Green cast on night scene (Fuji Superia)

[kevin_young2]

hi guys,

 

i recently took some photos from New York City. Among them are some

night scenes from lower manhattan. It turned out there is a green cast

on almost all frames. But my B&W turned out to be alright. I was

wondering if this has anything to do with the lights of the buildings,

or my negative film problem (it has been expired for 9 months and i

did not frid. stored it, yes, i know i was really idiot. just been

moving around a lot lately).

 

http://www.photo.net/photodb/member-photos?user_id=2030383

 

The two photos were from the same roll shot at the same night. The

negative was scanned with Dimage Scan Dual III with "Color Negative"

selected as the source. As you can see, one shows a lot more green

cast while the other one showed more magenta like. The negative seem

to be correct exposed. Just the color is off. As a matter of fact,

the magenta-like one is very close to what my Canon S50 digital camera

shots which was auto-white balanced.

 

Any one of your experts can point out what the problem is?

 

Thanks a lot

 

Kevin<div></div>

[sndr]

Fluorescent lights are generally green on film. Depending on what film exactly, you might get different colours though.

 

The green effect seems to be a combination of the wrong color balance (for most white fluorescent lights) and the response of most films on the very narrow spectral emission bands of the amber-coloured street sodium vapor street lights.

 

Both types of light need their own filter for daylight-balanced films, but yellow sodium vapor street lights seem to do much better with tungsten-balanced film

[wbowman]

Fluorescent lights (and most other types of artificial lights) always shift to the warm or cool

side of the spectrum. Personally I think the photo on the left with the green looks fantastic!

[edgar_njari]

Yea, If I were you I'd complain about the one on the right. The green one looks just right to me

[James G. Dainis]

I agree with Wesley and Edgar.

[kevin_young2]

Thank you all for you wonderful input!

 

What really bothers me was the totally different turn-out of the two

negatives. Taken under same condition, developed in the same lab on

the same day, scanned in the same scanner...but turned out to be one

greenish, one magenta like. I was confused...

 

So Wesley, my understanding about what you said is that the shift of

fluorent lights is unpredictable on the same color negative? It

can go either side of the spectrum? Did i get it right?

 

again, thanks a lot

[kevin_young2]

Sndr,

 

As you correctly pointed out, the sodium vapor street lights did better

than the fluorecent lights. My question is, with a scene having both

these two types of lights, what are the right filters to choose?

 

Is that why i searched on the internet, most photos of lower mahanttan

in NYC are b&w. As I mentioned, my b&w turned out to be acceptable too.

 

Thanks

 

Kev

[kevin_young2]

one more thing, as i mentioned earlier, the magenta-like (right one)

negative looks just like those that shot with my Canon S50 digital

camera. I set the white balance function to be automatic. Is this

just coincidence or something?

[roger_smith4]

"one more thing, as i mentioned earlier, the magenta-like (right one) negative looks just like those that shot with my Canon S50 digital camera. I set the white balance function to be automatic. Is this just coincidence or something?"

 

Do you have white balance enabled in your scanner software? If so, then your scanner might come up with a similar result as your digital camera.

[kevin_young2]

No, I am sure i did not turn on any scanner color correction function.

That's how i got suspicious about the quality of the film i was using.

I was worried the agitation of the film due to the heat might cause the

inconsistent variance.

[josphy]

Maybe the scanner was just trying to add some extra magenta in the second to "correct" all that green, but for some reason didn't in the first shot. I'd scan it again and just adjust the color to whatever you like. I don't think there is really a right or wrong color (within reason) for cityscapes at night because there are such a variety of different color temperatures/white balance issues.

[sndr]

There's no real way to deal with two different lights/color temperatures; unfortunately you'll have to choose which one's more important, or spend some time color correcting during individual areas post processing. On the other hand, it may present you with interesting aesthetic opportunities, as others have noted: your greenish picture looks great.

 

The other 'random' factor is the scanning software of the Scan Dual III. As far as i can tell (I have the same scanner) the Minolta software does something similar to 'auto color' in Photoshop: it tries to spread the histogram in each colour channel in the hope of getting at least two points right: white and black (It also does some fairly aggressive clipping). It may be that this color correcting goes wrong sometimes.

[discpad]

Most likely the film was processed correctly: It's hard to screw up C41 film. Also, the process requires pretty aggressive agitation; and processors are designed to supply this action.

 

What you are seeing is the attempts by the scanner software to white balance, i.e. set the white point.

 

MY SUGGESTION: Shoot some RAW files with your S50, and then open them up in Photoshop CS or CS2. Under the "Adjust" tab you'll see a pop-up for "White Balance" and two sliders for "Temperature" and "Tint" The "Temperature" slider shifts the image on the yellow-blue axis, while the tint shifts it on the green-magenta axis (which nicely compensates for the nasty green cast caused by fluorescent lights).

 

My Photoshop recipe, which works reasonably well for the RAW files from Fuji S2 & S3 cameras, is to look at the "Auto" and "As shot" values for temp and tint, and then average those values. [NOTE: When you shoot a RAW file, the white point is calculated but not applied: The values are appended on to the file as a tag, which Photoshop's RAW import filter reads and applies for the "As shot" choice.

 

These Photoshop "exercises" are actually quite important when it comes to film, as you can see, in real time dynamically, changes that in years past required shooting with a variety of filters and then processing the film.

 

-----------------

 

I ran both photos of yours through the Auto Color algorithm settings suggested by Bruce Fraser -- Ones that I find too aggressive, requiring me to "Fade..." to 50% under the Edit menu -- And you'll see that the overall scene basically fooled the auto white balance not only in Photoshop, but also most likely in your scanner software, too.

 

Hope this helps!

 

Dan<div></div>

[discpad]

Oops! I forgot to give you my Auto Color recipe! :( These work in Photoshop 7 and up.

 

Open up any file, and select Image --> Adjustments --> Levels (or Curves -- Makes no difference). Then, select Options... and you'll see a new dialog box popping up titled Auto Color Correction Options: THIS IS THE BURIED TREASURE!

 

Select "Find Dark & Light Colors" in the Algorithms section, and be SURE to select Snap Neutral Midtones box. Set the Target Colors & Clipping to 0.02% (the default of 0.1% is way too aggressive); and finally, be SURE to check the "Save as defaults" box. Click OK, then click OK in the Curves (or Levels); then close the file and "Don't save." [You don't have to save the file, but you have to click OK to the two dialog boxes else the settings won't "stick."]

 

Hope this helps!

[kevin_young2]

thank you all for your wonderful inputs, especially to Dan for taking time showing it to me and the nice recipe. The reason i posted the

question, is to learn. now it is safe to say that i learned more than

what i ever expected. this is a wonderful community.

 

now back to my original question, about the color inconsistency. one

thing suddenly struck me last night was that:

i somehow remembers that during the shooting, for a little bit time,

i probably had 81B in front of the lens by accident ( i mean not

intentionally). then i realized it and removed it from the lens. That

is likely how the magenta photo was produced. Yeah, i know. I screwed

it up...

 

That means Canon S50 auto WB function is actually putting on a warming

filter in this type of situtation to "correct" the light tempature.

As Dan demonstrated, the effect of 81B filter, S50 auto WB, and PS

Auto Color are essentially similar.

 

Again, thanks a lot. Will start to come this website more often

and post some of my pictures up for your critique which will

definitely help me improve .....