Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Alarming test: In-camera settings baked-in in post??!!

J

James Press

Well-known member
Joined
Oct 8, 2007
Messages
495
Reaction score
0
Points
0
Location
Sydney
Just doing some testing and was very alarmed by the results: unless I'm missing something, it appears that any metadata settings that get dialed into the camera (eg saturation; contrast; exposure; curve) DOES carry with the footage in post into RedAlert, and effects the footage in RedAlert EVEN WHEN YOU RESET THE METADATA TO DEFAULT (or any other setting).

We conducted 2 footage tests, shooting the same color chart in our studio, about a minute apart, with no variables except for adjusting some of the metadata settings in the camera to Footage number 2. In footage number 1, we set the camera to the RedAlert "default" settings; in the second footage we made some minor adjustments to some of the RedAlert default settings (Contrast= -0.2; saturation=0.9, exposure=0.1, brightness=0.1, 5400K, and slight S-curve).

We then opened the R3Ds in RedAlert; and adjusted the settings of footage number 2 (shot with the variation, "non-default" footage) back to default; made QT proxies; and imported the _H QT into FCP. So both footage now has exactly the same RedAlert settings (set to default), shot in exactly the same conditions, so it should look exactly the same, right?

Well, it doesn't, and this is what I can't understand. I've attached pictures of the two different shots accompanied by the vectorscope, and you can see immediately the difference in at least the saturation:

Can anyone explain this?? I understood that whatever you set the camera settings to was simply metadata, and was completely changeable in post, not effecting the footage. But this test suggests to me that the in-camera settings do effect what you can work with in post?? Hopefully I'm missing something...
 
they look pretty similar to me, even on the scope. why don't the shots line up perfectly, did you move the camera at all?
 
Hi ronx, if you check on the scopes, the position of the color patches is different--in the second shot they are more saturated.

The camera or tripod didn't move position, I just didn't frame them identically, but that shouldn't effect the color values of the chart, I wouldn't have thought.
 
James Press said:
Just doing some testing and was very alarmed by the results: unless I'm missing something, it appears that any metadata settings that get dialed into the camera (eg saturation; contrast; exposure; curve) DOES carry with the footage in post into RedAlert, and effects the footage in RedAlert EVEN WHEN YOU RESET THE METADATA TO DEFAULT (or any other setting).

We conducted 2 footage tests, shooting the same color chart in our studio, about a minute apart, with no variables except for adjusting some of the metadata settings in the camera to Footage number 2. In footage number 1, we set the camera to the RedAlert "default" settings; in the second footage we made some minor adjustments to some of the RedAlert default settings (Contrast= -0.2; saturation=0.9, exposure=0.1, brightness=0.1, 5400K, and slight S-curve).

We then opened the R3Ds in RedAlert; and adjusted the settings of footage number 2 (shot with the variation, "non-default" footage) back to default; made QT proxies; and imported the _H QT into FCP. So both footage now has exactly the same RedAlert settings (set to default), shot in exactly the same conditions, so it should look exactly the same, right?

Well, it doesn't, and this is what I can't understand. I've attached pictures of the two different shots accompanied by the vectorscope, and you can see immediately the difference in at least the saturation:

Can anyone explain this?? I understood that whatever you set the camera settings to was simply metadata, and was completely changeable in post, not effecting the footage. But this test suggests to me that the in-camera settings do effect what you can work with in post?? Hopefully I'm missing something...
Click to expand...

You are missing something... I'm busy right now so I'll let someone else help you out.

Jim
 
Thanks Jim-- that's incredibly generous, given all you all have on. I apologise in advance if this is a time waster...
 
First off, the proxies are proxies, and I wouldn't be too concerned about them.

If your DPX's aren't coming out right, then you might want to do some further digging.

Have you tried rendering out some test footage to UC or DPX/TIFF from both RedAlert and RedCine?
 
Hi Charles,

I'm using the proxies as that's our preferred (fast) workflow for web and DVD outputs. Others, like Mark Toia seem to get great results (for HD deliverables) using the proxies: http://www.reduser.net/forum/showthread.php?t=25431.

I'm just trying to get a "stable baseline" out of RedAlert so that I can get an image that looks close to "normal", that I will only have to do minor (ie primary) color correction to in FCP (ideally--if not, then in Color). We're shooting people talking around a table (see pics below) and not really wanting, if at all possible, to do much cc...
 
James, I don't know what to tell ya'. When I attempt what you describe, it works fine. I'm going to have to assume you're missing something. At the very least, have you rebooted your machine and tried again?
 
Yeah, it seems bizaar, so I'm sure it's something I'm doing (or not doing), but what??

More relevantly, could someone post their "standard" RedAlert settings--that would be a big help.
 
The adjustments you make to saturation, curves etc in camera are what gets mapped to the "default" position when you reset it in RedAlert and RedCine. But they ARE still metadata and you can verify that by setting the parameters the same across the two shots (not using the "default" button)

To see your RAW image at its neutral position set 5000K, zero out all the curves and saturation/contrast/colour parameters etc to be the same. You'll find that unless you altered something on the lens or adjusted shutter speed, your shots WILL be identical.

HTH

Paul
 
For all I know, any changes you make to Colour Space, Gamma and Curves in REDAlert will NEVER be applied to the Quicktime reference files (proxies). You are stuck with the settings you made in-camera.

The above named settings WILL be applied if you transcode the clips, i.e. render out to ProRes, DPX, etc.

Cheers,
Hans
 
I know this seems so simple but in the past I had to actually delete my camera generated proxies before I produced new ones from Red Alert. When I didn't delete first my newly exported proxies didn't take. This might have been fixed in a new version, but I've gotten in the habit of doing this first.

Are the adjusted proxies different from the camera generated ones?
 
Thanks for the considered thoughts--I'll get testing again now and post the results.

David-
We didn't generate proxies in-camera, only in post in RedAlert.
 
Hans-Georg Daun said:
For all I know, any changes you make to Colour Space, Gamma and Curves in REDAlert will NEVER be applied to the Quicktime reference files (proxies). You are stuck with the settings you made in-camera.
Click to expand...

That is correct for now, these features will be added in a future version of the QuickTime codec.

David Anderson said:
Are the adjusted proxies different from the camera generated ones?
Click to expand...

No, they should be exactly the same

Look settings are absolutely not baked in! It is only metadata. Only things that affect your recording are things like framerate, shutter speed, resolution, how you set your lens, etc.
 
You must log in or register to reply here.
Back
Top