Sweat Bees Foraging in Sourgrass Flowers

[Dalantech]

Sourgrass is a seasonal plant, usually showing up in my yard around November and by June it will be completely dormant, so it seems to like cooler weather and shade. The flowers are a major source of nectar for honeybees and most solitary bees, and a source of pollen for smaller species of solitary bees like Sweat Bees. I like the challenge of photographing the female Sweat Bees as they forage for pollen. They'll use their mandibles to strip pollen out of a Sourgrass anther and then push it down to their abdomen with their legs.

There are two challenges to photographing them when they are hyperactive. The first is being able to grab onto the flower they are foraging in undetected. It's easier to do on windy days cause the vibration that I induce when I grab onto the flower's stem is masked by the breeze. In this video I describe the technique and end up spooking the subject because there wasn't enough wind.

The second challenge is to compose the scene and get the area of acceptable focus where it needs to be. I don't focus stack, and it really doesn't matter cause it's not possible to focus stack a subject that's in motion anyway. They can take up to a minute or two to go through all the anthers in a flower, but it's only the anthers at the top that allows for a good view of the bee. So once she comes to the top of the flower I only have a few seconds, and even then there won't be many opportunities to catch her looking at the camera. If I can work out the lighting it would be an easier scene to shoot a video of than to take stills.

Tech Specs: Canon 80D (F11, 1/250, ISO 100) + a Canon MP-E 65mm macro lens (around 2.5x to 3x) + a diffused MT-26EX-RT with a Kaiser adjustable flash shoe on the "A" head (the key), E-TTL metering, -1/3 FEC). These are single, uncropped, frames taken hand held. In post I used Topaz Sharpen AI, Denoise AI, and Clarity in that order.

Sweat Bee Foraging in a Sourgrass Flower II by John Kimbler, on Flickr

Sweat Bee Foraging in a Sourgrass Flower by John Kimbler, on Flickr

Sweat Bee Foraging in a Sourgrass Flower III by John Kimbler, on Flickr

I usually end up with more images like this next one, cause the critter noticed me when I grabbed onto the flower's stem. Sometimes they'll pause at the edge for a fer seconds before doing their normal pre-flight cleaning and take off. Very common for the bee to be facing away from me, but since I've got the stem of the flower between my index finger and thumb I just gently rotate it until I get the subject where I want it in the frame.

Sweat Bee in a Sourgrass Flower by John Kimbler, on Flickr

[Beatsy]

These are beautiful. Thanks.

Interesting that you denoise after sharpen. I do all denoise right up front - even before levels adjustments. Am I missing a trick?

[Dalantech]

These are beautiful. Thanks.

Thank you!

Interesting that you denoise after sharpen. I do all denoise right up front - even before levels adjustments. Am I missing a trick?

I've tried stripping out noise first, but with my diffraction limited images sharpening first seems to preserve more detail. Both plugins sharpen and remove noise (just they're better at one than the other). A lot of people who use the same plugins are surprised that I'm using both of them. I normally stick with the settings that the plugins recommend after they analyze my images, only applying more sharpening when I'm shooting above 4x.

[Beatsy]

Ahh, I see ref denoise. I was wondering about the f/11. I see diffraction kicking in around f/5.6 at 2.5x when using the A7r4 and tend to shoot quite a bit wider as a result. Rarely over f/8, usually 6.3 and 4-5.6 if the subject/pose will stand it. Guess that's why 'denoise first' works better for me. I turn noise reduction off in sharpen.

[Dalantech]

Ahh, I see ref denoise. I was wondering about the f/11. I see diffraction kicking in around f/5.6 at 2.5x when using the A7r4 and tend to shoot quite a bit wider as a result. Rarely over f/8, usually 6.3 and 4-5.6 if the subject/pose will stand it.

I don't mind trading some fine texture detail for a little more depth, so that's why I do most of my shooting at F11. The MP-E 65mm has a floating lens group that adjusts the focus as the mag changes and it seems to work really well (sharper image means that diffraction effects get "amplified" less). I've noticed a difference in detail between F14 and F16 with that lens and I don't think that it can be attributed to just diffraction.

Maybe someday there will be a camera that can fire fast enough for a limited stack of a subject in motion, but until then I'll stick to single frames.

Guess that's why 'denoise first' works better for me. I turn noise reduction off in sharpen.

I need to experiment with Sharpen AI and Denoise AI more. Post processing is definitely my weakest aspect of photography at this point.

[Beatsy]

...Maybe someday there will be a camera that can fire fast enough for a limited stack of a subject in motion...

120fps video

[Dalantech]

...Maybe someday there will be a camera that can fire fast enough for a limited stack of a subject in motion...

120fps video

Wouldn't that be the same as shooting with a 1/120 shutter speed?

[Beatsy]

...Maybe someday there will be a camera that can fire fast enough for a limited stack of a subject in motion...

120fps video

Wouldn't that be the same as shooting with a 1/120 shutter speed?

Or faster, yes. Go 60fps or 24fps if that works. But you can crank the ISO if you need slower (more light) cos Topaz Denoise has your back

[Dalantech]

...Maybe someday there will be a camera that can fire fast enough for a limited stack of a subject in motion...

120fps video

Wouldn't that be the same as shooting with a 1/120 shutter speed?

Or faster, yes. Go 60fps or 24fps if that works. But you can crank the ISO if you need slower (more light) cos Topaz Denoise has your back

120 FPS would be one frame every 1/120 of a second. The problem isn't the exposure, the problem is the effective shutter speed. Even 240 FPS wouldn't have been fast enough for the first three shots that I posted in this thread. Those Sweat Bees were moving -I wasn't shooting them when they paused. I miss the focus a lot because of the lag between pressing the shutter and the subject changing position before the shutter trips. I think it would take somewhere in the 1000 FPS or higher range, and even then subject motion might be a problem from some angles.

Here's another example: Caterpillars feed by extending their heads and then curling up as they chew. So the trick was to focus right at the edge of the leaf and wait for the critter to bring it's head up to start the next "row". I got seven frames before it figured out I was close and stopped feeding. Shot horizontally but framed for a vertical composition (I turned the shot 90 degrees in post). Easier than trying to hold the camera in portrait orientation.

Tech Specs: Canon 80D (F11, 1/250, ISO 100) + a Canon MP-E 65mm macro lens (over 3x) + a diffused MT-26EX-RT with a Kaiser adjustable flash shoe on the "A" head (the key), E-TTL metering, -1/3 FEC, second curtain sync). This is a single, uncropped, frame taken hand held. I'm holding on to the stem of the plant with my left hand, and resting the lens on that same hand to keep the scene steady.

Caterpillar Feeding on Mint by John Kimbler, on Flickr

I need to re-edit that one cause the white balance is off (too warm).

Not only would the frame rate have to be ridiculously high, but the focus shift would probably have to occur in camera. The tech just doesn't exist to take that kind of focus stack -if it did I'd use it. Would be a couple of orders of magnitude easier than manually timing single frames of a moving subject.

[MarkSturtevant]

Very nice pictures. The paradigm of course is to denoise first and then sharpen. But I can't knock great results. I expect that dedicated software for bringing down noise can do a better job than what I have.
On a tangential subject, I have been researching the mirrorless micro-4/3 cameras from Olympus and Panasonic. They both have a feature where they rapidly take pictures when the shutter is pressed down halfway, dropping pictures off of the rear end of the buffer. When you see a moment that is picture-worthy, you take the picture. What is recorded is that moment, plus a lot of frames from just before. This increases the chances that you get the perfect shot.

[Dalantech]

Very nice pictures. The paradigm of course is to denoise first and then sharpen. But I can't knock great results. I expect that dedicated software for bringing down noise can do a better job than what I have.

Thanks! Both plugins sharpen and remove noise. I could almost get away with just running Sharpen AI, but using the bare minimum settings in Denoise AI after I use Sharpen AI cleans up my images really well.

On a tangential subject, I have been researching the mirrorless micro-4/3 cameras from Olympus and Panasonic. They both have a feature where they rapidly take pictures when the shutter is pressed down halfway, dropping pictures off of the rear end of the buffer. When you see a moment that is picture-worthy, you take the picture. What is recorded is that moment, plus a lot of frames from just before. This increases the chances that you get the perfect shot.

Might be a good feature for someone shooting an event, but the timing wouldn't be fast enough for what I'm ding. If that camera can take 20 frames per second then that's a 1/20 virtual shutter. Too slow...

[MarkSturtevant]

I don't understand how fps is the same thing as shutter speed. It is academic in any case as what you are doing is clearly working and I see no need to change it, but for example in the burst mode that I mentioned for the Olympus cameras, which they call Pro-capture, you can use a higher speed burst up to 60fps but it is single focus (the subject would hopefully be moved into focus while the burst of pictures are taken), and they have a lower speed burst mode of 18fps, but with continual autofocus.
These are the # of pictures taken per second. Then there is the shutter speed, which seems different to me. I haven't discovered if there are limitations to shutter speed while shooting a burst of pictures with these cameras, but the specs describe very high shutters speeds like 1/8000 sec for mechanical, 1/32000 for electronic. These cameras are pretty good at freezing movement! I am talking full raw pictures. Not video.
Of course lighting would be a consideration, such as continuous LED lighting. Finally, the smaller sensor sizes mean there is a greater intrinsic depth of focus. One can get more light by shooting at wider apertures.

[Dalantech]

I don't understand how fps is the same thing as shutter speed.

What would be the difference in shooting 60 stills at 1/60 of a second or shooting video at 60 FPS? Seems to me that either way you'd still have 60 frames per second, so the video speed acts as a virtual shutter speed...

If there is any movement in that second, and it doesn't matter if you're shooting video or stills, it's going to be recorded.

[Pau]

Mark, I agree with what you said except:

Finally, the smaller sensor sizes mean there is a greater intrinsic depth of focus.

Not really if you shot at the same effective aperture and magnification at the final image. Rik has explained it several times much better than I could.

What would be the difference in shooting 60 stills at 1/60 of a second or shooting video at 60 FPS? Seems to me that either way you'd still have 60 frames per second, so the video speed acts as a virtual shutter speed..

Nope, FPS of course will limit the exposure time, 1/60s will be the theoretical MAXIMUM exposure time at 60 FPS but nothing prevents to use shorter exposure times like 1/1000s

John, excellent pictures, BTW

[Dalantech]

What would be the difference in shooting 60 stills at 1/60 of a second or shooting video at 60 FPS? Seems to me that either way you'd still have 60 frames per second, so the video speed acts as a virtual shutter speed..

Nope, FPS of course will limit the exposure time, 1/60s will be the theoretical MAXIMUM exposure time at 60 FPS but nothing prevents to use shorter exposure times like 1/1000s

John, excellent pictures, BTW

Thanks!

Note: I've edited my response below cause I didn't explain it well enough. But my conclusion remains unchanged.

Don't think of it in terms of exposure, but in frame rate. If I pull two stills out of a 60 FPS video, and take a single frame at 1/60 of a second, the amount of time the subject can move will be the same, either between the video frames (1/60 of a second) or during the single frame exposure (1/60 of a second). So in both cases my "stopping power" is still 1/60 of a second. Due to differences is potential shutter speed between the video and the still shot a single video frame could be sharper (cause the speed of the shutter could be faster). But if you're trying to create a sequence of frames for a stacked image using video then the capture rate of that video becomes an issue if there is too much subject movement between frames.