Can Canon Log DR be matched with a custom profile?
- Thread starter jcs
- Start date
Tweaks: green & WB matched (CLog changes color & WB*), minor levels, etc. to match a bit closer before applying the main grade, which is an adjustment layer and applied the same to both clips.
* When switching from a custom profile on the EOS R to CLog, where WB is fixed, Canon changes the WB (cooler); all on the same camera. Thus, even though both cameras are set to the same fixed WB, must tweak CLog in post to get a closer match.
NorBro requested an exposure sweep with both cameras at the same time.
Settings for both cameras:
Canon 24-105 F4L at 50mm
F5.6
1/50
ISO set to match exposure on histograms
Full frame 1080p ALL-I
Same kind minor color/level tweaks in post before applying an adjustment layer grade to both clips at the same time (very easy to match even though I haven't tweaked the custom log profile for the 1DX II yet (was developed on the EOS R)).
Aputure Lightstorm LS1s from 10-100%:
Results are as expected- CLog's still a bit better, however I'm surprised it was this close. Could be useful to those needing to match the 1DX II, 80D etc. to other Canon cameras using CLog + maximizing usable DR on non-CLog Canon cameras.
Settings for both cameras:
Canon 24-105 F4L at 50mm
F5.6
1/50
ISO set to match exposure on histograms
Full frame 1080p ALL-I
Same kind minor color/level tweaks in post before applying an adjustment layer grade to both clips at the same time (very easy to match even though I haven't tweaked the custom log profile for the 1DX II yet (was developed on the EOS R)).
Aputure Lightstorm LS1s from 10-100%:
Results are as expected- CLog's still a bit better, however I'm surprised it was this close. Could be useful to those needing to match the 1DX II, 80D etc. to other Canon cameras using CLog + maximizing usable DR on non-CLog Canon cameras.
Last edited:
Thanks- the settings are fixed- I changed the light, histograms at low and high:
CLog requires 2x the ISO to match non-CLog profiles (all of the them, not just custom profiles). Lower ISO provides more DR- outdoors the lowest is 400 for CLog and 200 for non-CLog.
Guess it's cool you couldn't tell which is which without the settings
(would also be easier to tell if I did no post- however the purpose of the test was to show real, usable DR, and we'll always do post). In any case, I've offered to let anyone test and see results for themselves with their own tests.Neither video has been correctly guessed yet in this thread.
NorBro
Major Contributor
Not sure why C-Log's ISO is being bumped twice as much (and you have some D-Lighting activated?).
You're really not budging - thread after thread - in doing this test the correct way, lol.
Just so simple...leave the ISO, aperture and shutter the same and cycle through (with zebras). It's so easy, jcs.
You're really not budging - thread after thread - in doing this test the correct way, lol.
Just so simple...leave the ISO, aperture and shutter the same and cycle through (with zebras). It's so easy, jcs.
Not about budging- it's about time- already spent way to much time on this. Recording the back of the cameras doesn't come close to representing what's actually recorded. Furthermore, it makes the guessing game not very fun (as in impossible, since everyone knows which camera is which haha).
Good to see you still can't tell even with shots from the back of the cameras! Thanks for prompting me to do these tests, I was curious how it would turn out. I still give the edge to CLog, and am still quite surprised it was possible to get this close based on earlier tests and other folks trying and not quite getting there. Additionally, skin tones seem to survive OK after grading.
The most common complaint of camera tests online, "You did it wrong!!" lol.
Anyone wishing to "do it right" can PM me and I'll send so you can test.
This is impossible- if ISO, aperture and shutter are the same, what's left to cycle through? Neither camera provides zebras. The LCDs aren't calibrated and don't represent what's actually recorded.
Is your position that recording the back of cameras is the only valid way to compare DR?
Holding everything constant and increasing light shows what the cameras can do with respect to DR and clipping. The fact that I made minor tweaks to hide differences between the cameras to make guessing more fun doesn't invalidate real-world DR capability + the tech is available for others to test for themselves.
NorBro
Major Contributor
You do not understand that it is not about this silly matching at different settings.
One of the main purposes behind C-Log (or any Log) is so you can capture a certain amount of dynamic range at certain settings.
What is the reasoning behind setting C-Log to ISO 2500 and your profile to ISO 1200 and saying: "Look, they match!"
I want to see both cameras at ISO 400 with the same aperture and the same shutter. If you show the DR in one profile and then you adjust the aperture, ISO or shutter on the other one and then also add post-production "tweaks", how the heck is this a valid evaluation?
How can we possibly know anything? It's not about having fun.
This discussion goes back a month now and you refuse to provide the test I'm asking for and I'm kindly done with it.
I will not be commenting on it anymore and I hope you sell a million copies of your profile.
One of the main purposes behind C-Log (or any Log) is so you can capture a certain amount of dynamic range at certain settings.
What is the reasoning behind setting C-Log to ISO 2500 and your profile to ISO 1200 and saying: "Look, they match!"
I want to see both cameras at ISO 400 with the same aperture and the same shutter. If you show the DR in one profile and then you adjust the aperture, ISO or shutter on the other one and then also add post-production "tweaks", how the heck is this a valid evaluation?
How can we possibly know anything? It's not about having fun.
This discussion goes back a month now and you refuse to provide the test I'm asking for and I'm kindly done with it.
I will not be commenting on it anymore and I hope you sell a million copies of your profile.
G
Guest
Guest
I just want to weigh in. The way that Canon implements C-Log, is they shift the exposure. So, in order to fairly compare dynamic range of C-Log with a color profile, they should have different iso values.
For example, C-Log might be clipped in the shadows, while a colour profile still has detail there, and then as one begins to increase light, the C-Log will hold onto the highlight longer, and the colour profile will clip. But they both might share similar dynamic range.
This is not a test for middle grey. It is a total dynamic range comparison. C-Log always won in my tests, but not by enough to make me never use colour profiles.
Jcs is being straight forward enough here, and his offsetting the iso makes sense in terms of seeing the dynamic range differences.
For example, C-Log might be clipped in the shadows, while a colour profile still has detail there, and then as one begins to increase light, the C-Log will hold onto the highlight longer, and the colour profile will clip. But they both might share similar dynamic range.
This is not a test for middle grey. It is a total dynamic range comparison. C-Log always won in my tests, but not by enough to make me never use colour profiles.
Jcs is being straight forward enough here, and his offsetting the iso makes sense in terms of seeing the dynamic range differences.