Paul C. Buff, Inc. Technical Forum

Technical Discussion Forum for all Paul C. Buff, Inc. Products

Login

Post a reply
 [ 7 posts ] 

Tue Dec 22, 2009 10:45 pm

Joined: Wed Dec 16, 2009 1:17 pm
Posts: 4
Location: Overland Park, Kansas

Sure appreciate the help from this forum. I'm new to studio lighting and have been testing one of my brand new B800s with my Sekonic meter and my Nikon D300. I set up the B800 with a 60" sliver umbrella and shot a grey card at various light settings. I'm using ISO 200 and a sync speed of 1/250th. (Max on the D300 is 1/320th but the meter doesn't have that setting.) I've been careful to ensure that both the meter and the camera have matching settings. In addition, I put the camera on a tripod to reduce variables as much as possible and used the sync cord for all the metering as well as the shots.

I've adjusted the light to be exactly f/2.8, f/4, f/8, f11 and f/16 and shot the grey card at each of those settings. On the first three, the histogram was dead on (or close enough). But on the others, it was off. The histogram for f/8 centered at f/10, f/11 centered at f/13, and f/16 centered at f/20. Is this variation of the histogram with the metered settings normal?

Thanx,
Bruce




Top Top
Profile
 

#

Tue Dec 22, 2009 11:45 pm

Site Admin
Joined: Tue Dec 08, 2009 11:49 am
Posts: 1432

It's not that uncommon. Histograms can depend on white balance setting, flashmeters aren't always absolute. camera exposures aren't always absolute, ad infinitum. If you are shooting in RAW but your camera color temp isn't close to the actual light color temp the histograms can be in error.

I find color 200+°K temperature difference on my D300 on a test patch shot at f11 than on the same patch and light settings, but shot at f16. Canon is worse in this regard than Nikon. On my McBeth color cards I find 400°K differences from one grey patch shade to another in the same exposure. On my Color Meters, Minolta and Gossen, I find 300°K indicated color temperature variations depending on the distance from light to meter.

All these things have s RGB sensors, and none of them are completely linear. Digital displays can be a curse because their accuracy depends on many analog components and many nonlinearities. My digital speedometer says 33MPH when my digital GPS says 30MPH . . . which one do I argue with the cop?




Top Top
Profile
 

#

Wed Dec 23, 2009 8:49 am

Joined: Wed Dec 16, 2009 1:17 pm
Posts: 4
Location: Overland Park, Kansas

Thanks, Paul -- er, Luap, I mean. I figured that was probably the case. Sounds like I'll just need to be aware of these differences and compensate accordingly. However, you mentioned white balance, and I wonder if -- even for just a test -- setting a custom white balance would have changed the outcomes. I used the camera's flash WB setting just to be consistent with the test shots. Any thoughts?

Thanx,
Bruce




Top Top
Profile
 

#

Wed Dec 23, 2009 9:09 am

Joined: Sun Dec 20, 2009 7:48 am
Posts: 7

SkyGuy,
In order to generate an histogram the raw data from the sensor needs to be processed into some sort of color space, so that a luminosity value can be computed (luminosity here being used in a general sense as opposed to a strict color science sense - the details of what most cameras do are usually proprietary).
On Canon cameras (the brand I own), from my experimentation, the setting of the color space, color temperature and picture stile all affect the displayed histogram. It stands to reason that variation of color temperature in the light would similarly affect the results of the algorithm that computes the histogram, so that different histogram would result for different color temperatures even if the light flux and the reflectivity of the scene are unchanged.

In the end I tend to shoot with a fixed color balance (daylight) and with the default picture style as these setting do not really affect the result when shooting raw, and by calibrating the camera behavior - as you did - you get at least a somewhat consistent way to interpret the histogram.

For studio work, my sekonic L758 is usually dead on, and in those conditions I rely on the light meter much more than on the camera histogram.

Merry Christmas and happy shooting,

Stefano




Top Top
Profile
 
Website
 

#

Wed Dec 23, 2009 11:15 am

Joined: Wed Dec 16, 2009 1:17 pm
Posts: 4
Location: Overland Park, Kansas

Thanks, Stefano. I appreciate your input. I don't know why I hadn't thought about using the daylight WB setting since that's how the ABs are calibrated, but it makes sense. I just used the flash setting arbitrarily -- as mentioned, my main concern was reducing the number of variables as much as possible. So thanks for the explanation. Just for the fun of it, I think I'll repeat the tests using the daylight and a custom WB setting just to see what happens. If nothing else, it will help me learn. And merry Christmas to you as well!

Bruce




Top Top
Profile
 

#

Thu Dec 24, 2009 3:59 am

Site Admin
Joined: Tue Dec 08, 2009 11:49 am
Posts: 1432

skyguy wrote:
Thanks, Paul -- er, Luap, I mean. I figured that was probably the case. Sounds like I'll just need to be aware of these differences and compensate accordingly. However, you mentioned white balance, and I wonder if -- even for just a test -- setting a custom white balance would have changed the outcomes. I used the camera's flash WB setting just to be consistent with the test shots. Any thoughts?

Thanx,
Bruce


Setting a custom white balance will likely have some positive effect. As another poster said, the color space used by the histogram can effect the accuracy if it is very far different from the actual color.

I don't do a whole lot of shooting, but when I do, I always lean toward a little under exposure so I don't burn out highlights. With modern CMOS cameras it's easy to bring the exposure up to what you want without detrimental effects, but it's impossible to repair burned out areas.

Also, no meter or histogram can equal your own eyes in terms of what constitutes a proper exposure. For instance, if you meter a scene with reflective elements - jewelry, sequins, silverware, etc, the meter can't read the specular reflected highlights and they are certain to burn out.




Top Top
Profile
 

#

Thu Dec 24, 2009 4:00 pm

Joined: Wed Dec 16, 2009 1:17 pm
Posts: 4
Location: Overland Park, Kansas

That makes sense, Luap. I'm going to retest using the same criteria as before but this time I'll take a custom white balance as well. Thanks to everyone for your help. Maybe someday I'll actually know what I'm talking about!

Merry Christmas or happy holidays -- your choice!

Bruce




Top Top
Profile
 
Display posts from previous:  Sort by  
Post a reply
 [ 7 posts ] 

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 68 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum


cron