Handheld test: wild rose/bramble

[DanK]

I went for a walk yesterday to experiment with doing macro without a tripod using my new R6II with an EF 100mm macro. Macro is a very difficult challenge for IS, so I wanted to see whether the much better IS in this camera, relative to my old one, would let me do without a tripod.

The answer is "sort of". I couldn't go to very slow speeds, but I think the problem wasn't IS; it was my own motion. Even slight motion forward or back ruins a shot because of the very narrow DOF. So as a result, I found that I rarely could go below 1/90.

The shot below is a wild plant in the rose family, but I couldn't identify it as to species. It was shot at 1/125, f.8, ISO 1250. Given that I wasn't focus stacking (which I would have done with a tripod), I should have gone narrower than f/8, as you can see from some areas in the petals. Still, this leaves me optimistic that in better lighting, I can manage sometimes without support.

Comments welcome, as always.

[Manfred M]

No real surprises here Dan. IS has its limitations and the moment the camera movement exceeds the IS system's availability to compensate for that movement motion blur shows up. It looks like you have shown the limits of this tool. It's hard to beat what you can do with a tripod...

[DanK]

The issue is somewhat different for macro, for several reasons. One is that the motions you have to worry about are not entirely the same as those at issue with conventional IS. This difference has been reduced with 5-axis IBIS, but there is still the issue of motion back and forth. Given the amount of motion involved, IBIS cant manage that.

So, I had no idea what numbers to expect. Doing informal handheld shots around the house, I was able to get some sharp shots down to 1/6 second with a FL (if I remember right) of 105mm. I knew I wouldn't be able to approach this in doing macro, but I was a bit surprised and certainly disappointed by the extent of the difference. It took me a number of shots before I was able to piece together that forward and backward motion was a big part of the problem.

Still, I think this will give me tripod-free options I didn't have before. And since part of my motivation in upgrading was to be able to do without a tripod more to save weight, this is important to me.

I've only just started to experiment to see when in-camera focus bracketing is effective without a tripod. Certainly not for most macro, I suspect.

[Geoff F]

Sometimes I have to abandon my tripod to chase an insect which is flitting around in the brambles etc, which is where IS becomes desirable.

Lack of focus depth seems to be the chief problem with this image. Focus stacking which enabled you to get that left side bud in focus would have produced an acceptable result.

[DanK]

I think my original post was too vague. Here's the background:

Traditional in-lens IS adjusts for angular displacement (yaw and pitch) as well as roll. The problem is that as you get into macro distances, these movements matter less and less, and horizontal and vertical displacement of the sensor matter more and more. 5-axis IS adds compensation for horizontal and vertical displacement, labeled X and Y in this video from Sony: https://youtu.be/svbUXedWsbA.

If I correctly understood an email exchange with a Canon tech years ago, the "hybrid IS" in one or more Canon macro lenses is in fact 5-axis IS. He didn't use the terms, but he said I was correct that hybrid IS adds compensation for horizontal and vertical displacement of the sensor. I found that this hybrid IS did give me a few stops of improvement at macro distances.

Now I have a camera that has 5-axis IS that coordinates automatically with my in-lens IS, in this case, also 5-axis. Under some circumstances, this can yield 8 stops of stabilization. My question was what the practical impact would be in macro work.

The answer, if I can base it on only 30-some shots of two subjects, is that IS isn't the bottleneck. The limitation is motion forward and backward, which no IS (AFAIK) adjusts for. So, I can't really test how slow a shutter speed the IS will allow under these circumstances because forward and backward motion introduces blur before I get to the limit IS would impose.

It's easy to see why this is an issue. To take a random example, suppose you focus a FF camera with a 100mm lens at a distance of 3m. DOF is roughly 16 cm, so a movement of, say, 2 or 3 mm backward or forward wouldn't have a visible impact. Not so in the case of a macro image. Keep everything the same but focus on a flower 40 cm away. DOF in that case is roughly 1mm. So that's the limiting factor, not the 5 types of motion IBIS compensates for.

[Chataignier]

Interesting discussion. I've never had much faith in IS for macro work, though without any effort to study the technicalities, this was just based on experience.
Like Dan I concluded that back/forth is the main problem, sometimes due to wind moving the subject, but just as often, me swaying a mm or two when working without a tripod (which I mostly am, because chasing flying insects with a tripod is recipe for exteeeeeme frustration). One trick that works when the light is good enough to have fast shutter speeds is to fix the focus manually, then deliberately move slowly towards the subject while burst firing. My Fuji X-T3 allows me 20/s with the electronic shutter.
The other thing I use is flash. I dont like the head-on light from a ring flash, so I've built a rig that holds my cobra flash above the lens (also a Canon 100mm macro) with the head angled down at around 45° both to the side and above. It does not help with problems of focus due to movement, but the very short flash duration does freeze both subject and lateral camera movement even better than IS and used with auto focus yields good results around 30% of the time, which is quite good for hand held macro.

[DanK]

David,

You and I have ended up in a similar place, but I'm wondering whether my new camera will change that.

My standard approach has been to balance the camera on a monopod with a tilt head, focus at about the distance I want, rock back and forth slightly to achieve focus, and try to trip the shutter at the right instant. Needless to say, I usually fail. I use a highly diffused flash angled over the front of the lens. However, I've been doing this with single shots, separated by a fair amount because of the recharge time of the flash.

The new camera has a fast burst rate (12 fps mechanical, 40 fps electronic), and it has a "raw burst mode" that will take a burst of about 1 second and put the images in a separate virtual roll when uploaded. One option, which is also present on the OM-1, is to have the camera start storing and discarding images as long as the shutter is half depressed; it then keeps roughly half a second before and after the shutter is fully depressed. However, this can only be used with available light, of course. Also, my new camera has much lower pixel density than the old 7D I keep for macro, so I would have to get closer and use longer tubes to approach the same level of detail. Flowers are less of an issue because I usually opt for lower magnification than I do with bugs.

Dan

[William W]

I know the EF 100mm f/2.8L Macro IS USM.

Were you bending/leaning in when you took your 30-some shots of two subjects?

I believe stance is one of the important variables in this experiment.

WW

[DanK]

Stnadinfpg for some of the tests but leaning for this one.

Stance is indeed important, but I think that in general, shots like this need support, a faster shutter, or flash.


Sent from my iPad using Tapatalk

[Ken MT]

Stnadinfpg for some of the tests but leaning for this one.

Stance is indeed important, but I think that in general, shots like this need support, a faster shutter, or flash.


Sent from my iPad using Tapatalk

That would not increase the DOF, only a smaller aperture would

[DanK]

That would not increase the DOF, only a smaller aperture would

Sorry, I was too brief. Yes, you're right, those changes wouldn't change DOF. What they would change is camera motion front to back, allowing the focus point to remain where it should be, within the narrow band of good focus. I assume Bill's reason for asking about stance is that bending makes it harder to maintain position, making it more likely that the camera will move and cause the subject to go out of focus. That's what I was responding to. A longer version of my response would have been:

"Stance is indeed important, but I think that in general, shots like this need support, a faster shutter, or flash in order to keep the subject in focus during the time of exposure.

One of the characteristics of field macro is that subjects are often in very inconvenient locations, and you have to get the lens within a few cm of them. That often requires contortions. And even when that's not the case, even standing, I have difficulty maintaining that precise a position. Perhaps for younger or more coordinated people, merely standing rather than crouching would be sufficient, but I'm not among them unless the shutter speed is very fast or I have flash to freeze motion.

[revi]

And you often have to deal with subject motion (wind). No IS is going to help with that.

For the non-macro shooters: at "normal" focusing distances, changing the distance to the subject by 0.5mm or so doesn't matter much. In close-up and macro, 0.5mm can be more than the DoF you have. And IS doesn't help there (that would be a 6th, Z, axis)

[DanK]

For the non-macro shooters: at "normal" focusing distances, changing the distance to the subject by 0.5mm or so doesn't matter much. In close-up and macro, 0.5mm can be more than the DoF you have. And IS doesn't help there (that would be a 6th, Z, axis)

I think that's exactly right.