| Author | Thread |
|
|
01/22/2006 06:42:19 AM · #1 |
I finally got a Saturday with no work and no holiday plans so I've been reading and experimenting. One thing I did is photograph my grey card (18%) to see how accurate the meter is. It consistently comes up 1/2 stop under-exposed with the grey card filling the view -- no shadows, no hot spots.
So, I search the net and find several others stating thier Nikons underexpose as well, anywhere from 1/3 to a full stop under, but most usually 1/2 a stop.
Now, I'm not one to jump to a conspiracy conclussion -- even with circumstantial supporting evidence. The grey card I have is an off-brand, but was the only one available at the only camera store near here. When I first opened it my first reaction was that it looked darker than what I had alwasy pictured middle grey to be -- but having never seen a calibrated 18% grey card before, I believed it. It is a 'Delta 1' 18% grey card that says it is made by CPM, Inc; said to have an accuracy of +/- 1%.
Any other Nikon users (or other cameras, out of curiosity) care to photograph a grey card and see where the histogram spikes. I had to adjust with +1/5 EV to get the spike in the middle of the histogram, what adjustment do you need (if any)?
Needless to say, I'm perfectly willing to believe the grey card is indeed darker than it should be. ;)
David
|
|
|
|
01/22/2006 10:14:49 AM · #2 |
| I don't have a grey card but my d70 generally underexposes shots by about 2/3 of a stop when shooting RAW. I'm just making this up but given that it's a "consumer" camera I wonder if Nikon did that to try and "help" the average joe out there keep from blowing out their highlights all the time... |
|
|
|
01/22/2006 10:21:15 AM · #3 |
I don't think the d70 generally underexposes shots -- I think it tends to err on the side of the shadows when you have a really contrasty, badly lit scene...and that's why most people think it underexposes...at least, mine doesn't have any underexposing issues.
I don't have a gray card either David, or I'd check for you...good luck.
|
|
|
|
01/22/2006 06:53:37 PM · #4 |
Thanks for the replies so far. Just to clarify, I haven't found it mentioned overly much, but those that do mention it seem to view it as a Nikon thing, applicable to all Nikon cameras. The end result is usually to shrug, apply a compensation and move on. Fair enough, any complex tool needs a bit of calibration from time to time. So the problme for me is to now find out how much (if any) compensation I need.
Dan, the note you mentioned about it being in RAW that it underexposes prompted me to make a few more tests. The result is that every metering mode at every quality mode produces a histogram that is from 1/3 to 1/2 a stop underexposed when photographing my grey card.
It was while taking these test shots my brain kick in and I remembered anything metered should be photographed as middle grey -- that is just how the meter works. So I photographed a blank white surface (the inside of the box the D70 came in). The results were the same as with the grey card -- every shot is about a 1/3 stop underexposed.
I repeated the same tests with a piece of black foamboard and found it to be underexposed by 1/3 a stop in RAW -- did not bother with jpg since previous tests have indicated the jpg conversion alters the results.
So I need to compensate from +1/3 stop at the shadows and highlights to +1/2 stop at the midtones. I can easily assume the grey card is (as my first impression indicated) just a bit darker than it should be and apply a +1/3 stop adjustment across the board. This is easily calibrated with EV compensation, but I'm a bit anal about wanting dials to read accurately.
Anyone that wants to help me know if it is just my D70 or all Nikons by photographing a few solid toned featureless surfaces (such as a a piece of paper) to see how their results compare to mine, I would appreciate it. Thanks.
So now my question for all of you is, can I create and load into the D70 a custom curve to perform this calibration on every shot? Yes I've read the manual, and am looking thru it again to see if I missed it, but many eyes see what few do not.
/edit: to add that I have found the relevant section of the manual. Yes custom tone curves can be added, they are applied as the image is written to the card -- but it doesn't say if they are applied to RAW images. I'm thinking they are not, so won't generally be worth my time to make one (or get the software to load it), but I'll add this question as well, do custom tone curves affect RAW images?
David
Message edited by author 2006-01-22 19:01:54.
|
|
|
|
01/22/2006 06:59:25 PM · #5 |
| The 8700 also underexposes by about 1/2 stop on a 18% grey card. |
|
|
|
01/22/2006 07:07:41 PM · #6 |
Yes, you can load custom curves to the D70 - you have to use Nikon Capture.
Here is an article which may be informative: //fotogenetic.dearingfilm.com/custom_tone_curves.html
|
|
|
|
01/22/2006 09:19:06 PM · #7 |
Well, this is certainly turning into a much deeper well than I intended to dig.
Joe, thanks for the link -- it was very informative.
It turns out I've stubbled upon a very old arguement that has, as yet, no definitive solution.
For those who didn't read the linked article (and many linked threads), the problem arises due to a difference between how middle grey is calculated. The press people use a measure of reflectivity to calculate middle grey (which comes to about 18% reflectivity) while the meter people follow the ANSI standard and calculate middle grey based on luminousity (which comes out to about 12% reflectivity, 1/2 stop less).
After reading the article and many linked articles and forum threads (which in turn linked to many more) I understand why those I had come across before had simply shrugged their shoulders, made an adjustment and moved on. After looking at the options of how to make the adjustment, a custom curve certainly seems the best approach -- but I don't like that I have to purchase a seperate program to load one.
Does anyone know if Nikon Capture is the only program that is capable of loading a custom curve into the D70? Anyone have any experience with Nikon Capture as an image processing application? Is it worth the money as a RAW converter?
David
|
|
|
|
01/25/2006 07:50:10 AM · #8 |
I can help you with your last question...
I was underwhelmed by the speed of Nikon Capture at processing. I am using a Mac (and a slightly underpowered one, at that) so YMMV, but once I got Photoshop CS I switched entirely over to Bridge and Adobe RAW.
The CD that came with your camera should include a 30 day free trial of Nikon Capture, and if you can't find the CD the download is free and comes with the same 30 day free trial. So you can at least try it out, and assuming you can make the adjustments you need within 30 days, shouldn't need to buy the program just for that reason.
One warning to heed: They've released several upgrades to Nikon Capture, and they're all progressive, and you can't download the latest and greatest version from the website. So you HAVE to install 1.0, then patch to 1.1, then patch to 1.2, etc. Do it. It's worth it.
I haven't even started to research this yet, but like you, I'm a little bothered about having the EVComp +/- icon on for eternity, so if it works for you and you wouldn't mind sharing the steps to do it, I'd be very appreciative.
|
|
|
|
01/25/2006 08:28:54 AM · #9 |
Like livitup, I was not impressed enough with Capture to fork over the dough for it. It is annoying to have to pay so much to use an advertised feature of the camera.
But whatever. I did a little more searching and found "ToneUp" - I have not tried it, but it may meet some people's needs. If only there was a Mac version!
Here are some product details: ToneUp. |
|
|
|
01/25/2006 08:56:15 AM · #10 |
First - to your wondering about the gray card - it is NOT suppose to be mid-gray!
Do you have any idea what the 18 means? It means that this card only returns 18% of the light falling on it. This is why it is a dark gray.
If it was supposed to be middle gray, it would have been called 50 gray. Wouldn't it?
As for the digital cam's exposure - there is a certain "fix" of exposure in the cam. But it is not a certain number that you can messure.
A simple experiment can prove this:
If you take a film camera and shoot at normal exposure a full frame of a totaly white board, you will get a photo that is actually gray 18.
You will get the exact same result if you shoot an all black board.
This seems to be an error in exposure, wouldn't you say?
Well, this is how a camera knows how to expose at any scene.
Now do the same thing with a digital camera and you will find out that the results are slightly different: the white board will be a lighter gray then 18, and the black board will be slightly darker then the 18 gray.
If you think of those results compared to the theoretical results you have been expecting, you can understand that in the black scene - your cam was under-exposing, and in the white scene - it was over-exposing.
Thing is, the digital cam calculates the color it sees and tries to minimize the "errors" in exposure. After all, you want the wall to be white. Not gray 18, don't you?
So your cam's meter could be a bit wrongly calibrated, but to be honest - I doubt that very much.
Check the scene you are shooting, and compare results each time.
And if you want to be sure you're getting the perfectly exposed photo - just use braketing ;-)
Edit: one day I will learn to type
Message edited by author 2006-01-25 09:01:40. |
|
|
|
01/27/2006 05:18:40 AM · #11 |
Originally posted by Jinjit:
First - to your wondering about the gray card - it is NOT suppose to be mid-gray!
Do you have any idea what the 18 means? It means that this card only returns 18% of the light falling on it. This is why it is a dark gray.
If it was supposed to be middle gray, it would have been called 50 gray. Wouldn't it? ... |
No, grey cards are middle grey -- that is the whole point of them. They are a standard reference to allow a photographer to meter the exact middle tone in any light. They are suppose to be the one thing the meter will always place at the exact tone it already is -- middle grey.
It is called 18% grey because it reflects 18% of the light that falls on it. But the reflectivity of light is not measured linearly (1,2,3, etc), but on a log scale. This is why exposure stops are numbered as they are -- that is, each increase of a stop doubles the amount of light and each decrease halves the amount of light. Looked at another way, reducing the exposure by a half in the same light will produce a shade of grey one stop less bright.
Think of a light source as being pure white. A 100% reflective surface would be the same brightness as the light source -- and a 50% reflective surface would be 1 stop less bright. A very (very!) light grey.
Originally posted by Jinjit:
... As for the digital cam's exposure - there is a certain "fix" of exposure in the cam. But it is not a certain number that you can messure.
A simple experiment can prove this:
If you take a film camera and shoot at normal exposure a full frame of a totaly white board, you will get a photo that is actually gray 18.
You will get the exact same result if you shoot an all black board.
This seems to be an error in exposure, wouldn't you say?
Well, this is how a camera knows how to expose at any scene. ... |
The camera (or actually the meter) doesn't know how to expose anything -- it only knows how to measure the exposure needed to make the tone of the scene match it's reference tone -- middle grey. Adjusting the suggested exposure to properly expose the scene is something the camera can not do with any great reliability. Again, this is why a grey card is so valuable -- it is the one object that is always measured at the tone it is.
Originally posted by Jinjit:
... Now do the same thing with a digital camera and you will find out that the results are slightly different: the white board will be a lighter gray then 18, and the black board will be slightly darker then the 18 gray.
If you think of those results compared to the theoretical results you have been expecting, you can understand that in the black scene - your cam was under-exposing, and in the white scene - it was over-exposing.
Thing is, the digital cam calculates the color it sees and tries to minimize the "errors" in exposure. After all, you want the wall to be white. Not gray 18, don't you? ... |
If you meter is doing the above -- it is broken. Either that, or you have your camera set on one of the automagical metering modes that attempt to calculate the exposure for you and are including other tones for it to compare the scene against a database of potential scenes. No telling what exposure you'll get in that case. The exposure reading I was taken earlier were made with the camera set to the spot metering mode.
Originally posted by Jinjit:
... So your cam's meter could be a bit wrongly calibrated, but to be honest - I doubt that very much.
Check the scene you are shooting, and compare results each time.
And if you want to be sure you're getting the perfectly exposed photo - just use braketing ;-)
Edit: one day I will learn to type |
Read this article, "Meters Don't See 18% Gray" linked from the above linked article. It explains how meters are calibrated against a tone of grey slighting darker than 18% grey (about 12% grey, or 1/2 a stop darker than 18%) by calibrating against the luminance midtone instead of the reflective midtone. Not wrong, just calibrated against a value that has no real-world application for me. When I meter a scene I am metering reflective light, not luminance -- so all images are under-exposed by about 1/2 a stop.
Thanks for replying.
David
|
|
|
|
01/27/2006 05:23:14 AM · #12 |
Andrew, Joe and the others,
thanks for the help (and the links). :D
Looks like I'll actually have a second Saturday in a row off (first in half a year) so I plan on looking into custom curves more closely then -- will report my finding and results when I have some.
David
|
|
|
|
01/27/2006 06:28:15 AM · #13 |
I shoot with the White Wedding, almost exclusively. Brings up those midtones by probably a third, maybe 1/2. Never took the time to compare, because I like the results as is.
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/03/2025 05:08:38 AM EDT.
