Ad
Ad
Ad
Pages: « 1 [2] 3 4 ... 6 »   Bottom of Page
Print
Author Topic: Camera White Balance and Post Processing  (Read 16149 times)
b2martin
Full Member
***
Offline Offline

Posts: 130


« Reply #20 on: July 22, 2011, 07:02:17 AM »
ReplyReply

When I shoot a panorama, I use preset white balance - set this by pointing the camera at the light source with a BaLens cap and set the white balance.  I set the camera in manual mode,  use autofocus to focus the lens and then turn it to manual.  If I shoot multi row panorama, I generally turn on auto focus for the bottom row since objects are much closer than the top two rows.  I shoot panoramas with the camera in portrait mode.  Lens is generally 105mm and most images require 3 rows of 6 images each.  Image is saved in RAW format and I process using Photoshop CS5 ACR RAW comverter.  Since exposure and white balance are identical in each image, I select one of the images in ACR and then select all and adjust all at the same time. 
Logged
Bill Koenig
Sr. Member
****
Offline Offline

Posts: 339


« Reply #21 on: July 27, 2011, 03:19:53 PM »
ReplyReply

I checked the total time needed for my last few panorama shots (in auto WB) and found an elapsed time of just 5-10 seconds over 3-5 exposures. I've never had a problem with light changing that quickly.

That would work for just 3-5 exposures.
For most of the pano's I shoot, using a spherical pano head I'm doing vertical columns of as many as 5-6 frames, with 10 to 12 horizontal frames with a 85mm lens. Add in HDR and now you have three exposures per frame, setting a custom WB is essential.
Working fast, it could take 30 to 40 min, that said, your not going shoot pano's this big as the sun is setting.
Depending on how late in the day, and how fast the light is changing, by shooting vertical columns, adjusting exposure as you go from column to column has worked for me, but with AWB would it will fail


 
Logged

Bill Koenig,
milt
Jr. Member
**
Offline Offline

Posts: 69



WWW
« Reply #22 on: July 29, 2011, 07:53:04 PM »
ReplyReply

When I shoot a panorama, I use preset white balance - set this by pointing the camera at the light source with a BaLens cap and set the white balance.  I set the camera in manual mode,  use autofocus to focus the lens and then turn it to manual.  If I shoot multi row panorama, I generally turn on auto focus for the bottom row since objects are much closer than the top two rows.  I shoot panoramas with the camera in portrait mode.  Lens is generally 105mm and most images require 3 rows of 6 images each.  Image is saved in RAW format and I process using Photoshop CS5 ACR RAW comverter.  Since exposure and white balance are identical in each image, I select one of the images in ACR and then select all and adjust all at the same time. 

I'm missing something here.  Since the white balance setting (preset or otherwise) doesn't affect the RAW data and ACR doesn't pickup any white balance setting from the camera, what's the point of setting this preset?

--Milt--
Logged

Los Gatos, California | http://miltonbarber.com
Schewe
Sr. Member
****
Offline Offline

Posts: 5255


WWW
« Reply #23 on: July 29, 2011, 08:06:06 PM »
ReplyReply

I'm missing something here.  Since the white balance setting (preset or otherwise) doesn't affect the RAW data and ACR doesn't pickup any white balance setting from the camera, what's the point of setting this preset?

Actually, ACR/LR DOES pick up WB settings in the raw file. What it DOES with them is, however usually different than what the camera makers software would do since ACR/LR's white balance is unique...setting a fixed WB in camera would prolly be better for panos than letting the camera adjust the WB using auto. Note, you are correct that the WB setting in camera is simply a metadata tag and can be overridden in ACR/LR. That takes only a bit of work to WB one of the pano images and sync to the rest of the captures. So it's more efficient to not worry about WB when shooting and adjust it in the raw processing before the pano is done...
Logged
milt
Jr. Member
**
Offline Offline

Posts: 69



WWW
« Reply #24 on: July 30, 2011, 11:18:18 AM »
ReplyReply

Hmm.  I think I'm learning something new here.   Are you saying that ACR does pick up the information from a camera white balance preset and make it available for use during processing?  If so, and if that camera preset were established as suggested by a baLens (or a card), then that's a piece of information about the shooting conditions that isn't otherwise available during processing.  Also, is ACR able to do this for a pretty wide variety of modern cameras?

--Milt--
Logged

Los Gatos, California | http://miltonbarber.com
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3015


« Reply #25 on: July 30, 2011, 03:18:28 PM »
ReplyReply

Hmm.  I think I'm learning something new here.   Are you saying that ACR does pick up the information from a camera white balance preset and make it available for use during processing?

It reads the  "As shot" white balance settings from the EXIF data and uses that as a startingpoint.

Cheers,
Bart
Logged
Oldfox
Jr. Member
**
Offline Offline

Posts: 76


« Reply #26 on: October 24, 2011, 12:28:20 PM »
ReplyReply

I found this old topic after a discussion with a friend of mine. He insisted that my way of using Raw and Auto in camera settings is not the proper way. He uses a Color Temperature Meter and adjusts white balance in LR if necessary. I have been quite happy with using Auto so far. Sometimes I have changed the white balance a bit, but most of the time I have accepted the white balance suggested by Adobe ACR.

Today I made a simple test taking two raw images, one with Auto and one with Daylight. I processed the raw images in Adobe ACR and to my astonishment Daylight gave better result than Auto. It was a sunny day and I made the test in the middle of the day. The difference was not big, but the difference was easy to see.

Maybe I have to make new tests...

(Canon 5dmkII, Canon EF 24-105 f/4, Bridge CS4 3.0.0.464, ACR 5.7.0.213)
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 8073



WWW
« Reply #27 on: October 24, 2011, 01:48:51 PM »
ReplyReply

Maybe I have to make new tests..

I think so as the WB has no effect on the raw data. The suggestions made do in terms of what you initially see but you can over-ride that at any time. IOW, the metadata affects the converter, not the raw data.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Oldfox
Jr. Member
**
Offline Offline

Posts: 76


« Reply #28 on: October 24, 2011, 03:09:07 PM »
ReplyReply

Of course it hasn't. It's all about guessing or determining the 'starting point'.

As Guillermo Luijk writes in one of articles: "La conclusión es que no puede existir un balance de blancos automático que emule bien en cualquier situación el balance que hace nuestro cerebro"

"The conclusion is that there can be no automatic white balance that emulates good balance in any situation that makes our brain." (Translation by google, not mine...)
Logged
louoates
Sr. Member
****
Offline Offline

Posts: 678



WWW
« Reply #29 on: October 24, 2011, 04:50:44 PM »
ReplyReply

White balance has become one of those subjective evaluations that has been taken to extremes, especially in the stock photography business. I have had dozens of otherwise acceptable images refused because "...white balance may be incorrect". What that often means is that the reviewer doesn't like something about the image that he/she can't quite put a finger on. On the rejected images I often go back and double check my work flow and, sure enough, the white balance is perfect. My conclusion is that many reviewers live in climates other than mine (Arizona) and are not used to seeing really brillant sun and the colors that result, especially in landscape scenes. The same rejections often occur in low-light situations when the reviewer hasn't experienced that particular light condition and thus it must be "incorrect white balance."
Thankfully I've since ceased to worry about outside evaluations on this subject and simply process my images as I see fit.
Logged
PierreVandevenne
Sr. Member
****
Offline Offline

Posts: 508


WWW
« Reply #30 on: October 24, 2011, 09:53:49 PM »
ReplyReply

AWB is almost always "right". Anyway, as others have said, you can post-process raw data as you wish but the archetypal example is the candle light dinner. Technically correct white balance with yield hospital like shots, but we all know that the yellowish AWB default settings come up with is the good one :-)
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5255


WWW
« Reply #31 on: October 24, 2011, 10:37:43 PM »
ReplyReply

AWB is almost always "right".

No...it's often close but usually requires tweaking and this is really camera dependent. nFor example, AWB on a Phase One back sucks big ones...similar to Panasonic and other lesser brands (it's ok, I've got a GH2). But Nikon and Canon AWB is close but expect to optimize it series by series.
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1055



WWW
« Reply #32 on: October 25, 2011, 02:18:35 AM »
ReplyReply

Then you get into the issue of what influences the proper appearance of WB shooting in 2800K tungsten or sunset. It can mess up your perception during WB tweaks when you're not quite sure if the color cast of the surface in question is being influenced by the color cast of the light source OR by the spectral reflectance characteristic of the surface itself as I tried to point out in this thread...

http://photo.net/beginner-photography-questions-forum/00ZUlB

The OP specifically states the japanese woman was wearing white-ish makeup lit by the setting sun shot outdoors but she kept indicating no matter what she did with WB in her Raw processor the image just wasn't looking right. You'll note my demo edits didn't result in the make up looking white at all. I suspected the white pancake makeup lit by the setting sun reflected back something the camera recorded but the eyes did not see.

Would you say this is one situation WB shooting Raw did have some influence on the Raw data or maybe the processor? I was thinking she should've chosen "tungsten" WB preset within her camera to correct for the white makeup.
Logged
Oldfox
Jr. Member
**
Offline Offline

Posts: 76


« Reply #33 on: October 25, 2011, 03:02:31 AM »
ReplyReply

No...it's often close but usually requires tweaking
This is the lesson I learned. I have routinely accepted ACR's suggestion. From now on I will try focus more on white balance in ACR:
Logged
Guillermo Luijk
Sr. Member
****
Offline Offline

Posts: 1273



WWW
« Reply #34 on: October 30, 2011, 05:05:49 AM »
ReplyReply

The right WB is many times subject to discussion. What is the definition of a right WB?
- The one that allows to see the real colour of objects?
- The one that reminds us better the feeling we had at taking the picture?
- The one that pleases most our eyes?
...

There could be 3 different WB settings for those 3 definitions. IMO the right WB depends on the application. Just an example:



If you are so lucky to have Ferrari as a customer, they will demand a precise Ferrari red colour for their catalogue. The right WB here will be the one that provides the real colour of the Ferrarri. So using a custom WB over a neutral gray card will be the right WB.

If we are taking more emotional pictures of a warm sunset over a white building, we don't want the building to look white (its real colour) in the final image, but preserve some of the warm atmosphere. In this case a gray card would ruin our WB, and we would obtain the right WB by just tweaking it on the RAW developer to match the light and feelings we recall from the scene.

Two different ways to achieve the right WB. None of them would have worked in both situations.

Regards
Logged

stamper
Sr. Member
****
Offline Offline

Posts: 2391


« Reply #35 on: October 30, 2011, 05:13:58 AM »
ReplyReply

I have noticed that when I change the WB setting in ACR to different settings then the exposure changes thus affecting the histogram. If it changes in ACR then the setting chosen in camera will affect the histogram because it is a jpeg histogram. So choosing a WB in camera does make a difference?
Logged

hjulenissen
Sr. Member
****
Offline Offline

Posts: 1615


« Reply #36 on: October 30, 2011, 06:38:44 AM »
ReplyReply

The right WB is many times subject to discussion. What is the definition of a right WB?
Exactly.

Imagine staying inside a candle-lit cabin with your friends late in the evening at winter-time. Look out the window and you may see a snow-clad landscape with an effective color-temperature that is very different from inside the cabin (i.e. outside the "whites" are "blueish", while inside the whites are yellowish). Do you or any of your friends complain about the white-balance of the window or of nature being "wrong"? Of course not, they (subconsciously) seem to accept that the illumination outside is different, and as a consequence the color balance is shifted.

What happens if you take a picture of that scene, print it and hang it on the wall of the same cabin? Say that you have a camera workflow that prioritize "engineering philosophy", i.e. recreating the color perfectly within the predefined CIE space. How would that image look under similar conditions? Would you and your friends think that it had a blue cast?

I have tried to use this analogy to get a better grip at what WB really is and what people expect it to do. To me it seems that most people want WB to "move them into the scene". I.e. alter the WB of the image in such a way that the viewer perceives it the same way as if he had been there (assuming perceptual adaptive processes) if it is to fill the entire field of vision, and/or that it is a natural extension to however the viewing environment (lighting, roof, walls,...) is. I.e. i the viewing environment is yellow-cast, the viewers adapt to that white-point, and expect the image to have a similar white-point (under the same illumination - directly lit displays is perhaps more difficult)

-h
Logged
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1615


« Reply #37 on: October 30, 2011, 06:42:19 AM »
ReplyReply

AWB is almost always "right". Anyway, as others have said, you can post-process raw data as you wish but the archetypal example is the candle light dinner. Technically correct white balance with yield hospital like shots, but we all know that the yellowish AWB default settings come up with is the good one :-)
I am seldomly happy with the AWB of my Canon SLR. It does not bother me that much because I can rapidly find something better using either the WB "picker", or just tweaking the settings for a group of pictures taken under the same lighting conditions.

The most difficult is when a face/skin dominate the image, no "neutral whites" are present, and the background is dark/grey. For those cases, I wish there was a "skin tone picker" that let me mark the position of human skin, and then apply WB that made it match some averaged human skin chromaticity - as a starting point.

-h
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1055



WWW
« Reply #38 on: October 30, 2011, 04:40:36 PM »
ReplyReply

The selected camera profile can greatly affect the appearance of perceived WB in combination with adjusting the temp/tint WB sliders in ACR/LR.

With my camera Adobe Standard profile neutralizes the entire tonal scale with tungsten lit scenes that may look too yellow amber at default WB. With the right warm hue of overall WB this profile can create some unusual de-saturated browns and beige skin tone colors that at first look a saturated orange yellow with the default ACR 4.4 or a custom DNG profile.

Highlights will appear a dull (de-saturated) yellow amber while the rest of the image seems neutral. It completely changes the emotional ambience of the image in unexpected ways depending on the brand and type of tungsten light shot under and adjustments to WB in ACR/LR.

Logged
kers
Sr. Member
****
Online Online

Posts: 707


WWW
« Reply #39 on: October 30, 2011, 06:59:43 PM »
ReplyReply

I am seldom happy with the auto white balance of my Nikon d3x either...and even the exposure.. There needs still some improvement to be made
Logged

Pieter Kers
www.beeld.nu
Pages: « 1 [2] 3 4 ... 6 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad