I took a dragonfly out of a spiderweb and put it in my freezer. I finally got around to photographing it tonight. I figured that I'd go simple to almost guarantee good results and I'm not happy with the results.
I used my Sony Alpha 200 with a Tamron 90 mm 1:1 Macro lens set up on a tripod. I then mounted the dragonfly on a needle through a cork and placed it on the stage of a microscope and used the mechanical stage to move the dragonfly. I used a new standalone lamp with ping pong ball diffusion. The camera was set at F-8, 1 second, ISO 100. I used the remote to trigger the camera with no flash. The 30 images were stacked with PS CS5. I believe that for this image I also had the Raynox on there too. The image was cropped slightly to remove some empty space.
I'm not happy with the color, the lighting, the sharpness . . . nothing.
From the description above and the image below, what advice would you all offer to help me improve?
I usually have dark dust streaks on my images. This one has bright streaks. I don't think they're digital noise. It was shot at ISO 100. What causes those?
Mike wrote:I usually have dark dust streaks on my images. This one has bright streaks. I don't think they're digital noise. It was shot at ISO 100. What causes those?
Could be 'warm pixels' on the sensor. These usually begin to appear with exposures >1 second. How old is your Sony Alpha?
What was the WB setting on the camera when taking the images for this stack, and can you provide more details regarding the "standalone lamp" and the ping pong ball diffuser arrangement? Also, what was the exposure setting on the camera; was it set to Auto, Manual or other?
Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"
The camera was set to manual because I was trying to create a stack by adjusting the lens's focus ring manually but it didn't work well. So, instead, I decided to keep the camera still and move the subject.
White balance is a setting that I have never played with. It was set on "Shade +/- 0". I just switched it to AWB.
By exposure setting, are you speaking of the exposure time? I set that manually to something that gave an image with a reasonable amount of light. I'll put one of the untouched images below. They are very orange, but the light is very yellow and the ping pong ball seems to give it an orange tint. Plus, the dragonfly is brownish orange. Strange thing is, I don't see any bright spots in that image. Maybe they're too small until it's stacked?
You should definitely switch to using custom white balance. It's odd for a camera to go this far orange when faced with incandescent light, but very common for there to be some color cast. To use custom white balance, just replace your subject with a white card, or stick a white card in front of it, and go through the balancing procedure described in your camera's manual. Typically the procedure is something like take one picture, then push a couple of buttons to tell your camera "this should be white/gray/neutral", then maybe push a couple more buttons to tell your camera to actually use the custom white balance that it just determined. (One of the peculiarities of many Canon cameras is that they will happily let you define a new custom white balance but then continue using whatever you had set before. I have no idea about your Sony.)
If you have that lamp on a dimmer, be aware that the color will shift when you change brightness.
When you're short of light, a pingpong diffuser may not be the best choice because they tend to absorb quite a bit of light. Try using one or two layers of Kleenex tissue instead. That's almost as good a diffuser, and won't absorb nearly as much light. You may be able to remove the diffuser entirely, if you're using a frosted bulb. Adding a white card or some aluminum foil on the shadow side to provide reflected fill light will give you a little more light and brighten up the shadows too.
The bright streaks are undoubtedly warm pixels. Just like dust spots turn into dust trails, the stacking process turns warm pixels into bright lines that are a lot more obvious. Unlike dust spots, warm pixels change intensity depending on ambient temperature and how much your camera has been used over the last few minutes. They also vary with exposure time, longer is brighter, so adding illumination to get a shorter exposure reduces the problem. Sometimes warm pixels get worse over time, but frequently they were always present and just get noticed more as you run into situations where they appear.
I think you'll be a lot happier when you get the color right, even if everything else remains unchanged.
Regarding sharpness, I suggest taking a look at your original images to evaluate that. CS5's stacking method is noted for having trouble with complex geometry like you have here.
I think you're both right about the warm pixels. I had the remote in one hand and the knob in the other and was just going turn/click, turn/click, turn/click one after the other. I'll make some adjustments on the camera and slow down the clicking.
I see a dilemma, though. I want longer exposures to reduce vibrations but shorter exposures to stop warm pixels. I used my grant to order an external flash which would help solve both problems, but it wouldn't fit the hot shoe on my camera even though it was advertised "For Sony DSLRs." So, I have to send it back and get one that does fit.
The softness might be caused by diffraction. If you used a Raynox that is( and also depending on which Raynox lens it was), in which case it might be safer to use F5.6
"White balance is a setting that I have never played with. It was set on "Shade +/- 0". I just switched it to AWB."
Don't know the Alpha and I have never used a camera with "picture settings" but if your camera was set on "Shade", shade light has a blue cast from skylight so the camera would add more red or orange to counteract this? If you are using artificial light the colour temperature needs setting to that.
Our eyes correct for it, and normally your cameras auto light balance will try to correct, but if you have picture settings on your camera and switch to the "Shade" setting it will try and add more red to compensate for the expected blue light.
Set you colour balance to that of the artificial light you are using and see if that corrects the problem?
Thanks Dave and Realjax. The Sony Alpha doesn't really have picture settings like "sports" or "sunset." But it has about 5 white balance settings and the default happened to be that shade setting. The "+/-" on the shade setting was adjustable, but since I'd never really looked at that feature before, I haven't experimented with it. I've just been depending on photoshop to fix all of my color issues. But I should learn more about white balance. Thanks for the links.
Sorry, I overlooked that the camera was set on "shade" when you took those photos. That completely explains the strong orange cast.
As DaveW suggests, using a predefined setting that matches the illumination type would be a big improvement.
To get spot on, though, you'll need to use custom balance. That's because the color of incandescent light varies quite a bit depending on bulb type, wattage, and line voltage.
Auto white balance is tempting, but it often gets misled by colors in the subject. Depending on composition, there may be no way for the camera to tell the difference between an orangish subject lit by white light and a white subject lit by orangish light. Custom white balance is how you tell the camera which situation it's dealing with.
I looked at a review of the Alpha A200 but could not at first see the "Shade" setting mentioned anywhere. I found these though:-
"White balance. Of particular interest is the Sony A200's White Balance menu, which offers a very simple approach to a complicated subject. It's actually identical to the A100's White Balance system, but the interface is slightly easier now. Just use the up and down arrows to pick a white balance method, and use the left and right arrows to adjust the color bias of that particular setting. If you've chosen Tungsten, for example, but your light source is just a little off from the norm, hit the left arrow button to make the image a little bluer, or to the right to make it a little more yellow or orange.
If you know a little more about color balance, you can switch to Kelvin mode and dial in the right color temperature, and add green and magenta filters. You can use the A200 as a gauge by moving to Custom mode, which will ask you to take a picture of a white or neutral object and dial in the correct temperature and filter setting to match. There are no pretty graphics to accompany the adjustment, as is more common on other cameras, but it's pretty straightforward in practice."
"Sony’s implemented a new Function button on the rear. Labelled Fn, this fires-up a new user interface, splitting the screen into six sections which give you easy access to the flash, metering and AF modes, AF area, White Balance and Dynamic Range Optimiser (DRO) options. These give you relatively quick and easy access to the common settings."
However I later found the "Shade" colour temperature setting on the following link. You want to be set normally on AWB (Automatic White Balance) and then if that is far off the correct colour use custom white balance as in the video. See white balance settings in the link below:-
If you are set to "Shade" as said before, plus shooting with tungsten lighting you will get a pronounced orange cast on the image.
Edit: Rik just posted before me, but above I wondered why your camera had been set on shade in the first place? For normal photography use AWB, but as Rik says for precise control in a studio setting use Custom White Balance.
Here's a little experiment I did tonight thanks to your feedback. I put a little piece of cardboard under the trinoc scope. I have three images below. The top is with AWB. The second is with a custom WB based upon a white sheet of paper. The bottom one is with a custom WB based upon the white circle at the base of the microscope. Pretty clear differences. But I cannot tell which one is best. I think that the little white thing in the middle is the whitest on AWB. I was using a fluorescent ring light. The custom WB for the paper was set at 5700K M8. The custom WB for the white circle was 5900K M9.
Take a look at the histograms in Photoshop. Set it to Expanded View, showing Colors so that you can see the distributions of R, G, B separately.
In the first one, AWB, the three channels line up very well, with Blue just a bit low and Green just a bit high. That's typical of what AWB is liable to do with any uniform pastel color, because it assumes that the scene is neutral on average so it determines a white balance that makes the image be neutral on average also.
In the second one, custom white balance off the white paper, the Blue peak is significantly lower than Red, with Red slightly lower than Green. The custom white balance would make the white paper neutral -- all three peaks lined up -- so what the picture shows is that the cardboard has a yellowish cast compared to the paper.
In the third one, custom white balance off the white circle at the base of the microscope, all three peaks are well separated, with blue being the lowest and red being the highest. This means that the cardboard has a distinctly brownish cast compared to the white circle at the base of the microscope.
To make custom white balance work the best, you will need to find a reference material that really is neutral. I always find it surprising how much variation there is between various colors that are all "white". I usually use one of my favorite index cards, and I guess I've gravitated to that because it's easy to work with and I know those particular cards give pretty good color. You can also buy white or gray cards that are specifically made for this purpose, but they tend to be pretty expensive and I've never taken the plunge. High quality printer papers tend to have quite a bit of blue in them due to added brighteners, so I shy away from those.
Not sure if that is the same colour balance as ordinary fluorescent tubes, but if so it is always a problem correcting fluorescent lighting colour balance since they are not full spectrum light sources. However the newer fluorescent lights designed for photography seem to be different. I wonder therefore what the colour temperature of these are and whether they need the fluorescent setting on the camera anymore? No doubt the instructions that came with your ring light should tell you the colour temperature or camera setting required?
Rik, I appreciate the information about histograms, but I know so little about it that I don't even know if Red, Green, and Blue are supposed to overlap or if they are supposed to be separated. I've seen the histograms on my camera, but have never used them in PhotoShop before. I'm sure I can find them, I'm not so sure that I can figure out what they mean.
Dave, my camera is capable of shooting in Raw, and I did it once, but I really didn't see any advantages to it. I've been trying to improve the sharpness of my pictures and Raw didn't help me with that so I went back to shooting in jpeg. If I get the sharpness solved, then I might start shooting in Raw and worry about color more.
Thanks for all the advice. I'm more comfortable with white balance now.
"How many of you shoot RAW all of the time? Please raise your hands?
OK, I see about 15% of your hands raised. Those of you with your hands at your sides….in my opinion you are incorrect. I believe that 100% of you are shooting RAW images, 100% of the time. OK. I am playing with you, but hear me out. When your camera is set to RAW, the picture is taken, and a RAW file is produced, and then it is written to the media card. However, guess what? When your camera is set to JPEG, the picture is taken, and a RAW file is produced (in memory), and then the camera goes ahead and develops the RAW file into a JPEG and discards the RAW data."
Therefore Scitch if your camera is actually shooting RAW before it converts it into a JPEG, how can shooting JPEG potentially produce a better image than using the RAW data directly in say Photoshop? You can obviously potentially produce a better image from RAW than you can from in camera JPEG's, but not vice versa since the RAW data is the starting point in both instances.
If you had shot your above image in RAW you would have had more chance of rescuing it than using in camera JPEG's. See the link below. If that had been an in camera JPEG it would be beyond saving:-
Obviously our images are not usually that bad but the principle is the same, a RAW image can always be manipulated and improved more than an in camera JPEG since you have more of the original information to work on.
