't have much experience shooting log in real film set. simply speaking VA lets you expose in a traditional way (50% midtone), yes you will probably get "clipped" high/low in the histogram but you know they are still there, in the recorded image (if you exposed correctly).

Think of VA as nightvision goggle for soldiers, you can see things and move freely with it at night, but actually the darkness is still there which makes you safe.
 
View assist mode is showing that you have less headroom than you actually might have. If you were trying to light a scene to take full advantage of Canon log dynamic range by viewing the histogram in view assist mode, you couldn't!

I understand you want to use View Assist to make it easier to see what's going on a bit better, and maybe get a crisper image that is easier to focus?

But what I don't quite get is this: if the histogram would still correspond to Log while using View Assist, what would you do then? Expose to the right? Is that the headroom you feel is lost with the View Assist histogram?

If it is, then that's the wrong track. A normal exposure in Log should probably show a histogram that dies right at the last line before the right edge. That is: the last 20% of the histogram would end up unused UNLESS you are shooting into a light of some sort. Into the sky/sun, into lamps and so on... No reflected light should show up there. If you expose reflected light "to the right" you are going to compress your signal a lot where it matters. Lose tonality.

But it's quite possible that I misunderstand you...
 
Ok then it is my misunderstanding of how Canon Log works in regards to the histogram. 'Expose to the right' has always been my mantra. Will do a few tests in C log exposing to the far right to see the results. Curious if the below 400 ISO limited dynamic range is something that can be fixed in firmware. Thanks.
 
It's technically possible to expose to the right of course, it's just that you'll put parts of the image in the "highlights roll off zone" that belong lower in the histogram.

And the reason behind all of this is Canon's efforts to create a 12 stop picture profile with similar amounts of visual detail and tonality over the whole shadow-mids-highlights spectrum, in an 8-bit space: they've got quite a curve on the gamma, and it starts to flatten out pretty early when you approach the top section.
 
Back
Top