Questions surrounding UniWB & the D700...

<p>I was having issues with the especially Red channel on my D700. I kept "blowing it". Then a "professional" photographer on DPR told me to do one check by going in to NX2 & doing an extreme negative EV & tell him what I saw. What I saw was a nicely tapering off red channel. So the response was - - I'm not blowing my red channel & that the histogram in the camera & NX2 are tricking me.<br>

OK - so I can deal with that & go back to working on b3ecoming friends with the D700. But then he adds that I should add UniWB to the camera to then later on tell NX2 what I want it to see.<br>

Now, if I've understood this right - then I can suddenly expose my - lets take a red rose - by looking at only the red channel.....<br>

OK - did I understand that correctly. And is it only going to show me the Red Channel - or how does it affect me looking at the other channels?<br>

Since this is UniWB - White Balance - how will this affect my regular WB in the camera & in NX2?<br>

I feel I have been doing this for a few years now - but I've not ever had to go to UniWB before & honestly I don't want to.<br>

Others seem thrilled I started my thread - I feel like someone who doesn't know a thing.<br>

I'm confused. I'm irritated. I don't want to. If in fact my red channel is not blown - & I should ultimately be looking at the white one - the combined one - (but I was told to control my red channel.....)<br>

Please help me make sense of this......<br>

Thank you for your time & hopefully helpful responses....<br>

Lil</p>

 

<p>Lil... I don't know what Uni stands for but I found a solution for that problem. I used to think that my camera had a problem but it seems all DSL's suffer from blown reds. What I do is just to check the histogram (red chanel ONLY) and expossed according to it. If other colors are afected I can also take a normal expossure and then disaturade the red on NX2. I don't know if this is any help at all for you but for me it works just fine.</p>

<p>Rene' -<br>

I think that's kind of what the UniWB does - it shows the red channel from what I understand.<br>

So - does it come down to the fact that the red channel is the most difficult to control & by adding this "file with the information" mean that I will have more access to the red channel - easier if I put it so......<br>

I wonder - - - -<br>

Thanks</p>

<p>Lil :-)</p>

<p>Well, it seems it is the hardest chanel followed closely by the yellow one.<br>

you shouldn't let those little things irritate you! That takes off the fun out of the hobby! :) Give me a few minutes and I will send you a link of what other people told me and explained me how to work around the problem. i have to look for that thread! I'll be back! :)</p>

<p>Lil, I hope this will help a little. The histogram that you see on the LCD in the back of your camera is derived from the JPEG file and with all the in-camera settings for saturation, contrast, color space etc applied - even if you shoot RAW. Without tricks, there is no way to see the histogram based on the RAW image. So, for your "blown" red - the histogram shows that the JPEG has overexposed red channel. In NX, with your negative exposure correction, you see that in the RAW file, the channel was not blown but is recoverable - you get a gentle curve after the application of your exposure correction. </p>

<p>Now, it is my understanding, that UniWB allows you to see the histogram of the RAW file on the camera's LCD - provided that other settings are set properly too (like AdobeRGB as color space, no saturation, no sharpening, linear contrast). Now, the histogram will reflect the exposure of each channel in the RAW file and allow for better exposure control (ETTR, expose to the right)</p>

<p>Of course, since there is twice as many green sites than blue and red on your sensor, the image on your screen will now look green. UniWB makes the three factors that multiply the green, red and blue data all equal to unity - hence UniWB. Normally, the factors for blue and red - based on the fact that there is only half as many as green ones, means those factors are far from unity.</p>

<p>I have never tried this - but judging by the many posts on the web, it indeed works. When proocessing your RAW images later in NX or ACR, one of course needs to apply WB tothe image - and my current understanding is that the UniWB and exposure control affected in camera now results in the minimum possible factors for the red and blue data - maybe someone who uses this method can confirm this (or prove me wrong, I am here to learn).</p>

<p>Rene', there is no yellow channel - the three channels are red, green, and blue - problems with overexposing yellow are due to blowing the red channel.</p>

<p>Lil UNIWB is applicable to different brands of camers so you might be better off to look in fora that deal with post processing. There have been many posts here in the Nikon forum back in the days of the D70 and D200^^.</p>

<p>I suggest you search in the DPreview fora where I see a new thread about WB and correct exposure almost every day. For a start search for contributions e.g. by <a href="http://forums.dpreview.com/forums/postersprofile.asp?forum=1021&message=30944593">Iliah Borg</a> . Not easy reading but worth it to try and understand what happens.</p>

<p>The histogram you see in typical cameras is derived AFTER all settings for in camera processing are applied, even if you shoot raw. These settings are applied to present a "pretty" image to the photographer. The histogram that the camera displays after the shot is therefore often very misleading if you want to judge best exposure. Exposure is typically way off if you have one dominating color, especially a color that matches one of the sensor colors.</p>

<p>When shooting RAW in the attempt to make best use of the sensor and record as much information from the scene as possible it is advisable to set the gain of the three color sensor channels to a factor of 1 (<em>unity</em> ) and turn off all settings that process the raw image and change the display of the histogram. This way the histogram will represent the recorded channels in a more precise way.</p>

<p>If you look at images that were recorded that way these look just terrible. The colors and WB are way off. But that is "working as intended". You have to correct the image for final output by setting the best ratio of the color channels (plus a few other things) in post processing. This is what is normally done in camera with auto-WB. Now you have to do this in post processing. The advantage is potentially better detail, better color representation and less noise. However, all other factors involved must be well controlled in order to get the benefit.</p>

<p>Dieter..... NO YELLOW CHANNEL? Uhmmm! Well, there should be one! :)That's exactly how rumors start! hehe<br>

I am sorry, I was ready to leave my office and I replied in a rush! I meant that the color red is the hardest to expose right and the second one is the yellow. At least those are the 2 colors I have problems.<br>

I still don't know what UNIWB is!?</p>

<p>LIL.... please take a look at this thread! You'll like it coz there are a few birds! :)<br>

http://www.birdphotographers.net/forums/showthread.php?t=29869</p>

<p><strong>Rene'</strong> - yellow channel - - <strong>I want one.....</strong> LOL LOL LOL That's too great. :-) Thank you for giving me a great morning laugh. :-) You've put a huge smile on my face & I think you for that. In the middle of a bad morning headache I'm smiling. Now how can my day be anything thing but great. Thank you Rene'.<br>

<strong>Hi Dieter</strong> - thank you for putting this in such great terms. I know that the histogram is a jpg idea of the the NEF file. So adding this file to the camera should allow me to see the actual RAW file's histogram - but all settings to the camera would have to be 0/nothing. That'll add a few steps in editing.... It does sound like it does have some advantages... I will have to think a tad more about this - - true & correct histogram of RAW file - - or a pretty for me to show & look at on the LCD screen & less steps in post.... Thanks again Dieter.<br>

<strong>Hi Walter</strong> - I will search some forums, but I may well stay away from DPR on this subject. My thread there is fairly current & just guess who it was who has been talking down to me - - even though at one point I left a cordial message to him which made him change his tone a tad. It was rather obvious he was talking down to me. The attitude that if I didn't know this I should not have bought the camera was rather clear. Thank you Walter for such clear terms. In between you & Dieter I feel I understand things a lot better. I don't have many settings in camera - Sharpening at 5 - NL - WB to Sunshine (as that's what I do most...)I'm just wondering if it's not easier for me to just get my settings right in camera.... What if I install this & absolutely go nuts with what I see on the screen.... Can I undo the "damage"? Something I need to know.... Thanks Walter.<br>

<strong>Hi Rene'</strong> - you're back. :-D You're right - Red & then yellow are the most difficult colors. I've figured UniWB now - but I'm still debating if I really want to install that on my D700. To me it just sounds like I'm better off doing negative EV until I have a better idea where I have my Red Channel. Since in the long run the final product will be a jpg - - having an idea where that histogram is from the beginning seems a better idea.... Thank you for finding me the thread. I will check it out after I've fed the horses breakfast. Which is what I'm off to do now. Thanks Rene'<br>

Now I just played with the camera through the window out in the yard. If I have it set to -0.7 EV I get a pretty much correct histogram within the lines. Perfectly - then I have to go to -1.0 EV<br>

Now I know this is a jpg rendition of what I shot - but doesn't that sound a tad too much?<br>

Anyhow.... my learning goes on.... Thank you all for helping. Got to love the civilized photo.net - - - - on an other specific forum I would have been crucified...<br>

Thanks Lil :-)</p>

 

<p>Lil, I have actually never gone through the process - read about it more than a year ago - read that it wasn't that easy to get the setting into a D200 and some assorted software problems - so I forgot all about it. Looks like now the process just creates a custom WB setting - and you can choose not to use it or delete it if you don't like it. One option might be to adjust your exposure with the UniWB setting, and then switch to whatever WB setting you want and shoot with the exposure determined from the UniWB set manually. Not sure how practical that is.</p>

<p>I know that you use Nikon's software for processing - which uses the in-camera JPEG parameters - but one thing I definitely would not do is sharpening already in the RAW conversion. Sharpening should be done IMO as the last step in the workflow and specific for the intended use (web display or print).</p>

<p>As to your question re: negative exposure compensation - I need to do this all the time on my D200 and D300. Remember, since the histogram is off the JPEG and you shoot RAW, there is some leeway for recovery later on - that is, the in-camera histogram doesn't need to be fully contained. I think I have developed a pretty good sense from looking at the in-camera histogram and determine how much overexposure is tolerable.<br>

You are right that eventially the image will end up as a JPEG again - but I process in ProPhotoRGB color space and at 16 bit until I am ready to convert to JPEG and sRGB dolor space for web display. Sure, then sometimes the carefully guarded histogram shows some overexposure again but to no apparent detriment for the image itself.</p>

<p>Hi Dieter -<br>

I remember reading about the D200 which prompted me to test my D200 & go insane about the red channel. Roman Johnston took it upon himself to request one of my NEFs & simply told me he saw no problem. But it also prompted me to really mess with the D200 & get it where I like it. Now it's there - set - & not ever changed.<br>

The D300 just turned out to be so much easier. I have no issues with my D300 - I finally went to a -0.3 EV only because it gives me "warmer" colors. I only compensate or spot meter when I deal with a white bird or BIFs.<br>

Then I get the D700 & every shot I take the red channel is off the board. So I start negative EV - but then I start realizing that I have to do as much as -0.7 EV on a regular scene & get concerned.<br>

Then all this starts....<br>

Dieter, problem with NX & NX2 is that they really want everything done in their steps. The minute you do something in a different order everything you have done goes away & has to be re-applied. Easier to just do it from the beginning...<br>

I so wish they didn't have it set up that way - but unfortunately they do.<br>

So - to be honest, based upon what you, Walter, Rene' & I have been discussing - - I really see no reason to do this. I'd rather go back to shooting as I'm used to & just accept that I have to negative EV a lot.<br>

That way I don't have to worry & do things in ways I don't want to.<br>

Thanks for helping me sort this out.<br>

Lil :-)</p>

<p>Lil, out of curiosity - what are the changes you made to the D200?<br>

Back to UniWB. Using it won't change the fact that you have to apply negative exposure compensation as I understand it - just giving you a better idea on how much.<br>

It took me a while to get things right for shooting white birds - initial results with matrix metering where disappointing and I switched to spot metering - until the experts at BPN called me on it a few times - so I switched back to matrix and learned when to apply exposure compensation - seems to work most of the time now. But especially in bird photography, I can hardly take a shot with either the D200 or D300 without the need to compensate - but I never find the need to use spot metering. It appears that the D300 needs less help than the D200 though.<br>

I know I could set some permanent exposure correction in my cameras but decided against it - too hard to teach an old dog new tricks...<br>

I occasionally used Capture NX - but it is just painfully slow on my computer - can't say I explored it much at all.</p>

 

<p>I haden't heard about UniWB, but I just set my usual settings, look at the histogram if possible and expose accordingly. I can pretty much tell from the subject and from how the histogram looks how much leeway I have before I blow some channel in RAW. This probably upsets those who want a more numeric method of optimal exposure, but in practice many situations require quick reactions and at that point experience is needed to get things right.<br>

I also want to minimize the amount of tweaking that I need to do in post and especially like to get the WB right when shooting. It saves time and the WB adjustment in Capture NX is rock bottom in terms of ease of adjustment. Should probably get Lightroom, but haven't felt I need it enough (Adobe charges exorbitant prices here).<br>

One thing to keep in mind though is that with all this advanced electronics, it's sometimes very useful to go to manual exposure after having determined the correct exposure. This happens when the subject will be fairly constantly lit but the camera meter easily fooled. For flower photography, taking an incident reading and going with manual is often quite easy.</p>

<p><strong>Dieter </strong> - For one thing - Always have a -0.3 EV standard in Matrix metering. Secondly to always double check for highlights. I customized it to some settings a friend on DPR gave me. It's very colorful & happy now. ;-)<br>

If all it's going to do is still have me EV compensate etc - - why bother. I've noticed I'm getting better at it so I'm going to stay as is. I'll just work harder at getting my settings down.<br>

As for metering - I prefer to Matrix meter - unless I shooting at a bird in flight (BIF) then I like to spot meter because the sky just overpowers the bird.<br>

I'm on FM a lot & there there's a lot of Canon shooters - the fact that I'm not always in Manual metering really gets some of them. I guess they don't trust their metering in the camera.... I find I shoot very differently then they do. I had a young man - probably in his early twenties question how I shot the other day. When I told him I needed shutter speed with the Sigmonster he said he was sure I could shoot this way or that. He has absolutely no idea what he's talking about. I don't have VR & need fast shutter speeds. Granted I've shot with longer - but you just don't shoot at 1/30s with a Sigmonster..... Especially not birds. He had to prove he was right & put the camera (yes I let him test it) in ISO 1000 & 800. I went back to 2500 because that's what I chose to do.<br>

Thanks <strong>Oscar</strong> for joining the thread with your comment. I'm sure you're far more advanced than I am - but I do agree.<br>

Lil :-)</p>

<p>Lil... Well, a laugh is always good! In that thread they explained to me how to worked with the red channel. It seems fairly easy. If you look at the thread make sure you read all the way to the bottom! have a good day!</p>

<p>"My thread there is fairly current & just guess who it was who has been talking down to me - - even though at one point I left a cordial message to him which made him change his tone a tad."</p>

<p>Haha Lil - I know there are many experts in this field who probably do not just sound arrogant but probably are arrogant. (Very funny typing error just happened: I typed aggrogant - perhaps that describes it even better ^^.)</p>

<p>Nevertheless I follow several threads regarding the topic. As it happens sometimes arrogant people are no real experts but sometimes arrogant people are experts and it is still worth while to read the posts and just ignore the style. Often these posts contain real information, sometimes just calling for attention and applause :-P</p>

<p>Perhaps my comment that it would be "Not easy reading but worth it to try and understand what happens.." should have been more direct in the way: no easy reading because of the subject and because of some of the people involved^^.</p>

<p> </p>

<p><strong>Hi Rene'</strong> -<br>

Well I've also been Internet chatting with Jussi (I think he's here as well....) and he's given me some thoughts. So I may try it & see how I like working with it as Jussi says the colors actually come out better for some reason... Ah well.... ;-)</p>

<p><strong>Hi Walter</strong> -<br>

I think the man in question at one point was banned off DPR believe it or not. There was a rumor going around... He is a pro & I'm sure he knows a lot. But I still don't like being talked down to. There are pros & there are pros. I like getting advice from a pro - - but I prefer they treat me at least like a human being. ;-)<br>

Thanks for helping me sort this out.<br>

Lil :-)</p>

<p>I set my cameras to produce the best diagnostic previews, so the images can look blown, even if they are not. Usually, if I don't get some 'blinkies,' I'm underexposing.<br>

My D200 was factory set to underexpose by 1/3-2/3 stop, depending on the metering pattern. I corrected this with the meter calibration function, but that makes exposures hotter. It hasn't been a problem for me -- I always start with no compensation and that usually works fine. If scene contrast is high, negative compensation may be needed.<br>

I haven't tested or adjusted the metering offset on my new D700 yet. In 14-bit raw, I'm surprised by how much the highlights can be recovered. I shoot with high-contrast stage lighting, so I've learned to ride the exposure compensation based on scene contrast (often -1EV to -2EV). Sometimes I really 'blow' an image, usually due to inattention -- stage lighting can be variable and dynamic. Compared to my D200, the D700 seems much more forgiving of overexposure and subsequent recovery.<br>

My experience and opinions are based on Lightroom 2.2.</p>

<p>In a sense, that was an extremely interesting read. Thank you Lil.<br>

If anybody is interested, here is the thread on DPR Lil mentioned<br>

http://forums.dpreview.com/forums/read.asp?forum=1021&message=30847386</p>