Troubleshooting

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

enricosavazzi
Posts: 1474
Joined: Sat Nov 21, 2009 2:41 pm
Location: Västerås, Sweden
Contact:

Post by enricosavazzi »

kaiotaku wrote:...
It's almost like the lens is stopped down causing major diffraction, but it can't be, I'd see vignetting if that were the case. The camera reports f/2.8.

...

Not sure where to go from here except trying to rent another 200mm f/2.8 or use a bellows
If you are using the 200 mm as a tube lens, then the effective aperture of the objective is used, not the lens aperture. This means in practice far narrower (higher f value) than f/2.8. So it is correct, and expected, that diffraction is higher than what you would expect from an f/2.8 tube lens. Diffraction and vignetting have different causes and are not related.

For the above reason, you don't really need an f/2.8 lens. An f/4 or f/5 tube lens is already more than enough. A 200 mm f/2.8 is overkill, and may in fact perform worse than a slower lens. With the objective mounted close to the front element of the lens, only the central portion of the front element is used, so in practice the tube lens becomes much slower. If you move the objective away from the lens, on the other hand, you will use more of the front element surface including its peripheral areas, but 200 mm f/2.8 lenses almost inevitably trade IQ for lens speed, and an f/4 or f/5 lens of good quality likely provides a better IQ than an f/2.8.

There is no reason and no need to use bellows (assuming you mean bellows placed between the 200 mm tube lens and the camera). In fact, the 200 mm lens must be focused at infinity.
--ES

MarkSturtevant
Posts: 1946
Joined: Sat Nov 21, 2015 6:52 pm
Location: Michigan, U.S.A.
Contact:

Post by MarkSturtevant »

I am not a stacker (so I might say something a bit wrong), but I would still recommend trying what ChrisR had suggested earlier, which is to take a single shot of something very flat. This will show a bad image if there is something wrong with the optics right now. Carefully check the focus at all 4 corners before the picture is taken to make sure everything is in one plane. If such a picture comes out good, then some other problem is indicated. My suspicion then would fall on there being something askew in the mount or the movement in the rail.
Mark Sturtevant
Dept. of Still Waters

kaiotaku
Posts: 22
Joined: Sun May 27, 2012 10:15 pm
Location: Melbourne, Australia

Post by kaiotaku »

Thanks for everyone's input.

I've done some more thinking, reading and experimenting with my gear, and I've made some progress but I'm not yet completely happy.

So, these problems first emerged whilst using the studio strobe (but also earlier with the Sigma 150mm). I also noticed that I can see sharp detail in live view, but can't capture that in an image. I can take perfectly fine images with straight macro lenses, the issues are restricted to using objectives but several different combinations of prime lens / objective were all generating the problem. Subsequently, I noticed that if I use ISO 400 and underexpose, even ISO 800, I get a considerably sharper image, better looking than ISO 100. Those were all big clues. But I also did some reading about the studio strobe that I've been using, the Elinchrom D-Lite RX 4.

Overall, this has exposed a couple of false assumptions.

Studio strobe flash duration is faster than speedlite flash duration = false assumption. The Elinchrom D-Lite RX 4 is rated to 1/800 but that is at max power. I couldn't find data for the lower power settings, but it could be as slow as 1/100 according to something I read. I saw some frames that seemed to indicate vibration (but not others). With 20MP at 20x, on a not-super-stable wooden table and wooden floorboards, all prone to vibration, 1/100 is probably not always fast enough to quell signs of vibration. (Any) flash is sufficiently fast to suppress vibration when photographing stationary objects = second false assumption. I found data indicating that the Canon 580EX flash gives 1/4000 at full power but still 1/500 at 1/4 power, and I try to keep my flash power below 1/4 since I've read that variability can increase close to max power, hence it's not good whilst stacking. You won't see vibrations with flash in regular macro photography like 1:3 or 1:2 but at 20x all bets are off I guess.

But, the bigger issue is that of internal reflections. I don't think there's anything wrong with my optics, rather it's a terrible problem with - call it what you like - flare, glare, veiling, etc. I did several things to counter this, such as shoot on black cloth rather than on the Perspex stand, and put an improvised lens hood on the front of the objective. But most importantly, I flocked over the the gold threads at the back of the objective, and this produced a significant improvement in the image. Clearly there was some sort of internal issue there.

Image

Image

I then went back and retested the Mitutoyo 20x on the Sigma 150mm and it was significantly improved. So that lens does perhaps have some potential as a tube lens.

However, my images still aren't quite right. Not a sharpness issue so much as one of deteriorated contrast / microcontrast. They just have a veiled, dull, ghosted sort of appearance to them. When compared to the previous images, they have much worse clarity. And I still can't explain why I didn't cop these issues before, considering that nothing much has changed. Perhaps I have more than one 52-25mm ring lying around and one is better than the other somehow, I just don't know.

So overally, as far I can tell, the swap to the Elinchrom brought several interacting factors to the fore, the new strobe was putting a) more light into the system and b) putting it in slower, and this both overwhelmed/exposed the (rather low) tolerance of the system for internal flare and also (sometimes) allowed vibration to become visible.

It's a shame to lose the benefits of the studio flash. I suppose I could still use it at higher power settings to get faster duration and use some kind of
ND to bring the intensity down to what I need.

I'm still planning to switch to using a bellows or extension tube system using a Raynox, since I'm skeptical that I can completely fix the images using my current gear. Something is still badly wrong. But hopefully this info might be useful for anyone else who strikes similar problems. Aside from that, I guess I could try photographing somewhere with a concrete slab floor and solid benchtop, or swap in some different tube lenses if I can get access to them. But sadly, I have to go away for work for 4 weeks so everything must be shelved for a while.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Glad it's getting better!
If you start using bellows etc, you'll have another set of flocking and masking to do, of course.

Some time ago, after masking the rear of an objective like you did, (protostar) I also masked the camera lens . It might seem pointless, but if there IS any light bouncing about in the gap between the glasses, it'll go into the camera which you don't want.

How big a hole ?
The light from the objective to the camera lens is focused at infinity, but the cone still diverges from the objective, albeit in bundles of parallel rays.
The Mitutoyo tube lens is quite small (iirc) an inch or so across, and they specify a distance of about 100mm. OK, so that implies a cone going from objective (say a 10mm hole) to the tube lens (say a 20mm hole) in 100mm. Quite a narrow angle. So I figured I could mask off all but about the central 15mm of camera lens.
I'm sure it worked. Like I'm sure ny car goes faster when I wash it.

Yes Studio Flash tends to be slow. Some are fastest at maximum power, some the reverse.
Alien Bees eg Einstein is relatively affordable at iirc $500- $700 , and will give you 1/9000th (or so, depending what you read)
One can mention Hensel 400ws "SpeedMax", at 3.1ws (minimum)
1/66,600 sec., 31 flashes per second, $5000.
Moral - stick to camera flashguns/speedlights.


You wrote -
Canon 580EX flash gives 1/4000 at full power but still 1/500 at 1/4 power,
which looks wrong - hunting....
Chris R

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Yep, this is what I was expecting:
Canon 580Ex Flash Duration:

Here: http://www.photosbykev.com/wordpress/20 ... -duration/
Posted on July 12, 2008 by photosbykev

I was curious as to what I could freeze using my Canon 580Ex flashgun so I had a play in the lab using a photodiode and scope to measure the flash duration of the flashgun (averaged value over 4 flashes) at the various power output settings. The results were:

1/1 power = 1/1000 second
1/2 power = 1/2000
1/4 power = 1/4000
1/8 power = 1/9000
1/16 power = 1/15000
1/32 power = 1/21000
1/64 power = 1/30000
1/128 power = 1/35000


I expect cheap third party flashes would be similar. I'd avoid Yongnuo, having had three, all of which have failed or didn't work properly on arrival, and they won't put them right.
Chris R

kaiotaku
Posts: 22
Joined: Sun May 27, 2012 10:15 pm
Location: Melbourne, Australia

Post by kaiotaku »

Hi ChrisR,

Oh yes, I got the info from here, but I messed up badly:
https://gock.net/blog/2012/01/flash-dur ... l-strobes/
Serves me right for speed reading / relying on memory.

It's actually the reverse of studio strobe, in that the higher powers are slower and the lower powers faster.

Nevertheless, 1/2000 at 1/4 is much faster than the studio strobe.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Not all studio strobes are slow, the Adorama (Godox) XPLOR 600 and Flashpoint Rapid 600 have flash outputs that can be faster than Speedlights at certain power levels. Do a search here to find details on these strobes, or check Adorama (Godox) online manuals.

I don't know about the Elinchrom D-Lite RX 4, but most strobes I have don't get slower at lower power, they get a little faster if they are voltage control mode, and a lot faster in optical width controlled mode (this is how speedlights are controlled and the studio strobes I mentioned).

A simple test you can use with your Elinchrom is to capture an image of a high speed rotating fan blade. Put marks on the ends of the blades so you can access the blurring. Use a shutter speed and small aperture so only the flash exposes the fan blade. Vary the flash power and see how the blue behaves. If you put a scale behind the blade a little math can get you the actual exposure time since you know or can find the fan speed, and can figure the angular velocity of the mark on blade tip by the the blade radius to the mark.

If you have access to an oscilloscope then a 1N4148 or 1N914 $0.05 diode can act as a optical detector. Just parallel a resistor of 10~100K with the diode/resistor and set the scope trigger off the input channel. Place the diode/resistor in front of the strobe and record the waveform as you vary the power. I did this some time ago for various speedlights and strobes I recall, you should be able to find it on here with a search.

Best,

Mike
Last edited by mawyatt on Sat Oct 14, 2017 5:25 pm, edited 1 time in total.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Yes Kaiotaku, that's what I wrote. Alien Bees has a flash burst which is cut shorter like a Speedlight, unique or one of only a few to do that.

Question - you're using a Canon - which one and in what mode? Mirror up?
Try dim room, rear curtain sync, 2 second exposure, to help kill vibrations.
Chris R

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ChrisR wrote:Yes Kaiotaku, that's what I wrote. Alien Bees has a flash burst which is cut shorter like a Speedlight, unique or one of only a few to do that.

Question - you're using a Canon - which one and in what mode? Mirror up?
Try dim room, rear curtain sync, 2 second exposure, to help kill vibrations.
Chris,

Both the studio strobes I mentioned above do this, and can be very fast, the Rapid 600 is 1/19,606 second at the fastest setting (can even do ~1/28000 in burst mode). The XPLOR is 1/10,000 I recall.

These use IGBT technology which interrupts the flash tube current, same technology that's in modern speedlights. I have an use the XPLOR 600 since they first came out about 3 years ago (got them for this very reason), and more recently got the Rapid 600 for the same reason, very fast optical burst at higher power levels than speedlights.

Best,

Mike

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I just had a quick look at the Godox site - there appear to be new, much cheaper monoblocs with IGBT's , hence short duration flashes.
Chris R

kaiotaku
Posts: 22
Joined: Sun May 27, 2012 10:15 pm
Location: Melbourne, Australia

Post by kaiotaku »

Here's an example of where I'm at now.

These are both Nikon Plan 10x NA 0.25 (MRL00102) on Canon 200mm f/2.8.

Same species but different specimens. I won't show the whole image in this case because the species is unnamed, my co-author would be displeased. In any case, you can clearly see the decent clarity of the old image and the (still) horrible veiling in the current image.

Taken yesterday:
Image

Taken previously (12+ months ago):
Image

Those are 100% crops, straight from Zerene. I think the actual sharpness is perhaps comparable, as you'd expect, it's just that the glare is masking it. Believe it or not, that image is still better than what it was previously and prior to flocking the rear of the objective. Still poor though.

I photograph these specimens on SEM carbon tabs, the stub of which I shove into some Blu-Tak which rests on a scrap of black cardboard. This way I can rotate / angle the specimen easily, and the background is a very pure black. It had occurred to me that if some of the tack was protruding beyond the margin of the carbon tab to the extent that it becomes visible to the objective this could have provoked flare, but I doubt that this was the case.

Still quite mystified by this - same equipment, different result. But at least I think I've narrowed down the probable basic cause now.

kaiotaku
Posts: 22
Joined: Sun May 27, 2012 10:15 pm
Location: Melbourne, Australia

Post by kaiotaku »

ChrisR wrote:Yes Kaiotaku, that's what I wrote. Alien Bees has a flash burst which is cut shorter like a Speedlight, unique or one of only a few to do that.

Question - you're using a Canon - which one and in what mode? Mirror up?
Try dim room, rear curtain sync, 2 second exposure, to help kill vibrations.
I'm using a Canon 70D, mirror lockup, 1/200.

Regardless of how harmful they are, I don't hold much hope of quenching vibrations so long as I live here. The wooden floor is rather unstable and just makes it impossible. In live view at 10x I can see all kinds of jittering going on, even when nothing much is happening. I'm planning to move house in about 2 months anyway, I might need to keep a solid slab floor in mind as a prerequisite for the next place.

I remember a previous thread here in which someone was having trouble with his images, but eventually traced it to vibrations in the building itself. Moving the rig from the 6th floor to the basement solved the issue, which I found amusing.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Chris,

Godox has a SK300II (Adorama Studio 300 AC) which isn't IGBT controlled but does achieve a 1/800 at full power and 1/2000 at 1/16 power output.

These are quite good for macro use and only cost $120, and include a built in receiver.

Best,

Mike

kaiotaku
Posts: 22
Joined: Sun May 27, 2012 10:15 pm
Location: Melbourne, Australia

Post by kaiotaku »

Actually, I just noticed that some batches of carbon tabs are very slightly smaller than the stubs on which they sit, leaving a thin ring of exposed metal around the edge. This metal, exposed to intense flash, could easily be the cause of the flare, since it'd be pumping bright light into the objective at an oblique angle from outside the frame. This is my best theory for now, but I won't be able to test it for a few weeks.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

I'm glad to hear that you've figured out the cause of blurring shown in the early part of this thread.

About the current problem...
kaiotaku wrote:In any case, you can clearly see the decent clarity of the old image and the (still) horrible veiling in the current image.
What I'm seeing in the images now does not look to me like veiling glare due to stray light bouncing around. Here is the histogram of the whole image, shown next to a slice of the top of the image. In the upper left corner of the image, the background is nearly black, values only 7 on a scale of 0-255. In the histogram, the gap at bottom is pretty narrow, not much wider than my reference images at http://www.photomacrography.net/forum/v ... hp?t=35350 (4th panel, bottom two rows).

Image

I'm wondering now if the obvious difference in local contrast is perhaps due to difference in illumination (new illumination seems a lot softer, more diffused).

Another possible issue is "Brightness" adjustment in Zerene (at Options > Preferences > Alignment). If you have that selected, then what actually happens is that the offset and gamma of each frame get adjusted so that the mean and variance of luminance values in each adjusted frame match those in the first frame processed. If that first frame happens to be unusually low contrast, say because it's entirely out of focus, then subsequent frames get adjusted to match. The nearly black background argues against this as a cause, but the only good ways to tell are to either look closely at individual frames, or turn off Brightness adjustment and see what effect that has. (Best is to turn off Brightness adjustment and stack with DMap, which should work pretty well for subjects with no complicated overlaps, like these seem to be.)
... if some of the tack was protruding beyond the margin of the carbon tab to the extent that it becomes visible to the objective this could have provoked flare, but I doubt that this was the case.
Remember also that microscope objectives have quite a wide entrance cone, around 29 degrees wide for NA 0.25 and 50 degrees wide for NA 0.42. It is surprisingly simple for stuff to intrude into this cone where it gets broad in the background. But that doesn't explain loss of contrast in opaque foreground areas. I'm still suspicious of illumination for that.

--Rik

Post Reply Previous topicNext topic