Problem when shooting flowers


Status
Not open for further replies.
Yeah, but there are a couple of problems here.

First, contrary to the suggestion above, you can't have a histogram that "seems wrong". It's a visual representation of the data in the image. Either it's saturated, and you will see a spike/hump/peak/collection pegged at the right edge of the histogram, or it's not, and you won't see that. If the histogram is "wrong", then the machine (computer, software, camera, whatever) you're using to compute & show the histogram is wrong, and I'm pretty sure lightroom isn't broken. Is the implication that the histogram is being shown incorrectly in Nikon cameras?

Second, you can't use an image processing program to take an 8-bit jpg image that is "saturated" in and "unsaturate" it. The attached picture above, reduced in "exposure", isn't a helpful illustration.

An exercise to "reduce the exposure" on an 8-bit jpg that has areas of saturation (i.e. pixels in which one or more of the color channels has the value of 255), no matter the software you're using, will have no effect on the saturated areas. True, some of the pixels "near" the saturated pixels may not be totally saturated (i.e. slightly below 255), and when you "reduce the exposure" of the image those pixels will darken and you may think this means you're eliminating the saturation. But the saturated pixels, which will invariably be areas (collections of pixels), will all hold exactly the same value.

In other words, if you have an area of red saturation, that means you have an area with a group of pixels that have a value of 255 (in an 8-bit image) in the red channel, meaning you've lost all "detail" in that areas since no difference exists between one pixel and another. If you reduce the brightness/exposure of that image in 8-bit space, then that group of pixels will still have exactly the same value in the red channel, even though it may no longer be 255. Even though the saturated area of pixels may no longer be bright red, you'll still have a splotch with no detail (but with some other color).

By the way, the above applies no matter if you process the image is 16-bit space or 8-bit space, presuming you started with a jpg (8-bit by defnition). In other rods, you can't just load up an 8-bit jpg in photoshop, switch to 16-big mode, and expect to overcome this issue. You CAN start with a high-bit image (like RAW, which is generally 12-bit, or a 16-bit image) that appears saturated in 8-bit space, but which has not fully saturated in 12/16-bit space, and lower the exposure to eliminate the saturation at 8-bits. I'm at a loss to explain, but I hope the concept is clear.

The histogram included above clearly shows that the TS's issue is NOT saturation. He must be getting thrown off by either depth of field, proper focusing, camera shake, or something else. But it's clearly not saturated in the red channel, else the histogram would show the red channel shoved all the way to the right.

TS, try it yourself - increase the brightness/exposure until you SEE the histogram shove up to the right side. That's what saturation looks like. If that doesn't look like what you're worried about, then consider other factors.

Thanks,
Eric
 

hmm.. i understand what u guys are getting at.. will go try it out .. another question that might be related or unrelated to this topic.

is the UV filter necessary for the 50mm lens ? if im outdoor i would think so , but if im indoors and since this lens is supposed to be good in low light conditions, i would not need it ?

I recommend you keep the UV filter on the lens at all times. I think most people agree that UV filters have NO PRACTICAL EFFECT on image quality (unless you're shooting at high altitude, perhaps). Search the web and you will find many many arguments (about the practical effect of UV filters, or lack thereof). But, since they also have NO LIGHT LOSS, there is no reason not to keep this (presumably cheap) piece of glass screwed to the front of your lens. It's a hell of a lot cheaper to go buy a new UV filter if it hits something and breaks, than it is to have the front element on your lens replaced. It can also prevent dust from forming on the front element, and if it gets dusty/dirty, I'm a lot more comfortable wiping/blowing the filter than the front element.

Think of the UV filter as a cheap clear protector for your lens.
 

is the UV filter necessary for the 50mm lens ? if im outdoor i would think so , but if im indoors and since this lens is supposed to be good in low light conditions, i would not need it ?

Some people see it as cheapest protector. But be aware that a) the lens has been produced to work best without filter; and b) filters (especially the cheaper ones) will induce flare and ghosting under certain conditions. No point having a protected lens but the image is spoiled, right? In such circumstances a lens hood is much better. It helps to protect the lens physically (it's elastic plastic that can catch some bumps) and it shields the lens from stray light.
Something to read:
http://toothwalker.org/optics/flare.html
http://toothwalker.org/optics/lenshood.html
 

I just pointed out the picture is overexposed : because it is.

overexposure-of-red-channel-ffbe6b.jpg




If you don't want the picture to be overexposed, you have to take care of the red color channel when picturing red subjects :

red-channel-if-co...-snapping-ffbf75.jpg





In such case, i'm not fan of post-production corrections. The very simple solution is to have in mind you can easily overexpose red subjects even if your histogram seems normal and well-centered, and to expose carefully on the field for red color channel before snapping.
 

Last edited:
Please do note that whatever post processing you are doing to that hibiscus flower, it ain't what is suppose to look like.
 

I just pointed out the picture is overexposed : because it is.

overexposure-of-red-channel-ffbe6b.jpg




If you don't want the picture to be overexposed, you have to take care of the red color channel when picturing red subjects :

red-channel-if-co...-snapping-ffbf75.jpg





In such case, i'm not fan of post-production corrections. The very simple solution is to have in mind you can easily overexpose red subjects even if your histogram seems normal and well-centered, and to expose carefully on the field for red color channel before snapping.


Yeah, except that the histogram the TS posted does NOT show the red channel saturated. Maybe he's saturating it in his post. But the adjustment you've posted results in an image that has shifted from having saturated red in the flower, to saturated, darker red (magenta?) in the flower. You can't fix a saturated 8-bit image in post.

Maybe we're saying the same thing. As you very correctly point out, you should, particularly with flowers, expose carefully in the field for the predominant color/tone.
 

umm.. i think some of u may have misunderstood. nothing was done to the photo after snapping, i merely used lightroom to obtain the histogram so i could show it to you guys.
thanks for the answers on the filters too.

umm but could anyone further explain this: "The very simple solution is to have in mind you can easily overexpose red subjects even if your histogram seems normal and well-centered, and to expose carefully on the field for red color channel before snapping." ?
 

umm.. i think some of u may have misunderstood. nothing was done to the photo after snapping, i merely used lightroom to obtain the histogram so i could show it to you guys.
thanks for the answers on the filters too.

umm but could anyone further explain this: "The very simple solution is to have in mind you can easily overexpose red subjects even if your histogram seems normal and well-centered, and to expose carefully on the field for red color channel before snapping." ?

Well this of course presume that the issue is in fact saturation, and based on the histogram you showed earlier it doesn't seem like it is.

But generally speaking, what the other poster means is that when shooting flowers you should pay particularly close attention to potential saturation in the dominant color of the flower. If you're red flowers, watch for saturation in the red channel. If you're shooting blue flowers, watch for saturation in the blue channel.

A histogram can be represented in a couple of ways. Some cameras will shoe only a "white" histogram, which only shows the overall tone of the image, combining the rgb channels into one visual representation. Sometime in this mode, you can fail to see that one of the channels is close to saturation. If your camera show all three channels separately, which it probably does, then you can judge the exposure of each channel. Take a look at the lightroom histogram you posted earlier. Your camera's display is probably very similar. You can see that the individual color channels are all shown, meaning you can tell the exporue of red, green, and blue by looking at the histogram curves of those colors. If your shooting a red flower, just make sure the red portion of the histogram isn't pushed up against the right side (saturated). The other channels will be less imprtant relative to the dominant color in the flower.

In practice, it's more like "snap, then check" in these days of digital cameras. If you see saturation, reduce exposure and try again.

You may also find that of you properly expose for the flower's dominant color, the overall picture is rather dark. In that case, just fix that in post if necessary. Lightroom works great, and it's easy to get the hang of.

Also, shoot RAW. You can recover overexposed images often if you have the raw file.

Take Care,
Eric
 

Yeah, except that the histogram the TS posted does NOT show the red channel saturated.


I beg your pardon as i insist. The histogram the TS posted seemed normal but it is not true. I have been trapped myself many times before i understand why. RGB histogram appearence captured in Lightroom fools us, just as white histogram can do.

If you take TS's picture or my red poppy picture, and open them in Lightroom : everything may look OK. But just try, open them in other software : the red color channel is no doubt against the right side, it's to say satured. Photo are overexposed. It's very tricky, indeed.

histogram-in-lr2-1000518.jpg


histogram-in-dpp-10007b3.jpg


I won't insist any more in this thread : just want to warn buddies and do my possible to help .

Wish all of you a pleasant further discussion,
YP.
 

Last edited:
I'm your youngest sister so i beg your pardon to insist. I don't want to have reason at any price. I just want to share my very small experience. The histogram the TS posted seemed normal but it is not true. I have been trapped myself many times before i understand why. RGB histogram appearence captured in Lightroom fools us, just as white histogram can do.
If you take TS's picture or my red poppy picture, and open them in Lightroom : everything may look OK. But just try, open them in other software : the red color channel is no doubt against the right side, it's to say satured. Photo are overexposed. It's very tricky, indeed.

Have you discussed your question with the LR forum? It's not a direct feedback channel to Adobe but some developer are reading there as well and there are plenty of 'gurus' around. If the histogram is showing different information than another program it would be good to know why. Sometimes people use a different scale which can make a diagram look different although the input is the same.
Another idea that came up my mind: colour correction filters. If we use warming filters to reduce the blue cast of an image then the opposite would also be possible.
Using Google I found some other discussions on the same topic. From the example pictures not only in this thread it seems that Hibiscus is a very tricky flower to photograph :bsmilie:
http://photo.net/nature-photography-forum/00TGYd
http://photography-on-the.net/forum/showthread.php?p=7671520
http://www.pentaxforums.com/forums/.../42876-how-reduce-red-channel-saturation.html
http://forums.dpreview.com/forums/read.asp?forum=1034&message=18988029

Hope that helps.
 

Status
Not open for further replies.