house ant, take 7, new diffuser set up... comments welcomed!

[ctron]

As with all images I post, I invite comments and tips for improvement. Diffusion, although a seemingly simple concept, has turned out to be the "Achille's heel" for me regarding my macro studies. I sincerely hope I'm starting to make good progress here.

About the ant first. It's all I have available right now and I have continued to use these tiny house ants because I find them to be challenging for this set up (I should really go with the 10x mag instead of 6.7x). When the ant is cooled, its body has the tendency to curl into sort of a "reverse" fetal position with the face and rear at the peak of the curl. I was able to even the body out a lot gently pushing the ant's center down while it was on its back. I would still like to position the legs as they should be but haven't found a satisfactory way. Remember, this isn't one of the larger ants, very tiny specimen. Any ideas about body manipulaton for a proper image pose welcome!

Note the attached image and new diffuser set up. I based this design on an already simple and existing design recommended in one of my diffusion posts. This was the first time I actually had to diffuse the camera flash itself, which I use as a trigger for the external Yongnuo flash, I suppose because the poly cup is now directly facing it. A yougurt container was placed over the camera flash and then the tissue paper carefully moved between camera flash and open cup/ objective area.

I angled the Yongnuo flash so that it impacted the outer cup surface and not the specimen directly in any way.

The raw stack seemed ok, but required some post processing. For one thing, the face was slightly dark. This wasn't all that critical, however, but a little bit darker would have been. Stacking with Dmap in Zerene always seems to produce a brighter, more uniform result, but I seem to always end up with smooth areas that aren't easily retouched without using a bunch of images.

One problem I keep running into is when to stop taking images for the stack. I keep stopping a bit too soon and this makes the extreme end portion of the image blurry. I suppose I might be able to mark start/ stop areas on the course focus knob on the Nikon block(?).

I can't really go by the back of the camera histogram for stacking. If I try to keep highlights where they should be, the final stack is too dark. Seems like I have to almost clip the highlights, but the stack doesn't turn out overexposed. Not sure what's going on there.

There's quite a bit of retouching I have to do in Photoshop afterwards to get rid of darker, blurry areas particularly around limbs, backside of ant, or other areas.

One big drawback is that I have to run the flash at almost full power, just one step down from full. The cup seems to diffuse (or at least I think it does... please be the judge for this image), but robs a lot of the light. I guess this is no problem really, but I am going to have to go with either AC adapter or battery pack for the flash. I find flash light difficult to get right and you have to already have a stack to see if your manipulations were successful or not. I tried the Ikea lights and they seem too dim; now I have some 10 w LED flood lights on the way-- perhaps I can use shorter shutter for sharper results.

To get the face of the ant brighter before post-processing, should I move the flash to the opposite side of the camera platform?

Again, thanks, in advance, for any commentary.

[Lou Jost]

That's really a big improvement!

[ctron]

That's really a big improvement!

Thanks, Lou!

[Dalantech]

I would honestly look for something else to use as a diffuser. That cup is doing an excellent job -of blocking the light from your flash. In any setup there is going to be some vibration, and you really want to keep your flash duration as short as possible to freeze motion. There are some who think that the duration of the flash, no matter how long, is going to be short enough to freeze motion but it's a myth. Although not as obvious as stopping a bullet as it passes through an apple, or a balloon in mid pop, macro is still a form of flash based stop motion photography. Motion as little as half the width of a pixel while the flash is firing will rob you of detail.

For help with diffusion see the Apparent Light Size article at Strobist. Not only is that cup doing more to block the light, as a diffusion surface it's not large enough to give you really good specular highlights. This is the light quality you should try to achieve <-- Not my photo, but someone I follow at Deviant Art. His work recently made the cover of the Arabic edition of National Geographic magazine.

[ctron]

I would honestly look for something else to use as a diffuser. That cup is doing an excellent job -of blocking the light from your flash. In any setup there is going to be some vibration, and you really want to keep your flash duration as short as possible to freeze motion. There are some who think that the duration of the flash, no matter how long, is going to be short enough to freeze motion but it's a myth. Although not as obvious as stopping a bullet as it passes through an apple, or a balloon in mid pop, macro is still a form of flash based stop motion photography. Motion as little as half the width of a pixel while the flash is firing will rob you of detail.

For help with diffusion see the Apparent Light Size article at Strobist. Not only is that cup doing more to block the light, as a diffusion surface it's not large enough to give you really good specular highlights. This is the light quality you should try to achieve <-- Not my photo, but someone I follow at Deviant Art. His work recently made the cover of the Arabic edition of National Geographic magazine.

Yes, I read the article yesterday but, for me, it caused me to have more questions than answers. Plenty of folks have chimed in about diffusion, and I have also looked at a bunch of such set ups, but I'm finding it really hard to get right and the concept is, in fact, holding me up from making a permanent set up.

The biggest drawback from what I'm finding is that I don't really know how well or poorly diffusion changes make until after I've snapped the 300 or so frames and a stack is made. Trying to just take a couple of images here and there before stacking isn't cutting it. For example, someone suggested covering the outside of the poly cup I'm using with aluminum foil (all sides but the top so the flash can still get through), but I won't know how well/ poorly this did until after a stack is done. Unless someone has a better way.

I also really don't know how esthetic or not the so called "specular highlights" are. I happened to like the original image I posted where I placed the foil along the inside of the container, but I guess the highlights weren't right. So it kind of makes it hard to proceed from an unknown esthetic perspective.

One thing I should point out, and this may be contributing to the difficulties with this tiny ant, and that is that the pictures I'm posting are crops. In the originals, the ant fills maybe 30-40% of the frame. From my birding days, unless the bird filled 60+ percent of frame (80% preferred), it was automatically tossed out as CCD's/CMOS chips have a much harder time with resolution with smaller input. I knew this, so was pleasantly surprised at the sharpness I saw when using the objective.... but being so small (not filling half frame or more) might have something to do with my lighting issues.

My flash has been hard on batteries, especially while running at these higher powers, so today I'm switching it over to a DC power supply. Initial tests, after a bunch of failures yesterday, have shown that I'll be able to use a spare PC supply I have here to power it from the 5V leg. Amazing how much current a flash needs, nothing else I tried in the voltage range would power it.

Anyway, thanks for your input, Dalantech, and I welcome suggestions on diffusion changes leading to a consistently workable system because, so far, I just haven't found it. This set up pretty much mimics Nikonuser's set up from a couple of years back and was suggested by one of the moderators. He used what looked like the same type of poly-foam cup, but maybe not.

[Dalantech]

...My flash has been hard on batteries, especially while running at these higher powers, so today I'm switching it over to a DC power supply. Initial tests, after a bunch of failures yesterday, have shown that I'll be able to use a spare PC supply I have here to power it from the 5V leg. Amazing how much current a flash needs, nothing else I tried in the voltage range would power it...

Because that cup is doing a better job of blocking the light from your flash than diffusing it, so you're firing it close to or at full power...

Instead of working on stacks why not just take some single frames, at about the same distance you're shooting at now, while you work on your diffusion? Might have to use a standard macro lens so you can stop the aperture down. Honestly the quality of your light will make a difference in the level of detail that you can record with the camera. It's actually easy to lose fine texture detail to the flash -easy to blow it out with strong specular highlights (too much micro contrast). So working on your lighting will improve the detail in your images.

Wrap you head around how the flash works, and how to tame it, and it will make a huge impact on the quality of your photos...

[ChrisR]

@ctron : Obviously the amount of light going through the cup depends on your polystyrene cup thickness & density, they vary. But the dominant illumination using this setup often comes from the reflected light inside the cup. It bounces around and comes at the subject from a wide angle.

[ctron]

One thing I didn't mention that could be relevant here is that I've been using nothing but jpegs for stacking, mainly for convenience because the program works a lot faster. At the same time though, raws are taken. I wonder if it would be worth it to stack with the raws instead?

[Lou Jost]

When I convert my RAW images to tiffs to use in Zerene, many artifacts related to dust, hot pixels, and tiny specular highlights go away. The jpg compression scheme often makes a one-pixel problem into a five- or 9-pixel problem.

[ctron]

Thanks, Lou, I may go back to the raw to tiff conversion at some point. With a lot of images, it takes a lot of extra time and space but if I have a halfway decent stack in jpeg, I may try it soon.

Well, my latest stack was a failure and I won't even post it. I'm getting better results with a small "killing jar" with acetone instead of the freezer as the bugs tend to remain in a more natural position in the former. However, the larger ant I used today, carpenter ant, was all shrivelled up by the time the stack was done. My guess is that it got in some of the acetone and was there for too long causing dehydration I suppose.

[Pizzazz]

Hi ctron


Lots going on with this thread, and I wanted to share an alternative to
several of your points.

First, the cup is also absorbing light, and as mentioned, the thickness
will have an effect. Try using a sheet of white paper, in an arch over the
subject. it might allow you to move the flash a little further away to gain
more control and introduce some shadows.

Second, when placing your specimen in the freezer, take it out after a
minute or two, then place it on a small piece of foam so you can spread the
legs out. The foam will allow the legs to remain in a more natural position
because the structures on the legs will grab the foam, then place it back in
the freezer to complete the process. Once out, you might want to allow it
to dry/firm up so when you remove it from the foam, the legs will remain in
place.

Mike

[ctron]

Hi ctron


Lots going on with this thread, and I wanted to share an alternative to
several of your points.


First, the cup is also absorbing light, and as mentioned, the thickness
will have an effect. Try using a sheet of white paper, in an arch over the
subject. it might allow you to move the flash a little further away to gain
more control and introduce some shadows.

Ok, I'll give it a try later today. I have to say though that I have never been more baffled by the use of flash. It seems like no matter what I do or which direction, I'm getting the so-called specular highlights and washed out details. However, that being said, I have not yet tried paper, so will give that a go.

Second, when placing your specimen in the freezer, take it out after a
minute or two, then place it on a small piece of foam so you can spread the
legs out. The foam will allow the legs to remain in a more natural position
because the structures on the legs will grab the foam, then place it back in
the freezer to complete the process. Once out, you might want to allow it
to dry/firm up so when you remove it from the foam, the legs will remain in
place.

Unfortunately, this didn't work with a carpenter ant I captured today. I tried both fast freezing and much more slowly. If I managed to get its legs to attach to the foam, the minute it got too cold, it would detach and curl up almost like a ball. If I again warmed it, it would eventually spring back to life. The process would be repeated several more times, but the same result each time.

I think one thing I overlooked yesterday was that I was using my wife's expired pill jar as a killing jar. I hadn't washed it out and there were still traces of the diuretic she takes. Maybe that's why the ant from yesterday shrivelled up like a prune.

Unless I can finally get some good results with flash, I'm very tempted to go back to constant lighting. The Ikea lights I tried just weren't bright enough after diffusing to get the exposure low enough for maximum sharpness. Now I have some 10 watt LED flood lights on the way that I read about in this forum. With those, apparently shutter is much better at 1/500 or better. It still wouldn't be as sharp as a flash result, but perhaps I can tease out more sharpness in post processing.

I'll keep working with flash in the meantime, but I have to say that my enthusiasm is quickly waning.

[ChrisR]

It may be worth trying ethyl acetate, or cherry laurel leaves, as ant killer. This is not from my experience, but literature suggests that these leave a specimen in a pliable state.

[Edit - what I meant really was that I haven't used anything else, other than EA, to compare it with. The relatively few bugs I've had a go with, have never really seemed particularly pliable!]

[rjlittlefield]

I use ethyl acetate often. It definitely leaves specimens pliable, not rigid. But "pliable" is not the same as "moldable". Although you can move the limbs into any position you want, they will go right back where they want as soon as they are no longer being held in place.

--Rik

[Chris S.]

Ctron,

With a single flash an a polystyrene cup, you have one of the best setups for studio lighting insect that you could ask for. However, you want to do essentially the opposite of what you're doing here: Instead of shining the light through the cup, shine the light into the cup and let it bounce around inside. So use no diffusion at all--just let the light bounce from a wide range of directions--and you should obtain nice, even light. Bounced light can be much easier than diffused light, and can produce better results.

• To do this:
  
  1) Cut a lengthwise portion out of the edge wall of the cup, and shine your flash into it.
  
  2) Importantly, block any portion of the flash from shining on the ant directly. Only light bouncing off the inner surface of the cup should hit the ant. You can use part of the polystyrene you removed to make the light blocking "flag."
  
  3) Since you'll no longer be shining your light through the polystyrene, you'll lose far less of it, so you can probably set your flash for at a lower power, which will improve recharge time and save batteries.

I would normally suggest shining the light from underneath the subject, but your subject support is already taking up that space. So you can put the open portion of the cup, and your flash, above the ant instead. But if you do this, photograph the ant upside down and rotate the final image 180 degrees. As you likely know, our eyes/brains treat the direction the light is coming from as "visual up," since we evolved in a world where the sun was above us. In this lighting arrangement, the side of the ant nearest the flash is actually the least lit, as the flash is blocked in that direction--so visual up is opposite the flash.

One problem I keep running into is when to stop taking images for the stack. I keep stopping a bit too soon and this makes the extreme end portion of the image blurry. I suppose I might be able to mark start/ stop areas on the course focus knob on the Nikon block(?).

I have a slice of masking tape on the fine focus wheel of my focus block, which helps me count revolutions.

Also, I try to shoot "through" the subject, including some extra shots to front and rear. Most of the time, I throw these away. But sometimes they save me from having to reshoot when I miss on one end or the other.

Further, some members of our community mount an iris as close as possible to the rear of the objective. Then at the ends of the stack, they stop down for a few pictures (which may be more widely spaced). The stopped down images have greater depth of field, though less resolution, and can provide a more pleasing transition from in-focus to out-of-focus elements.

I can't really go by the back of the camera histogram for stacking. If I try to keep highlights where they should be, the final stack is too dark. Seems like I have to almost clip the highlights, but the stack doesn't turn out overexposed. Not sure what's going on there.

If you're using the live-view histogram, don't trust it overmuch, as in my experience they aren't all that accurate. If you're shooting an image and then checking the histogram, this can be much more accurate. Make sure you are using an RGB histogram, and not single-channel or luminosity histogram. Also, at magnifications like these, small highlights go in and out of focus shot by shot. When such a highlight is out of focus, the bright bits are averaged with less-bright bits, so OOF highlights appear darker than they will be when in focus. I make it a point to set my exposure with an in-focus frame, and test it on a few frames with different points of focus.

Not sure if any of the above answered your questions?

Cheers,

--Chris

--edited typo