Ad
Ad
Ad
Pages: « 1 [2]   Bottom of Page
Print
Author Topic: Rosco #002 gel 'ed light and Adobe's fail to handle such simple light  (Read 8316 times)
BartvanderWolf
Sr. Member
****
Online Online

Posts: 3468


« Reply #20 on: December 26, 2012, 04:39:22 PM »
ReplyReply

Bart, could you assign your monitor profile to that screengrab, save and repost and even better yet if so desired convert to sRGB? There's no embedded profile in the image you posted.

Hi Tim,

Sure. Attached the same file and name, now as JPEG with my displayprofile attached (requires conversion to ones own profile or sRGB to be meaningful). Then I've attached the same Raw conversion saved directly from Capture One as sRGB JPEG output (the Reds/Oranges are less oversaturated).

I also tested the Capture One ProPhotoRGB 16-bit/channel TIFF converted to SRGB in Photoshop route, to satisfy my own curiosity and to see if Photoshop produces a different result from the direct Capture One sRGB output (different CM engines and potentially JPEG quality, but they are close).

Cheers,
Bart
« Last Edit: December 26, 2012, 04:50:22 PM by BartvanderWolf » Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5427


WWW
« Reply #21 on: December 26, 2012, 05:14:42 PM »
ReplyReply

I'm not a long standing pro photographer going back to the film days (right now just a digital imaging enthusiast), so if Rosco says they've been around 100 years according to that video, I'll just take their word for it.

Rosco has been the default gel supplier for motion picture and theatre for a really long time. I keep an entire cabinet stocked with color temp (CT) correction and color effects sheets and multiple rolls of CT for putting in windows. Only studio photographers (or others who do special lighting) would be likely to know the brand.

As for what the OP was doing by putting the 002 gel over the light and expecting to get a perfectly balanced WB results, I don't know...but if it hurts when you do that, I would suggest not doing that, ya know? I do know that ACR/LR's WB tool is designed to "tween" between D65 and Standard Illuminate A (2856 K) and to be used to correct WB  spectral radiance of a black body following Planck's law...pretty sure it wasn't designed to correct for a whacked out pink gel that transmits very little light between just under 460-580 nanometer...spiky, incomplete spectral illumination is not really optimal for digital sensors...

Edited to change transmits "almost zero" to be "very little light" between just under 460-580 nanometer. I was looking at the wrong pink Rosco gel when I wrote that–my bad (but doesn't change the overall post).
« Last Edit: December 26, 2012, 06:15:55 PM by Schewe » Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1147



WWW
« Reply #22 on: December 26, 2012, 05:44:36 PM »
ReplyReply

Thanks, Bart. I'm going by Lab readouts in Photoshop to tell me how off the color patches are. I'm assuming the gold tinted yellow patch is a result of WB for the Rosco gels or did X-rite change the pigment formulation for yellow to a gold color? My GretagMacbeth CC chart shows yellow as L=82, a=4, b=80. Your posted image shows yellow L=70, a=25, b=73 with slight variances between the three image showing the same yellow.

Thanks, Jeff, for the clarification on the history of the Rosco gels. Now that quite a few Cinema styled video footage is being recorded off DSLR type devices which surely have better WB features, is the Rosco gels still useful by and large for other digital applications or are they still primarily useful and functional for film only?

That YouTube video doesn't give desirable results but it's better than the before but then I never get before shots that bad under fluorescent lights with my DSLR anyway.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5427


WWW
« Reply #23 on: December 26, 2012, 06:02:17 PM »
ReplyReply

Thanks, Jeff, for the clarification on the history of the Rosco gels. Now that quite a few Cinema styled video footage is being recorded off DSLR type devices which surely have better WB features, is the Rosco gels still useful by and large for other digital applications or are they still primarily useful and functional for film only?

CT gels are still as useful for digital as they were for film (still & motion) and every play I see at Steppenwolf (a theatre we go to here in Chicago) uses colored gels. They are useful for correcting mixed light scenes or to add effects. Say you are shooting tungsten interiors and you have window light coming in...no real good way to correct that in post but if you hang CT warming gels over the windows you can correct when shooting. Think if them being useful for "local" color corrections or effects. If your scene is from a single white balance though, there's no need to worry about the gelling–although I have added cooling CT filters over tungsten lights to cut down IR and help with the white balance–sensors under tungsten light get very little blue light which can lead to difficulties if you need real accurate color.

In any event, I don't suggest shooting with a really pink gel and expect do get a great white balance :~)
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #24 on: December 27, 2012, 10:02:09 AM »
ReplyReply

I find it a strange coincidence that these Rosco filters are getting mentioned several times between two photo discussion forums within the last month namely at LuLa and Photo.net when I've never heard of the brand for the ten years I've participated in both forums.

Rosco is a big name in that area... Rosco, Lee, GAM, Apollo... may be one, two more
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #25 on: December 27, 2012, 10:05:13 AM »
ReplyReply

Maybe complaints to Rosco's gel/filter formulation should be addressed to them rather in this thread and/or complain directly to Adobe about this.

there is no issue w/ the gel - it was selected on purpose... the only issue here is with a raw converter, which apparently is incapable to deal with the raw file where you have proper saturation in all channels... take RPP for example and you will get proper output.... or for example SilkyPix can WB the image quite pefectly just by using WB patch...
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #26 on: December 27, 2012, 10:07:30 AM »
ReplyReply

In any event, I don't suggest shooting with a really pink gel and expect do get a great white balance :~)

Jeff... the right text is  "I don't suggest shooting with a really pink gel and expect do get a great white balance in ACR/LR" - proper converters have no issues at all.
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #27 on: December 27, 2012, 10:13:37 AM »
ReplyReply

spiky, incomplete spectral illumination is not really optimal for digital sensors...

on the contrary, Jeff, on the contrary... that gel makes spectral illumination matching CFA properties to get quite even raw channel saturation (albeit ideally the red channel shall be aligned with greens/blue channels - and I still have red trailing, but less than 0.5 EV - but that is not bad at all) than you have w/o it from the light source behind it (flash) and it does not make the original spectrum incomplete (as you can very clearly see from the gel specs posted).
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #28 on: December 27, 2012, 10:19:07 AM »
ReplyReply

Edited to change transmits "almost zero" to be "very little light" between just under 460-580 nanometer. I was looking at the wrong pink Rosco gel when I wrote that–my bad (but doesn't change the overall post).

not very little, not at all... I bet you were looking @ gel that cut part of green spectrum to zero - but the one that was used #002 - is just matches CFA on sensor to let more blue and red through relative to green... and by increasing the power from a flash the total amount of light across the whole spectrum is perfectly good, as you can clearly see from a raw histogram, and no gaps in spectrum, no gaps.

so what we have here is - we have good raw data and N1 raw converter does not work with it... that by itself is very very strange...
Logged
Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #29 on: December 27, 2012, 10:22:03 AM »
ReplyReply

I do know that ACR/LR's WB tool is designed to "tween" between D65 and Standard Illuminate A (2856 K) and to be used to correct WB  spectral radiance of a black body following Planck's law..

indeed and that is the issue... you have a real light, produced with a $6 gel that is being sold in any shop that deals with equipment, software and products for theater, film, television and architectural purposes, from a good source of light behind it and you can't WB it... how come ?
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7257


WWW
« Reply #30 on: December 27, 2012, 12:40:17 PM »
ReplyReply

Hi,

Can you explain the intention?

Best regards
Erik


indeed and that is the issue... you have a real light, produced with a $6 gel that is being sold in any shop that deals with equipment, software and products for theater, film, television and architectural purposes, from a good source of light behind it and you can't WB it... how come ?
Logged

sniper
Sr. Member
****
Offline Offline

Posts: 580


« Reply #31 on: December 28, 2012, 08:12:03 AM »
ReplyReply

If the intention was correct white balance why not use a custom wb in camera?  and why shoot with a pink gel in the first place?  I', courious.
Logged
Steve House
Full Member
***
Offline Offline

Posts: 223


« Reply #32 on: December 28, 2012, 09:41:15 AM »
ReplyReply

Why shoot with a gel on the lights that gives them an uneven spectral distribution in the first place and then try to white balance the colour shift away? Isn't the logical course of action to remove the filter from the equation?  The filter is not shifting the colour temperature of the light, ie the 'white balance' of the source (note the listing for "MIRED shift" on the spec sheet posted), and even if the neutral grey patch was balanced out to grey, other colours in the image are still going to be off due to the distorted spectrum they are illuminated with.  Your thread title is not quite accurate - because of the shape of the transmission curve, the colour of the light coming through the filter is anything BUT simple.
« Last Edit: December 28, 2012, 09:44:24 AM by Steve House » Logged
BartvanderWolf
Sr. Member
****
Online Online

Posts: 3468


« Reply #33 on: December 28, 2012, 10:13:59 AM »
ReplyReply

Why shoot with a gel on the lights that gives them an uneven spectral distribution in the first place and then try to white balance the colour shift away? Isn't the logical course of action to remove the filter from the equation?

Hi Steve,

I agree, unless the combination of the illuminant+gel produces something closer to a normal spectrum than the illuminant alone. In this example it clearly didn't, because it required a combination of 4813 kelvin and -28.8 Tint (!) in the above conversions in Capture One Pro V7 to get a bit closer to reality for the gray patches.

The OPs observation that ACR has more of a problem correcting this poor input quality of lighting is what it is, an observation (and maybe something for Adobe to look into).

Cheers,
Bart
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1147



WWW
« Reply #34 on: December 28, 2012, 11:10:26 AM »
ReplyReply

And while Adobe's looking into that non-problem maybe they can explain how I could WB that jpeg of a heavily green color cast rock climbing wall in the linked Photo.net discussion I provided. We're not suppose to be able to white balance gamma encoded jpegs with editing tools engineered to operate their best on linear data, but apparently we can.

Epic fail on Adobe's part for spreading this kind of misinformation about the functionality of their own software. It's scandalous!

Wonder what the spectral distribution graph for that rock climbing wall image comprised. Apparently the OP's Rosco gel'ed DNG image must've been filtering a scene lit by a green/yellow light according to the Rosco provided spectrum curve. In what place on this earth would a photographer encounter such a light since the OP doesn't even mention the light being filtered.

Another solution looking for a problem.
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2101


« Reply #35 on: December 28, 2012, 11:17:22 AM »
ReplyReply

The unusual lighting is causing the WB to fall outside the ACR/LR standard supported range, which is why the click-WB doesn't work properly.  You can use Adobe's DNG Profile Editor to adjust for this by remapping the WB range.  The key is to use the White Balance sliders in the Color Matrix tab (third tab) of the DNG PE.  Please see Tutorial 4 in the DNG PE documentation, here:

http://wwwimages.adobe.com/www.adobe.com/content/dam/Adobe/en/products/photoshop/pdfs/cs6/DNGProfile_EditorDocumentation.pdf

The tutorial uses the example of an infrared-modified camera, for which similar issues arise.
Logged

Vladimirovich
Sr. Member
****
Offline Offline

Posts: 1280


« Reply #36 on: December 28, 2012, 12:06:16 PM »
ReplyReply

...

thank you ! finally I can go back to using ACR now
Logged
Pages: « 1 [2]   Top of Page
Print
Jump to:  

Ad
Ad
Ad