Lack of sharpness

scitch · Post by **scitch** » Fri Jul 16, 2010 6:34 pm

For the life of me, I cannot get any sharpness out of my photos. This stack was done on the Konus Crystal-Pro trinoc through the photo tube with the adapter and 10X widefield eyepiece to a Sony a200. I adjusted the photo tube so that it is parfocal with the stereo eyepieces, but then the camera is not parfocal (nor parmagnified if that's a real word). I also used the 10 second timer and 8-second exposure time to reduce vibration. The camera is pretty solidly connected, so I'm not sure that vibration is the issue. This was lit from all four sides with the microscope's halogen light diffused, two flashlights, and a reflector.

Other suggestions? It doesn't appear to be a stacking or a post-processing issue, the originals aren't very sharp either. All I did in photoshop is auto-levels and resize. Occasionally, I also sharpen, but it usually makes it worse. The sensor is still dusty, and I'm awaiting the arrival of some swabs to fix that.

Another question: Does it do any good to turn the camera's vibration reduction on? I think that the a200 has sensor-compensation so it seems like it should help.

Here's one at lower magnification, but cropped a little. Same disappointing results.

rjlittlefield · Post by **rjlittlefield** » Fri Jul 16, 2010 8:16 pm

You have my sympathy -- fuzzy pictures with no obvious cause are always a tough problem.

Let's go back to basics. A few questions:

1. When you look by eye through the normal eyepieces, do you see much more detail than these photos show?

2. How much more area do you see through the normal eyepieces than appears in these photos?

3. Have you tried photographing through the normal eyepieces, using a normal lens on the camera? If so, how does the detail captured in those photos compare to what you see by eye?

4. Have you tried photographing by flash, rather than continuous illumination? Usually you can get enough light from the camera's built-in flash by jury-rigging an aluminum foil tube from the flash down to the specimen.

Regarding the camera's vibration reduction, I would expect better results with that turned OFF. The reason is that (I expect) the sensor shift is based on accelerometers built into the camera body, and when the camera is looking through a microscope, movements of the camera body do not affect the image in the same way they would looking through an ordinary lens. But it's an easy test -- try it both ways and see what happens.

--Rik

scitch · Post by **scitch** » Fri Jul 16, 2010 9:12 pm

1) Yes, the insects look great through the eyepieces, even through the photo tube.

2) At 45X, I can see about 4.2 mm through the stereo eyepieces. Through the phototube with the 10X eyepiece, I can see about 3.8 mm. Through the camera, I can see about 1.8 mm.

3) Can you explain how I'd do this? Do you mean, just put the camera lens back on and hold the camera up to the eyepiece? I have not tried it, but I could if that's what you were describing.

4) I couldn't figure out how to get the light from the flash to the subject. It's a good 14 inches away with the adapter tube and all. I had never heard of the foil tube, though. I can give that a shot. Is there a picture of one anywhere that you know of? I have an external flash on the way and that should help.

5) I will try turning the "Steady Shot" off and doing some experimental shots.

Thanks for the recommendations, I'll let you know how it goes.

Mike

rjlittlefield · Post by **rjlittlefield** » Fri Jul 16, 2010 10:19 pm

scitch wrote:3) Can you explain how I'd do this? Do you mean, just put the camera lens back on and hold the camera up to the eyepiece? I have not tried it, but I could if that's what you were describing.

Exactly. Officially this is called the "afocal" method. I guess the name is because the light coming out of the eyepiece would never focus anywhere by itself. See HERE for some further discussion of how this setup works.

4) I couldn't figure out how to get the light from the flash to the subject. It's a good 14 inches away with the adapter tube and all. I had never heard of the foil tube, though. I can give that a shot. Is there a picture of one anywhere that you know of?

See HERE for an example of the concept. The last one that I made was done by wrapping a piece of aluminum foil around a nominal 1x2" stick of wood (actually 3/4" x 1-1/2"). One end of the tube gets taped against the camera's flash, the other end of the tube gets flared out so the light shines on the subject. I'd show you a picture, but it's about 1200 miles away right now.

--Rik

scitch · Post by **scitch** » Sat Jul 17, 2010 12:59 am

I'll give it a shot. My daughter already did it with a cheap Discovery Channel handheld digital camera through the eyepiece. She's forward thinking.

So, I've been experimenting with a lot of different methods to see which one works for me. I've tried microscopes (both stereo and compound). I've tried objectives at the end of a tube. I've tried objectives at the end of a bellows. I've tried diopters on the camera lens. Tonight, for the first time, I tried a camera lens reversed on the end of a bellows. Below are the results. Not horrible, but not spectacular either. I can tell that I skipped some areas and need to have finer control of the movement. For lighting, I used a sheet of paper curved over the camera running from the flash to the end of the bellows. I could do the same with foil on this setup, but I thought paper would be gentler. And I turned off the camera's image stabilization. This is a stack of about 30 ISO 100, 1/3 second images. Not the worst image I've ever taken, but could still use some work.

rjlittlefield · Post by **rjlittlefield** » Sat Jul 17, 2010 10:28 am

This looks pretty good. The next thing to do, if you haven't yet, is to play with different aperture settings until you find the one that gives the sharpest image. Probably it will be 2-3 stops down from wide open.

I worked for years with reversed camera lens on bellows. It's not a bad approach at all. Images won't be as sharp as what you can get with a high end macro lens, and much less sharp than what you can get with a well behaving microscope objective, but still definitely respectable. See HERE, third panel.

--Rik

scitch · Post by **scitch** » Sat Jul 17, 2010 11:35 am

I had read somewhere that the technique is to set the aperature in the camera and then remove the lens while the power is still on and it will lock that aperature in. But the stock lens that came with the a200 doesn't work that way. It is a mechanical sliding lever in the back of the lens that is spring loaded and closes down when the lens is removed from the camera. So, in this experiment, I used a paperclip to open it most of the way. I can play with changing that. Is the tradeoff more crispness, but less light and DOF?
The working distance is better than an objective and that makes it easier to light.

scitch · Post by **scitch** » Sat Jul 17, 2010 12:05 pm

Nevermind that tradeoff question. I went and read the thread you recommended. So, now my question would be, what's the advantage of a smaller aperature? Seems like larger has all of the advantages.

rjlittlefield · Post by **rjlittlefield** » Sat Jul 17, 2010 1:16 pm

scitch wrote:Is the tradeoff more crispness, but less light and DOF? ... Never mind that tradeoff question. ... So, now my question would be, what's the advantage of a smaller aperture? Seems like larger has all of the advantages.

The tradeoff question is a good one -- let's go back there.

For each single exposure of a 3D subject, the image you get is affected by three different phenomena:
1. Diffraction adds blur over the entire image. Smaller apertures (larger f-numbers) add more blur.
2. Geometry adds blur to out-of-focus regions. Larger apertures (smaller f-numbers) make out-of-focus regions be more blurred.
3. Aberrations add blur over the entire image, often worse in out-of-focus regions. Larger apertures expose larger aberrations, hence more blur.

So...

Sharpness in the plane of best focus is a tradeoff between diffraction and aberrations. Larger apertures give sharper images only up to the point that aberrations start to dominate. For typical camera lenses, this occurs at 2-3 stops below wide open. Look again at http://www.photomacrography.net/forum/v ... .php?t=424, second panel, second column -- the one labeled "Mamiya Sekor 55mm f/1.8". The overall blur at f/8 and below is primarily due to diffraction. (BTW, that "f/8" is a nominal setting. The effective aperture was more like f/50, due to bellows extension.)

When you are stacking, you always have the option of working at whatever aperture gives the very sharpest image. However, for many applications all that resolution is not really needed. Examples would be if you're only going to publish on the web at 500 pixels, or in a textbook at 1-1/2" wide. In those cases, it is more efficient to use a smaller aperture and correspondingly larger focus steps, giving fewer frames in the stack. The workflow becomes 1) decide how much resolution you need, 2) choose the smallest aperture that gives just that much resolution, and 3) choose a focus step that just avoids visible banding.

When you are not stacking, and your subject is 3D, then you have to balance loss of sharpness due to diffraction against loss of DOF. In this case there is an optimum aperture that gives just barely the sharpness you need, over the greatest depth. That aperture is likely to be quite a bit smaller than the one that will give the absolute sharpest image in the plane of best focus.

Making sense?

--Rik