sirimiri wrote:
Corposant, what did you digitize that shot with? I've pushed Portra a stop in 120, and aside from the shadow splatter, it can be made to look very presentable.
RPL scan, ironically. I didn't even really look at it on my monitor until now, and I can see what's causing the uproarium. Time to launch CS5 and...
voila!
It seems like I had some kind of odd color shifts on this roll - I'm wondering what the 1600 roll will bring.
corposant wrote:
Indeed - the artificial lighting wasn't wonderful, and the push in processing did not help. I was also curious to see how much of the highlights and shadows I could keep - kind of amazing, really. I shot a bunch of Portra 400 at 1600 last week - I am wondering how wonky it looks.
That photo I posted above was Portra 400 pushed to 3200. Still holds up pretty well!
I don't see any lost detail in that shot, in spite of the heavy contrast. I wonder if setting the black and white point and playing with the grey point might get nicer results?
alkanphel wrote:
That photo I posted above was Portra 400 pushed to 3200. Still holds up pretty well!
It looks very good. Did you actually push it 3 stops to 3200 or less? I regularly shoot Portra 400 up to 1600 without pushing (on 120) and it turns out fine. I shot it 3200 once but pushed 1 stop and didn't like the shadows much, I'll try pushing it two or 3 stops next time. That film is amazing, especially for medium format where you generally have a choice of 3.5 shutter speeds...
Joan, if you have the 25, the most reasonable set up is 35, 50, 85, and if you wanna go ultra wide, 18. I personally prefer the 35/2 and 50/2, but all ZM are great, there is no bad choice.
Depending on the scan settings and whatnot you can vary the color quite a bit. But the scene was way more orange. I don't think that straight scan is representative of what the Portra negative actually holds. I know Portra is not supposed to be saturated but here is a straight unedited raw file from a D300s: http://img6.imageshack.us/img6/9784/068svs.jpg
Under-exposed for the sky which increased the saturation a bit but still way more orange than the straight Portra scan. The sky was incredibly orange that evening which is why I don't think the scan is right.
Zaitz - nothing wrong with correcting a scan. Remember, the automatic color corrections in most software will just as readily correct out an *actual* color cast (like orange light) as it will the negative base or a cast due to a color temperature mismatch. At least it will attempt to. So if you remembered the scene orange, and particularly if the D300s version shows that orange, then by all means correct it orange.
The best way I've found to *not* correct out any color casts that were in the actual scene is to do a raw scan, invert manually, and set the black point based on the negative base. Usually though a bit of extra correction is called for after that anyway.
Simon Kennedy wrote:
Nice Zaitz!
Have you tried Vuescan/Colorperfect? The best thing I have found for colour negatives (except for drum scan...)
Schneider 72mm, Fuji Pro 160NC, 5x4".
Thanks! I actually did try colorperfect/colorneg. It definitely gave better tonality but the overall color was similar to the straight Epson Scan. I had the trial version so it had the extremely obtrusive watermark but I could see that the tonality was better.
kidtexas wrote:
Zaitz - nothing wrong with correcting a scan. Remember, the automatic color corrections in most software will just as readily correct out an *actual* color cast (like orange light) as it will the negative base or a cast due to a color temperature mismatch. At least it will attempt to. So if you remembered the scene orange, and particularly if the D300s version shows that orange, then by all means correct it orange.
The best way I've found to *not* correct out any color casts that were in the actual scene is to do a raw scan, invert manually, and set the black point based on the negative base. Usually though a bit of extra correction is called for after that anyway.
Oh, and nice photo ...Show more →
That is pretty much what I was thinking too. The color may be so strong that the software thinks it's an incorrect cast. I haven't shot much more Portra either. Unfortunately I overexposed the Velvia and shot it about 30min too early that day. But it also shows the orange, just one of those days with crazy lighting. I actually left work to photograph it so I know it was quite intense.
From all the searching I've done, Portra seems to have a pretty definable look and quality. It seems to be tough to know what the true negative actually holds when dealing with it on a PC. Processed in a darkroom, is it 'easier' to extract the 'true' color information? Or is it similar to scanning in that you need to make several big judgement calls and work the color yourself? Not that it matters really.
Thanks for the help. I tried the black point method on the raw scan and it works well, or similar to the other methods in this case in that I end up with a decently neutral file. But the orange is not there at all. Actually, just now messing with it, one of the best methods so far for me is using the color channels in Curves to 'correct' the color. I was able to get it neutral and then work in more color. Looks like anyway I do it I am going to have to make up my own mind on what is 'right'. Why can't transparencies have more dynamic range .
I've not printed color in the darkroom, but it should be a somewhat similar experience. The response of the paper is the same, but you set the basic color balance with filters. So you could easily compensate for a color cast with a different filter pack. So yes, you need to make judgement calls just like in scanning. I believe you use a basic filter pack as a starting point, to cancel out the orange mask, but then you usually adjust it for your final print.
One way to help to *not* compensate for actual color casts is to take a shot of a gray card in proper daylight balanced lighting. Scan and correct that photo. Then scan your photo with color casts with the *exact* same settings and use the corrections from the first photo on the second one. When I say use the exact same settings, that's what I mean. If your scanner is adjusting gain/exposure times on the RGB channels or something similar, they need to be the same. If you google 'vuescan lock film base' you should find some tutorials that talk about locking exposures between different scans.
^The easiest way I found to work in a color darkroom, once you get past the absolute inky blackness required, was to use the Kodak color print viewing kit to judge results and adjust the enlarger's dichroic filters based on the directions given per Kodak's filtration overlay frame that you most pleased you.
I'd never want to wet-print color again, though. It's exacting and laborious.
Zaitz wrote:
From all the searching I've done, Portra seems to have a pretty definable look and quality. It seems to be tough to know what the true negative actually holds when dealing with it on a PC. Processed in a darkroom, is it 'easier' to extract the 'true' color information? Or is it similar to scanning in that you need to make several big judgement calls and work the color yourself? Not that it matters really.
Thanks for the help. I tried the black point method on the raw scan and it works well, or similar to the other methods in this case in that I end up with a decently neutral file. But the orange is not there at all. Actually, just now messing with it, one of the best methods so far for me is using the color channels in Curves to 'correct' the color. I was able to get it neutral and then work in more color. Looks like anyway I do it I am going to have to make up my own mind on what is 'right'. Why can't transparencies have more dynamic range . ...Show more →
Zach - interesting architecture there - I definitely prefer the Portra to the digital shot.
There's nothing wrong with processing film shots digitally. Aside from manipulating color channels in "Curves," you might also want to use a Selective Color Adjustment Layer. These are great for either isolating color casts or adding new ones, especially in shadows, where things can get interesting.
You should also take into account who is giving you this advice - I often know how to edit the imperfections in my work but sometimes just don't bother out of laziness, complacency, or just because I like it wonky (see the train platform discussion on the previous page).
Portra 400 is a lot more muted in terms of saturation - you may want to see if you can get ahold of any of the "VC" sheets that still may be out there. Portra 160 VC is one of my favorite all-time films in 120 format. Great colors, without turning people orange, and very tight grain.
.