comparison

rjlittlefield · Post by **rjlittlefield** » Fri Jan 17, 2014 11:20 am

robirdman wrote:The previous 2 images were the same stack, just p& D methods in Zerene.

Other way up, I think: DMap on the top, PMax on the bottom.

Here are 2 area crops of the better, first image.

The bold contour lines are a direct result of extreme variation in brightness across frames. What "depth map" methods do is to simply select pixel values from input frames based on their estimate of which image is in best focus. The map is always continuous, so if the software estimates frame 7 in one place and frame 12 in another, then between the two, pixels will be selected from frames 8, 9, 10, and 11 also. If frames 8 and 11 happen to be unusually bright, then the collection of pixels chosen from those frames will result in two bright "contour lines".

PMax works in a completely different way that is far too complicated to explain here. Suffice to say that it works mainly with local differences in pixel values, so it's intrinsically much more resistant to variations in overall brightness from one frame to another.

Zerene Stacker is programmed to correct for minor differences in brightness between frames, but it can't handle the extreme variation shown for this collection of photos.

I strongly recommend turning off exposure metering (TTL). Even if everything is running perfectly, you can still get random variation in the chosen exposure if you're unlucky. (2.5 with small noise sometimes rounds to 2 and sometimes to 3.) Just run your flashes in manual mode.

The fact that you're getting an occasional bright frame instead of an occasional dark one suggests to me that something is going wrong in the metering. That could still be caused by the amount of charge on the flash; if the metering pulse ends up being dim, then the exposure pulse will end up being bright.

Helicon warned about the bightness differences, but Zerene didn't.

Yeah, that's one of a sizable collection of "detect bad input" features that's on the to-do list but never seems to make it all the way to the top.

--Rik

rjlittlefield · Post by **rjlittlefield** » Fri Jan 17, 2014 11:21 am

By the way, I'm thinking this thread should get moved to the Technical Discussions forum. It's really out of place here in the Gallery. I'll move it after you've had a chance to see this and reply.

--Rik

robirdman · Post by **robirdman** » Fri Jan 17, 2014 12:07 pm

Yes can be moved.

Yes the images got reversed because for some reason I kept getting a message that the first was too large., even though it was only ~300KB. I had loaded ones before that were about 1Mb, so I don't know why, I was changing the name in the save and made it even smaller and then it finally loaded.
I was doing manual at first but when I did the ruler test series, I switched to TTL as I had to keep changing the setting as the apertures changed. then I only had to adjust mainly for the last couple of extreme exposures.

I did change back to manual, earlier today and ran a stack on a small Bembidion beetle. the exposures looked good and regular but the stacks were so poor that I just deleted everything. Nothing was sharp, not sure why.

rjlittlefield · Post by **rjlittlefield** » Fri Jan 17, 2014 12:39 pm

robirdman wrote:...for some reason I kept getting a message that the first was too large., even though it was only ~300KB. I had loaded ones before that were about 1Mb, so I don't know why.

The upload process has a strange quirk. Images that exceed the allowed pixel count (1024x1024) are automatically downsized without comment, but images that only exceed the allowed file length (300 KB) are rejected. That's a hard limit, by the way, if "~300KB" means 320 KB, it won't work.

Nothing was sharp, not sure why.

That's a puzzlement, for sure.

Shooting with consumer flash on low power, motion is not likely to be a problem. (Consumer flashes get shorter on low power; studio flashes are likely to get longer.)

If you're shooting with a microscope on front of that 200, then I strongly recommend to set the 200 on infinity focus. That's the way the objective is intended to work. Focusing the 200 closer may get you some added magnification, but it also drags the objective out of its design space. That may work OK, or it may not, but in either case that's an aspect to be explored after you have your other problems under control.

--Rik

rjlittlefield · Post by **rjlittlefield** » Fri Jan 17, 2014 12:54 pm

Addressing a question asked earlier:

robirdman wrote:I also have a D300s, but why use that if I have a D4 and D3.

You might want to use the D300S because it has smaller pixels with only 5.6 microns pitch (12 megapixels on a 23.6x15.8 mm sensor), while the D4 and D3 have larger pixels with 7.3 and 8.5 microns pitch (16 and 12 megapixels on a 36x23.9 mm sensor).

Under good shooting conditions using the same optics, smaller pixels translate directly to more detail resolved on subject. If you want maximum detail, you need to use a sensor that's matched to your optics. In many cases that's the D300S.

That said, if your images are not sharp in the center using the D4, they would look even worse with the D300S.

--Rik

robirdman · Post by **robirdman** » Fri Jan 17, 2014 1:49 pm

The motion I am concerned about is the camera shake. When I look through the viewfinder after each shot, there is a vibration that takes several seconds to stop. I am concerned that if the flash fires during the vibration, the lens may not be in the exact same position (besides forward) when it fires, so I am giving 6 seconds for settling between.

The head shots were with the 4x objective. 200mm macro at minimum. I had on a previous series somewhere less as I thought I'd have a smaller images but it didn't seem to change size. The images then had vignetting, so I thought it should be at the x1 macro setting.

On the earlier series today, it was 200mm with 6T closeup. Per a suggestion, I shot first manually moving the rail 2mm steps from the pin head to the top of the beetle. Then auto distance for the rest. I think it was about 49 steps.
I redid using 207 steps from the top of the pin to the farthest part of the beetle. Unfortunately, I am now having a download problem with the card. Normally it is very fast, but now it says it is taking 3 hours, so that's it for today.

This still is a small image and maybe I need a 2x objective. What size does the 4x on 200m yield? It seems like about 6x, and too much for this whole Bembidion.

I'll try the 300s sometime. Each day I think I'll try this for a while and then the whole day is gone.

NikonUser · Post by **NikonUser** » Fri Jan 17, 2014 3:10 pm

I am totally perplexed with the problems you are having.
On the face of it you are attempting what is perhaps the simplest macrophotography assignment. It should be a no-brainer.

Anyway, Nikon flashes seem to behave strangely when one is used as the commander. I got around the problem by using a SU-800 as the commander and put a SB-800 and a SB-900 on remote.

Incidentally if the D4 has live view then attaching it to a HDMI TV via a HDMI cable will show you your images in great detail; great for critical focusing and for checking the image immediately after the exposure

robirdman · Post by **robirdman** » Fri Jan 17, 2014 3:46 pm

Still waiting for the last series to finish downloading.
Nikon tech first told me (incorrectly) that there program doesn't preview)

I got a trial ControlmyNikon and its comparison chart showed many more capabilities than Nikon Camera Control. I had to email tech support because so many things in the video of stacking didn't follow, without changing settings in preferences, and then I keep getting an unexplained error when trying to view a shot. This they can't explain.

My first problem in the Liveview is that I am not getting enough light to see detail. The other is that when the camera is tethered to a computer through USB, the software isn't integrated with Stackshot and the rail won't move.

The laptop does have HDMI I think but I have to get a cord.

Perhaps part of my problem is the distance of each step to move. Is there a formula, as I am just guessing when it is in between. For example, if the field is 25mm and the mag is 1.8, and F11: 1.4x is .14 and 2x is .067. I should be closer to the latter but I am not sure. If I choose intermediate aperture such as f14 then more interpolation (guessing) is needed.

johan · Post by **johan** » Fri Jan 17, 2014 4:09 pm

Beware, unless laptops have changed in the last couple of years, as far as I'm aware the don't usually have HDMI in, it's usually HDMI out. You'll want to use a PC monitor, TV or field monitor with HDMI in. I think there are some laptops with HDMI in but you might want to doublecheck.

May I make a couple of suggestions, much in line with other people's.

A) ditch the white flash coverings and instead surround the specimen with a diffuse dome covering its entire sky like a simple foamcup or if you don't have that handy, tissie paper. Have a look at this especially the diagram. What you're doing is most like 'flash', you want to be more like 'diffuser'.

B) do some testruns with your flashes but strictly on manual settings, 1/16 or 1/32 usually works for me. Doesn't hugely matter what you shoot, you're mainly after consistency of output. I make my shots every 10-20s or so with flashes on manual and that works just fine. My hunch would be that the mush you're seeing in some places is because your base images vary quite a bit and that causes mush if they don't register their bit of the photo correctly.

Once you have A and B sorted do a stack from beyond to behind using JPG highest quality settings (might be slow). I always err generously on both of these to have breathing space. Then select a sensible portion of these to the boundaries you want and run a stack using your software of choice.

robirdman · Post by **robirdman** » Fri Jan 17, 2014 5:06 pm

What a wasted day. The result of stacking 207 at f14. No better than the first discarded series.

rjlittlefield · Post by **rjlittlefield** » Mon Jan 20, 2014 2:17 pm

NikonUser wrote:I am totally perplexed with the problems you are having.
On the face of it you are attempting what is perhaps the simplest macrophotography assignment. It should be a no-brainer.

I agree. Using this equipment, producing a sharp clear image should be a matter of aim, focus, shoot, stack.

I am completely baffled by this continuing stream of blurred results.

I look forward to learning what the problem turns out to be, but at this time I have no further ideas.

--Rik

robirdman · Post by **robirdman** » Mon Jan 20, 2014 6:02 pm

I am wondering if it has something to do with the small size of the beetle compared with the size of the pin. Every time I try one about this size, I have bad results. Even trying my method of just shooting a single at high aperture, I don't get a image comparable to what I usually do. Perhaps the pin sticking up a quarter inch or more, just creates a blur, also scattering the light, over most of the beetle.

ChrisR · Post by **ChrisR** » Tue Jan 21, 2014 3:12 am

Then - try pushing the pin head as far as possible towards the beetle?
Snip the head off a pin, dot it with matt black?

For steps, for effective apertture and sensor in use, try this:
http://www.photomacrography.net/forum/v ... 606#126606 .
It works out when classic depth of field calculation is overwhelmed by small-aperture induced diffraction blur, and gives a figure for DOF, which is a good start point. Don't be too worried about the warning at the bottom of the spreadsheet; it's quite a strict judgement, and it's often impossible to achieve an optimum effective aperture with the optics available.

Broadly
- there's no harm in stepping finer (and you will tell more easily if the subject has moved between shots)
- if you're going in steps too wide then you should still have sharp bands - apparently lacking in this last beetle.

Repeating - you said you're getting 6x using a 4x objective on a 200mm. That's wrong? Sounds like the 200mm is focused other than at infinity.

johan · Post by **johan** » Tue Jan 21, 2014 4:14 am

Chris S. wrote: I think you should stop stacking--temporarily--and concentrate on getting a single shot that is very sharp within its limited zone of focus. This is a much quicker way to find and eliminate problems you may be having, rather than doing stacks of what may well be flawed source images.

--Chris S.

Another mystified observer here.

I think this from an earlier post is very good advice.

Get 1 nice and sharp and the rest will follow.

robirdman · Post by **robirdman** » Tue Jan 21, 2014 3:08 pm

Oh brother. I spent the day on this one beetle and then prepared my post with several images, and then it said the website timed out and I lost everything, as it wouldn't reload. Then the network device wasn't lighting and a replacement didn't either, though it did on my laptop. Then I tried to shut down , and it just stuck there, so I just powered it off. An hour and a half later I am ready again.

After the last dismal results, I skipped a couple of days and spend time photographing the only breeding male White Winged Scoter that I have seen in Chicago, since I started bird photography in 1087. I had much better results.
Today I returned to the stacking with a larger beetle about 1/2" in size. I have since got a small ultrasonic cleaner, and I also got precision tweezers and an Optivisor so I was able to get a broken off leg on this one closer to where it belongs.

I was pretty happy with the results as far as documenting this collection. by comparison, before I started this stacking, I had just gone ahead and photographed about a hundred beetles, one after the other, before I saw what a problem cleaning them was and what a mistake it was to leave the labels on, because it would take more than an hour to clean up.

For example here is what I was dealing with before, with the single shot approach. A really unusual Carabid beetle, I have as Panagaeus.

I ran a stack with the 200mm macro and 5T close-up. It was 61 steps at .245mm, f18. I was initially highly pleased with the result as far as documenting the collection. but when I looked at higher mag, I saw that the head was lacking detail and there was bad overlap of the antennae. I ran another with f16, but the results weren't that different.

I then switched to the 60mm lens with 5T. When I reviewed the stack, I noticed that each shot moved slight to the right until there was a jump between the last and first. I found that when I put the camera back on the rail, it wasn't aligned. but the results didn't look that different.
I decide to do a really fine run with 207 steps at .035mm, f8. Some improvement, but not what I would expect with almost 8 times the steps so I don't know how I can improve. Still the overlap and lacking head detail.

I did both P&D on each series but am only including both here on the last one.