Introduction and Request for Assistance

[jessbussert]

Hello folks!

I'm Jessica. I've been a beginner macro photographer for about 4 years now. About that time I became fascinated with the smallest blossoms I could find, usually those from weeds and grasses. The reason I call myself a beginner is because I have a bunch of lower end and homemade equipment that I use for most of my shots. Unfortunately, my results reveal my humble methods. Regardless, it is what I have to work with. I'm hoping that you all might be willing to offer me some advice in order to improve my output.

First off let me tell you about my kit. I shoot with a Canon t2i running Magic Lantern, an EFS 18-55 lens, and a collection of dumb extension tubes. Since I have no aperture control through the inexpensive tubes, I use a laser-cut black paper disc with a 5mm hole cut into it for my aperture. I use a home built focusing rail that gives me about 8cm of travel in approximately .02mm steps. For stacking I predominantly use Photoshop CS5.5 but I've recently started playing with Helicon and Zerene as well. That said, I tend to have my best results with Photoshop and a bunch of manual clean-up.

Regarding the paper aperture, the 5mm hole size seems to give me the best results. I've tried larger but the resulting blurring of foreground elements seems to confuse all of the software products mentioned above. I've also tried manually setting my lens aperture to the recommended f8~f11 range with the same results. Before I learned about pinhole diffusion I had also tried very small aperture sizes (<1mm) with the expected results.

For lighting I usually use a collection of single LEDs on pose-able wire arms that let me nicely configure a dramatic lightscape for my images. I can use these as spots, backlights, floods, and more, and they really help my subjects pop nicely. I much prefer the look I get using this method to other, more diffuse light sources. I have bounced a flash now and then, but the LEDs are my go-to solution.

When I do a shoot I usually take 50 to 100 images at around .2mm steps going front to back through my subject. This has given me some lovely results. After trying to educate myself so as to improve my process I've recently changed things up a bit. For one of my more recent experiments I took 350 images at .02mm steps using the lenses own aperture set to f8. My final results were rubbish after processing with all three modes from Helicon and both modes in Zerene. I could probably get something from Photoshop but my initial processing didn't look any better than my original process using fewer images. I am totally confused by this. It was my assumption that using an "optimal" f-stop and a 10x smaller step size I would be able to get substantially better results. The opposite seems to be the case.

What suggestions do you all have? I'm posting two images for your consideration. That said I'm a little self conscious after looking at the amazing results you all come up with.

Any help or suggestions would be greatly appreciated. Tossing out all my old kit and buying new isn't an option at this time. Now, I would be happy to take on any old cast-off equipment that you might have lying around... *grins*

Thanks in advance,
Jessica

[rjlittlefield]

Jessica, welcome aboard!

As quick introduction, I'm the fellow who wrote Zerene Stacker, and I answer all the emailed support requests about that software. I'm also insatiably curious about most things macro, and there are several aspects of your situation that I find very interesting.

In terms of what you're doing right now, several things jump out at me:

1. The images that you've shown look quite good. Kudos for that.

2. You're shooting at greater than 1:1 (about 2:1 = 2X in your examples), but I don't see anything about reversing your lens. If in fact you're not reversing your lens, then you're almost certainly getting some significant aberrations that could be eliminated for the small cost of a reversing ring.

3. Laser cutting a paper aperture is a technique that I don't recall reading about before. I would be interested to hear more about that process. However, the use of any external aperture with an 18-55 mm lens on extension tubes is likely to introduce aberrations that again are not necessary. There is a simple though not intuitive method for setting the built-in aperture of Canon lenses for use on non-automatic tubes. See Shooting with a reversed 18-55 mm Canon kit lens for explanation.

4. Using un-diffused small light sources can be the source of problems that are quite difficult to debug. See Reflections of hard and soft light in a spider's eye for some explanation of why that is. I note your preference for "pop", which I also appreciate, but be aware that there can be tradeoffs to be evaluated. You may end up getting better results from adding some small diffusers to minimize artifacts.

5. The report of getting best results from Photoshop CS5.5 versus both Helicon Focus and Zerene Stacker is unusual, to say the least. I don't know what's causing this result, and I'll be very interested to learn what's happening. In general, Photoshop mostly makes "errors of omission" -- leaving blurred areas that were sharp in some source -- while Helicon Focus and Zerene Stacker mostly make "errors of commission" -- doing a better job of retaining focused detail but also introducing halos of various sorts. It's possible that your result stems from personal preference about what sorts of defects are less annoying. But I'm suspecting that a) something about your shooting process is causing HF and ZS to introduce more artifacts than usual, and b) you haven't figured out yet how to reduce those, either by modifying the shoot or by setting controls in the software.

6. If you're using 0.2 mm steps at 2:1 and nominal f/8, as suggested by various bits of your post, then that combination is not far off optimal. Effective aperture typically is about nominal*(magnification+1), so at 2:1 and f/8 you would be running at effective f/24 [because 8*(2+1)=24]. Referencing Tables 2A or 2B at https://www.zerenesystems.com/cms/stack ... romicrodof, the combination of 2X and nominal f/8 or effective f/24 indicates a per-step DOF of slightly over 0.2 mm. 350 images at .02 mm sounds like significant overkill. I'm not surprised that more frames gives you about the same results as fewer images. But I am curious to know what leads you to say "optimal". Was the magnification much different from 2X?

As a next step, I would be very interested to see some images that show in more detail what your source and final images look like, so that we can understand better what difficulties you're running into with the stacking. Both whole frames (resized to 1024 pixels wide) and 100% crops (not resized, but cropped to 1024 pixels wide) would be very helpful. If you want to post those in Technique and Technical Discussions, there is no limitation on how many you can post.

I'm looking forward to more information!

--Rik

[jessbussert]

Hi Rik,

First off, YOU ARE 'DA BOMB! Thank you, thank you, thank you for all the information!

To address a few of your points let me begin by saying that I'm not reversing my lens. I'm only using extension tubes. I had heard of that method but had assumed that you only did it if you didn't have the tubes. Now I know better. I'm still confused as to why I'd get more or less aberrations depending on the orientation of the lens. It's the same glass so why would it differ? Regardless, I'll be buying a reversing ring in the morning!

Regarding the lenses own aperture, I hadn't been doing it correctly. After reading the post you referenced I'll be able to try again. In the absence of this the only way I had to create an aperture was the lasercut method I described above. Basically what I did was place my two shortest extension rings closest to the lens and placed my paper aperture between these two. I experimented with different sizes and found 5mm to give the best results. I have no way to effectively correlate my 5mm hole to a proper aperture number. It does seem to work with my current setup.

While I love the "pop" my LEDs give me, I often get a glittery look from my flower petals that I assume is from the lighting. They sometimes almost look like sugar confections as is apparent in the first image I posted above. I might try to use a diffused light as my flood and still incorporate some additional lights as spots. My background is landscape photography and I've got to have some drama in my lights!

I think the reason I'm getting the results I do in PS is just because I'm more familiar with it. I load and align my stack, duplicate the layers, focus blend them, and merge the results. This gives me a good starting blend sitting on top of my stack of aligned images. After I get to this point I simply start working my way through the stack, replacing any soft areas in the blend with corresponding sharper elements from below. Finally, depending on my subject I may have to do a bit of clone stamping to eliminate any foreground halos found on background elements. Suffice it to say it's a pain in the butt! I'd love to learn a better method which is why I've started investigating other software. Any tutorials or assistance in this regard would be greatly appreciated.

You taught me something else helpful regarding the step length. I had assumed that the more steps I made the better the final results. Now I know that isn't necessarily so.

I'm hoping to ultimately achieve much greater magnification and utilize a pan-and-scan stitch method to image my entire subject. Currently my rig only has one powered axis but I plan on making a different one with the three axes in order to automate the pan-and-scan as well as the focal point. I've got big plans for my small subjects!

I'm attaching a few photos of my current rig for your amusement. A servo in back drives a 1/4-20 threaded rod and is controlled by an Arduino that also triggers an external remote shutter release. My subject sits on a three axis gimbal I hacked together from a webcam mount. The mounting for the camera is designed so I can rotate the camera around its shooting axis for different framing of my shots. And lastly, my lights are just a bunch of LEDs on bendable, wire arms. All of this is powered by a bundle of NiMH batteries so I can take the whole rig into the field if I wish.

I'm drooling over all the lovely kit people have access to on this forum. I'm in a place that if a part isn't cheap I just need to hack it or make it on my own. Its Po' folk MacGyvering at its best!

If you'd like to chat further or help me best figure out Zerene I'd welcome any help I can get. If you PM me I'll send you my email. Thanks for all the help so far!

Sincerely,
Jessica

[Lou Jost]

That's actually a pretty ingenious set-up!

[jessbussert]

*blushes*

Thanks for the kind words.

-J

[ChrisR]

Welcome to the forum Jessica!
It seems to me you've done very well coping with things for yourself, with just a few unknown unknowns cropping up.

The rest of the tutorials and explanation on the Zerene site will probably fill a couple in. There are several FAQs here, too, it's worth scanning through many of them.

Some bits (- apologies if you were already aware!):

Lenses tend to work better the way they were designed for, with reference to the focused distances in front and behind. ("Conjugate distances"). So if you reverse a lens for close work, those are the better way round as you go close. Though your images "ain't broke", so...

If you aren't using Live View through the stack, try it. It prevents the first shutter curtain wobbling the camera, because it's already open. Pause of a few (~3) seconds between the shots should let the mirror vibrations settle.

Be careful with undiffused LEDs.
You know about pinholes and depth of field. If you use a point light source, it can reflect off parts of the subject as a dot of bright light. That effectively goes through a tiny part of the lens, so it's like you're using a tiny aperture, just for that light source. That means it's in focus though a lot of the frames in the stack in odd directions, but actually not very sharp (due to diffraction, like the pinhole image). That can mess things up, as you can imagine. A sheet of Kleenex very close to the subject (rather than close to the light) will have a massive effect on the light.

Living in the US is good for eBay bargains! There are for example sub $20 lenses which work very well at around 4x. Adapters are often cheap from China too. (By the way your APS sensor is 22.2mm wide so 4x means ~5.5mm subject width)

It's an inspiring subject, so we'll be watching with interest

[rjlittlefield]

I'm still confused as to why I'd get more or less aberrations depending on the orientation of the lens. It's the same glass so why would it differ?

It's a matter of where the focus planes are located with respect to the glass. Any lens behaves best when the focus planes are located at specific distances from the glass elements. For example a lens that is designed for landscape photography will behave best when one focus plane is far in front of the lens and the other is close behind the rear of the lens, in the position where the sensor would be located at infinity focus. As you force the lens to focus at other distances, the lens develops aberrations because some of the light rays do not get bent by the ideal amounts. The problem gets progressively worse with increasing difference between as-designed and as-used focus arrangements.

The purpose of reversing a lens is to reduce the difference between as-designed and as-used focus arrangements. It allows a lens that works well at say 0.5X to also work equally well, reversed, at 2.0X (2.0 = 1/0.5).

[the added paper aperture] does seem to work with my current setup

External aperture will always work OK at image center. The problem is away from center, where using an external aperture will force light rays that form the image to pass through portions of the lens that are unusually far away from center. Those outer portions of the lens are usually not as well corrected as more central portions, so the image quality degrades. The effect is usually not dramatic, but it can definitely be significant. See FAQ: Stopping down a lens combo for more discussion.

To clarify, when I wrote "Laser cutting a paper aperture is a technique that I don't recall reading about before", I was referring specifically to the bit about laser cutting. Adding some external aperture is common in some special situations, but most people either use a commercial iris or paper that has been punched or cut with scissors. Laser cutting sounds great, but I would not even know where to look for such a device or service.

While I love the "pop" my LEDs give me, I often get a glittery look from my flower petals that I assume is from the lighting.

It's from the interaction between the lighting and the subject. The underlying issue is that small subjects tend to be locally shiny, as if they were made of many small polished parts stuck together. When viewing without magnification, our eyes cannot resolved the individual parts and we perceive the surfaces as matte. But with magnification, the individual parts do resolve, and then we see the glittery effect which is quite real. To compensate, one can adjust by using more diffused illumination. It's very much the same issue that comes up when shooting glassware in product photography. Small lights can be great when shooting fabric, but for wine glasses you probably want to bounce the light off a big piece of posterboard.

There are other, more subtle effects. It turns out that for small subjects, you can actually see finer detail with fewer artifacts by using more diffused lighting. For a couple of extreme examples, see http://www.photomacrography.net/forum/v ... 8613#58613 and http://www.photomacrography.net/forum/v ... 487#123487 (the block of pictures labeled as "3. This is also the dime, this time shot with the JML 21 and with one additional tweak. The first two frames are with undiffused illumination and a focus tweak; the last three have same focus and vary only in diffusion.").

I see that ChrisR has also posted, while I've been typing. The effects that I'm talking about are the same that he is, just with different words.

I'm attaching a few photos of my current rig for your amusement.

That's very elegant! My first stacking rigs were quite crude in comparison. (For your amusement, see my early setup in Figure 6 at http://www.janrik.net/insects/ExtendedD ... summer.htm )

--Rik