Image quality under three lighting setups

[papilio]

Dicranocephalus wallichi bourgoini scarab

EFSC continuous light focus-stack, 5X through CFI60, Canon 6D body
Full-resolution image is HERE.


Not too long ago I switched from using a speedlight to an AlienBees studio strobe with a 16-inch beauty dish diffuser, as I needed the faster burst rate of the strobe in order to shoot stacks of live subjects as quickly as possible. After having used the studio strobe for some time and comparing those images with ones which I had taken with the speedlight I noticed that the strobe images were surprisingly soft. After some experimentation it became clear that the 1/1000-second burst from the strobe (as compared to ~1/10,000 second with the speedlight) was not effectively freezing camera vibrations.

The three full-resolution crops below illustrate the definition and image quality of the two lighting setups described above as well as a third image shot with the Canon 6D's Electronic First Shutter Curtain feature and continuous illumination, provided by two Ikea LEDs poorly diffused by a layer of toilet paper. All images are unsharpened.

Studio strobe image


Speedlight image


EFSC image
[size=0]
[/size]

This EFSC image really took me by surprise. Not only is definition significantly improved over even the image shot under the fast burst of the speedlight, but the nature of the light and the subtlety of the structures beneath the membrane are strikingly different. In my opinion the image is really quite lovely in comparison with any other stack which I've done to date.

Since I had modified my setup considerably in order to reduce the vibrations of the rig I believe that, for the first time in my work, environmental vibrations dominated. While I have not determined the source of them they caused significant softening of the image under the continuous lighting in my initial shots taken with a 1/200-second exposure time. I increased the ISO and reduced the shutter time to 1/1000 second, and the softness which persisted seems to confirm that this exposure duration, which approximately matches that of the studio strobe burst, does indeed fail to freeze vibration.

Only when I increased the ISO up to 1600 in order to allow exposures of 1/2500 second did the image begin to approach pixel-level definition, something which I've not previously obtained.

While my live subjects do not allow me to use continuous lighting and therefore limit me to flash illumination, which is currently incompatible with the EFSC technique, it became clear to me what a wonderful method this is when shooting stacks of dried specimens. Thank you Charles Krebs!

[ChrisR]

Errm, the EFSC one looks least sharp to me.

[papilio]

Errm, the EFSC one looks least sharp to me.

Hmmm ... Now that you mention it Chris I do see what you're referring to, and in fact you may be correct. But to my eye overall clarity of the structure is improved. My interpretation of the images is, I think, that the flash images are more harsh rather than sharper, but that may be just me.

I wonder whether each person's eye/brain interprets different aspects of an image in it's evaluation?

For instance. take a look especially at the left sides of the images.

I'll be quite interested in seeing what a general consensus finds in these images.

[papilio]

Oh dear, Chris! Looking at the original outputs of ZS it appears that I've mistakenly posted the sharpened versions of the first two images!

Give me an hour or so to get new images uploaded to flickr and replaced in the post, I'd be interested in whether or not your impression is the same.


[EDIT] Page revised, photos replaced. Sorry!

[ChrisR]

Do you think the different parts look different because they have transparent surface layers? If you look up around 1 o'clock the efsc one is still off, but as you say, around 8-9 it's clearer.
If you moved the light about, I'd guess the sharp areas would move as well - ?

Recently I've been photographing things in plastic bags, a nightmare! For 6 shots in a row, all with flash bounced off something big, even avoiding reflections it's just cloudy. Then one comes clear, and it's hard to say why, and the next one's different.

[papilio]

Do you think the different parts look different because they have transparent surface layers? If you look up around 1 o'clock the efsc one is still off, but as you say, around 8-9 it's clearer.
If you moved the light about, I'd guess the sharp areas would move as well - ?

Recently I've been photographing things in plastic bags, a nightmare! For 6 shots in a row, all with flash bounced off something big, even avoiding reflections it's just cloudy. Then one comes clear, and it's hard to say why, and the next one's different.

As Rik has mentioned, the world of photomicroscopy has strange happenings sometimes.

This thing really has me puzzled! When I look at 1 o'clock I see greater detail, for instance double tips on some of the receptors ... but perhaps that's false detail.

I'll probably shoot the stack again, moving the lighting as you suggest, and see what happens.

Thanks for taking a second look, Chris!

[Chris S.]

Michael,

I've also been wondering whether the double tips on some of the receptors are real or false. It will be interesting to see what you determine.

Something I'm really puzzled about is why there is so much less purple fringing (which I take to be chromatic aberration) in the EFSC stack. One possible explanation might be that the EFSC image has enough additional detail that this detail out-competes the purple fringing during the stacking process. In this case, the reduction in CA seems greater than the increased detail would account for--at least according to my gut feel. But I can't think of any other reason for this reduction.

If you check the individual, unstacked, input images of the EFSC and one of the other stacks, is CA equally present in both?

Cheers,

--Chris S.