- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
Brain dead I must be....I have ViewNX2. Photoshop CS4..Rawnalyze..Adobe DNG convertor. How do I measure the gray card RGB values in ViewNX2?
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
New for RAW - View NX2 v. 2.2.3 - anyone used it?
- Thread starter Mikedigi
- Start date
Yeah, I think 2.10.4 is the latest. I am unsure what to do about your D7K files. You can try converting them, But at the moment I don't know how that affects the data.
Still...you should be able to see the effects between ViewNX and Raw Therapee and other free converters that can show you a demosaiced image without any additional processing. Even with additional processing, there should be a big difference in brightness.
.
Remember that the camera is trying to make everything gray. As you've probably read me write in the past, take a picture of a black wall, it comes out gray...take a picture of a white wall, it comes out gray. So then the next question is...which gray is the camera trying to make? Well, all you have to do is to take a picture of any evenly lit, neutral surface, and the camera will adjust exposure to make that surface gray. Use that surface to set a custom white balance.
The gray that you get is the gray that the camera is trying to turn the world into. I find that the resulting gray level varies based on the conditions of the exposure...but it's always 100 +/-10 when using Spot metering. For Matrix metering the results are either the same or sometimes 1/3 stop more. That's because Spot and CW metering are calibrated to 12.7% reflectance, while Matrix metering is supposed to produce 18% gray (about sRGB value 119.) So a high Spot reading may overlap with a low Matrix reading.
In Raw Therapee, the Neutral profile will show you the gray level that matches what's in the RAW data. Even with one of the postprocessing profiles applied, the values don't change that much...certainly not like it does in ViewNX.
.
Jack Hogan
Veteran Member
The 8 bit RGB values are shown on top of the image, to the right, in parentheses next to the coordinates. However, don't lose sleep over it because, when unmodified, those values are modified by a number of parameters which have nothing to do with exposure compensation. I think Graystar is mixing apples and oranges.
Jack
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
I figured it out and see exactly what you are saying.......But, When I open up the NEF in RawTherapee, it is indeed darker but nowhere close to what my eye saw in the actual scene. The ViewNX image does have higher RGB values but the scene is exactly (much closer) to the actually brightness/exposure I saw when the shot was taken. Will do a gray card next.....and look for the 100/100/100.
Jack Hogan
Veteran Member
Yes, of course, brightness - not 'exposure'. Different raw processors apply different brightness and contrast curves by default when rendering an image. All cameras (including Nikons) and raw processors (including VNX2, CNX2 and RT itself) show a brighter image than RT neutral because RT Neutral is dull and bland and useful only in very limited circumstances.
The difference between adding brightness through curves vs 'exposure' compensation is huge: in the first case you do not lose any information and any change (even if performed under the hood) is reversible; in the second case you may lose information and changes may be irreversible - a big no no.
The conceptual error you are making is assuming that Nikon does not apply any curve to the data before it passes it to your custom PC. It does.
Cheers,
Jack
That is exactly right. Remember, the camera is trying to make everything a specific gray. The only subject that will photograph correctly is a gray card that is the perfect shade of gray! THIS is why we meter gray cards! If you meter the right shade of gray, everything else comes out correct.
The well-known Kodak 18% gray card is about 1/2 stop too bright, and the instructions with the card tell you increase exposure by 1/2 stop. Why? Well, bright objects need more exposure, right? (white wall turns gray and needs +EC to become white again.)
Lastolite makes a 12% gray reference that doesn't need any EC.
.
I meant shouldn't be a big difference.
There's nothing wrong with the application of default curves. However...1. You can disable all that and apply your own in all processors except ViewNX, and 2. the default application of curves doesn't mean that the image automatically gets brighter. With RT the image may be brighter or it may be darker...depending on the postprocessing profile you select. So it's not a foregone conclusion that the image will be brighter.
The difference doesn't matter if it's causing highlights to blow out. ViewNX appears to be adding exposure compensation.
All RAW processing is reversible. No RAW processor modifies RAW data.
If you had fully read my previous post you'd know that my complaint against View NX is that it IS applying processing to the data before it gets to my picture control. There's no way to remove that processing. And that processing appears to apply a large amount of positive EC to an image when it is unnecessary.
.
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
Ok, did it with a grey card just as you said and you're right....readings in Raw Therapee(RT) approx 100/100/100. ViewNX2 140/140/140 Colors in both are spot on...but the Raw Therapee image is indeed much darker than the ViewNX2 version with the ViewNX2 version being near perfect to the actual exposure of the real scene as my eye sees it. To get the RT image to match the VNX2 I have to add .7 EC to the RT slider. I have to Trying to learn how to interrupt that. Thanks again for the free lesson
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
For me it does...no dif now changing magnification and everything is very sharp. ViewNX2 2.2.3 with ATI 5870 card and 11.10 drivers.
Jack Hogan
Veteran Member
But that's exactly the point: curves are reversible and do not cause highlights to blow out, while exposure comp can cause blown highlights.
True, but when you say that Nikon Cameras and software apply exposure compensation under the hood (remember, exposure compensation is irreversible) you are saying something else: potential blown highlights, which is not the case. This is the reason why I took exception at your comment in the first place.
It's not EC. And I wasn't just talking about the raw processor, I was talking about your camera as well.
See...that's a huge difference, equal to about one stop in brightness.
Yes...I don't know what gray card you have, but like I said, an 18% card is too bright by 1/2 a stop, so an image of one is always a little darker than what you see. If you were to take a picture of a darker gray card, you would get the exact same results...100 and 140. In fact, try it with something black and with something white...you should get the same results (though be prepared for long exposure times with the black object.) Again, it's because the camera is trying to make whatever is in front of it into the same shade of gray.
I solved the problem by not using ViewNX. Highlight recovery is much better!
No prob!Thanks again for the free lesson
.
Jack Hogan
Veteran Member
Thank you for using the word 'brightness' and not 'exposure' compensation
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
Thanks again....just shot with a pure black and a pure whit. Raw Therapee does result in a 100/100/100 and VieNX2 in a 140/140/140. Problem is the RT seems to push brightness by 1/2 to 1 stop to far. In this case ViewNX2 seems a more accurate starting point vs RT neutral. I'll have to explore RT more. The learning continues.See...that's a huge difference, equal to about one stop in brightness.
Yes...I don't know what gray card you have, but like I said, an 18% card is too bright by 1/2 a stop, so an image of one is always a little darker than what you see. If you were to take a picture of a darker gray card, you would get the exact same results...100 and 140. In fact, try it with something black and with something white...you should get the same results (though be prepared for long exposure times with the black object.) Again, it's because the camera is trying to make whatever is in front of it into the same shade of gray.
I solved the problem by not using ViewNX. Highlight recovery is much better!
No prob!Thanks again for the free lesson
I used "brightness" due to the visual focus of the exchange. But numerically speaking, the addition of 1 stop of EC is equal to 139 in sRGB, which is equal to a doubling of the values from the RAW data. I can make an image in RT look just like the rendition in ViewNX by applying EC.Thank you for using the word 'brightness' and not 'exposure' compensation
And I'm not sure why you think curves won't blow highlight data. They can do so very easily.
.
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
Based on the exploration here I've been thinking....I've never had a problem with how the D7K exposes and have never seen a need to add a negative exposure except in specific cases. Many set one globally. My workflow is Raw shot/check histogram and make an adjustment if necessary/ViewNX2/ then CS4. The results have been very accurate so far. That would make me think that differences in workflow can have a great impact on how one might set up their camera (and be a valid reason in doing so). I'll have to account for that in the future.
Exactly. This is a great demonstration of how a reflective light meter works. The meter always makes the same presumption as to what's in front of it.
That statement I don't understand. The black and white cards prove that neither has an accurate starting point. Accuracy in the starting point is gained by correctly metering a proper, 100, 100, 100 gray reference (because that's what your camera wants.) Then RT will be accurate, and ViewNX will appear overexposed. Alternatively, you can get a 140, 140, 140 gray reference and then ViewNX will appear correct and RT will be underexposed. Why? Because the brighter reference will cause the camera to underexpose (metering snow underexposes and you need +2 EC.) But you'd get more noise this way.
Are you talking about the default processing of RT? Then I would agree that it tends to boost the EC up. It's one of the auto-levels tools that's doing it. I don't use that profile anymore, but when I did I would just restore the Exposure Compensation back to 0.
.
- Messages
- 30,186
- Solutions
- 112
- Reaction score
- 12,428
Just a question then, and I'll stop bothering. If you shoot a scene (being carfule to get the exposure and WB right) and the result in ViewNX2 is always accurate in terms of brightness/color/and perceived exposure...while the same in RT requires EC, why use RT or neutral? I'll start exploring RT features more as I'm sure the answer is there...............I withdraw the question, Have a great dayI used "brightness" due to the visual focus of the exchange. But numerically speaking, the addition of 1 stop of EC is equal to 139 in sRGB, which is equal to a doubling of the values from the RAW data. I can make an image in RT look just like the rendition in ViewNX by applying EC.Thank you for using the word 'brightness' and not 'exposure' compensation
And I'm not sure why you think curves won't blow highlight data. They can do so very easily.
Keyboard shortcuts
- f
- Forum