Uniwb walkthrough. Went ahead and did it after hearing a few

File: MAGENTA_WHITE_BALANCE.jpg (2KB, 256x256px) Image search: [Google]

2KB, 256x256px

>>3005259
Eventually you end up with a magenta color. This magenta hue is what the camera 'sees' as grey (2x green photosites)
RGB values are 172,64, and 91, respectively

File: 00000019-20170102-210000-RawDigger-ScreenShot.png (819KB, 1286x1070px) Image search: [Google]

819KB, 1286x1070px

Now results. Here is the raw over exposure displayed in rawdigger. This is with slight OE [image 19, 1/10]

File: image1(2).jpg (854KB, 800x567px) Image search: [Google]

854KB, 800x567px

>>3005264
And here is the blinkies using the jpeg on back of camera [image 19].

[EXIF data available. Click here to show/hide.]

Camera-Specific Properties:
Equipment Make	Apple
Camera Model	iPhone 6s Plus
Camera Software	Snapseed 2.14.141428617
Sensing Method	One-Chip Color Area
Focal Length (35mm Equiv)	29 mm
Image-Specific Properties:
Pixel Composition	Unknown
Horizontal Resolution	72 dpi
Vertical Resolution	72 dpi
Image Created	2017:01:18 22:59:39
Exposure Time	1/4525 sec
F-Number	f/2.2
Exposure Program	Manual
ISO Speed Rating	400
Lens Aperture	f/2.2
Brightness	4.9 EV
Exposure Bias	0.3 EV
Metering Mode	Pattern
Flash	No Flash, Compulsory
Focal Length	4.15 mm
Color Space Information	sRGB
Image Width	800
Image Height	567
Exposure Mode	Manual
White Balance	Auto
Scene Capture Type	Standard

File: 00000018-20170102-210000-RawDigger-ScreenShot2.png (1006KB, 1286x1070px) Image search: [Google]

1006KB, 1286x1070px

Raw digger OE for image 18 (1/8)

File: image2(1).jpg (141KB, 538x390px) Image search: [Google]

141KB, 538x390px

Jpeg display on back of camera for image 19 (1/8)

[EXIF data available. Click here to show/hide.]

Camera-Specific Properties:
Equipment Make	Apple
Camera Model	iPhone 6s Plus
Camera Software	Snapseed 2.14.141428617
Sensing Method	One-Chip Color Area
Focal Length (35mm Equiv)	29 mm
Image-Specific Properties:
Pixel Composition	Unknown
Horizontal Resolution	72 dpi
Vertical Resolution	72 dpi
Image Created	2017:01:18 23:00:43
Exposure Time	1/538 sec
F-Number	f/2.2
Exposure Program	Manual
ISO Speed Rating	400
Lens Aperture	f/2.2
Brightness	4.9 EV
Exposure Bias	0.3 EV
Metering Mode	Pattern
Flash	No Flash, Compulsory
Focal Length	4.15 mm
Color Space Information	sRGB
Image Width	538
Image Height	390
Exposure Mode	Manual
White Balance	Auto
Scene Capture Type	Standard

File: IMG_3566.jpg (963KB, 800x692px) Image search: [Google]

963KB, 800x692px

>>3005268
>>3005266
And here is the blinkies using auto wb, which looks pretty good, but the blinkies are very inaccurate

[EXIF data available. Click here to show/hide.]

Camera-Specific Properties:
Equipment Make	Apple
Camera Model	iPhone 6s Plus
Camera Software	Snapseed 2.14.141428617
Sensing Method	One-Chip Color Area
Focal Length (35mm Equiv)	29 mm
Image-Specific Properties:
Pixel Composition	Unknown
Horizontal Resolution	72 dpi
Vertical Resolution	72 dpi
Image Created	2017:01:18 23:26:26
Exposure Time	1/239 sec
F-Number	f/2.2
Exposure Program	Manual
ISO Speed Rating	400
Lens Aperture	f/2.2
Brightness	3.6 EV
Exposure Bias	0.3 EV
Metering Mode	Pattern
Flash	No Flash, Compulsory
Focal Length	4.15 mm
Color Space Information	sRGB
Image Width	800
Image Height	692
Exposure Mode	Manual
White Balance	Auto
Scene Capture Type	Standard

>>3005259
Auto-ETTR metering in CHDK is better.

>>3005272
but i love my gr. anyway, website found magenta color uniwb for Canon 350D to be
>R=162, G=64, B=104
gr is
>R=172, G=64, B=91
so i think an approximate magenta hue is more than sufficient to make highlight clip warnings far more accurate

>>3005270
>>3005268
TLDR: look at the difference in the highlight warning. Same exposure, same light except one is with 'uniwb' and the other is with correct wb

>>3005276
Sorry, so what is the point of this? To get your in camera histogram to most closely match a raw histogram?
If you're going to have to roast all the same channels in post to get a colour balanced image anyway, what's the point?

>>3005327
What do you mean roast all the channels in post?

>>3005327
It's just helping with highlight clipping. It has its problems but it's better than before. More important than uniwb is how many stops above a spot metered subject, using the in camera meter, clipping starts. Around 3 stops but I'm gonna test it tomorrow. None of this matters if you have alight meter. I don't and don't care to carry one since GR is pocketable

Umm... just so you guys know, the blinkies and the histogram are based on the JPEG preview generated, not the RAW. They are mostly just a guideline and you should know there is some leverage left in your RAW for processing.
This uniwb is bs.

>>3005374

thanks for your input, guy who just found out about uniwb.

>>3005375
Why do you think you need it in the first place?

>>3005378

To avoid clipping highlights, my man.

And it's not "why I think I need it", since I've been using uniwb since before you were even "into photography". It's just a simple fact: uniwb is a useful metering tool.

>>3005382
You don't need it to avoid clipping, you need practice and a quick look at the histogram. Knowing how much leverage your system has you can pretty much nail your exposure.

>>3005388

>tells me to look at the histogram
>doesn't realize this undermines his entire argument against uniwb

Neck yourself before you wreck yourself my man.

>>3005401
So you are saying you lack the experience and therefore you need to make up for it with some awkward discoloring method instead of just going out to practice more?

>>3005261
>(2x green photosites)
No faggot that is not because of quantity. Idiot.

>>3005448
If there weren't 2x sensels then I don't think uniwb would be relevant. The magenta hue produced in the test cancels out scaling of the other channels. White balance affects jpeg but not raw, so using uniwb is like looking at a jpeg with no white balance. You would want linear curve too for your jpeg. The reason it is green is because if all sensels are 1x multiple, and there are 2x many green sensels, then green dominates.