My venerable Olympus E330 stopped working after many years so I picked up an Olympus E5mII. Naturally, the first thing to try was the built in focus bracketing in a few panoramas.
The lens used was the 60mm macro. All were shot in natural light using aperture priority mode. Magnification was 1:1 (or more correctly, the angle of view for each stacked frame was the same as the angle of view at 1:1)
Petunia
83mp
960 images
Print size: 33"x30"
Grass
124mp
990 images
Print size: 58"x25"
Begonia
103mp
1850 images
Print size: 36"x33"
p.s. I'm experimenting with using the Prophoto color space, so the colors in these photos may seem flat in some browsers.
Several panoramas
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Great! I just got this same camera and lens. Silent mode is REALLY silent, nothing moves except the iris closing down, and there is even a way of locking the iris at its stopped-down position so it doesn't move either. (Not sure if this works when focus bracketing is enabled...)
I hope you post more about your experiments with ProPhoto color space. AdobeRGB is definitely limiting the saturation of some golden-yellow tones in my pictures (and these tones are common in flowers). I can see the limitation when I set a wide-gamut monitor to emulate AdobeRGB.
I hope you post more about your experiments with ProPhoto color space. AdobeRGB is definitely limiting the saturation of some golden-yellow tones in my pictures (and these tones are common in flowers). I can see the limitation when I set a wide-gamut monitor to emulate AdobeRGB.
- rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Really nice work, elf!
A couple of technical things...
1. I don't understand what this sentence means: "Magnification was 1:1 (or more correctly, the angle of view for each stacked frame was the same as the angle of view at 1:1)". I understand "angle of view" to be, well, an angle, like 25 degrees. That's independent of magnification, so it doesn't tell me what the subject size is. Thinking hard, I can guess that you mean the front of the stack was 1:1, but then I'm not sure exactly what "1:1" means. Some people use "1:1" to mean a 24x36 mm frame, i.e., 1X on "full-frame" format (itself a pretty vague term when you think about it). But I don't know if that's what you mean, or if you're talking about 1:1 with respect to the small sensor of the E5mII. Knowing a little about how you work, I'd guess the latter, but I'm not confident.
Can you clarify, please?
2. As far as I can tell, only the second two images are ProPhoto; the first one is still sRGB. Does it look the same on your end?
--Rik
A couple of technical things...
1. I don't understand what this sentence means: "Magnification was 1:1 (or more correctly, the angle of view for each stacked frame was the same as the angle of view at 1:1)". I understand "angle of view" to be, well, an angle, like 25 degrees. That's independent of magnification, so it doesn't tell me what the subject size is. Thinking hard, I can guess that you mean the front of the stack was 1:1, but then I'm not sure exactly what "1:1" means. Some people use "1:1" to mean a 24x36 mm frame, i.e., 1X on "full-frame" format (itself a pretty vague term when you think about it). But I don't know if that's what you mean, or if you're talking about 1:1 with respect to the small sensor of the E5mII. Knowing a little about how you work, I'd guess the latter, but I'm not confident.
Can you clarify, please?
2. As far as I can tell, only the second two images are ProPhoto; the first one is still sRGB. Does it look the same on your end?
--Rik
Lou:
The main reason for using the Prophoto color space is that it is more forgiving when editing. Most monitors and printers can't use the full gamut, so outputing in Prophoto isn't as important. I didn't try to manipulate the colors in these images, so there are out of gamut areas. The little prints I've made of these look quite nice in spite of the out of gamut issues.
Rik:
It's hard to describe the magnification for a focus stack. These images all start at 1x or 1:1 (at the sensor), but I don't know what the actual magnification is at the rear of the stack. In retrospect, angle of view suffers from the same problem in that the angle of view of the first image is not the same as the last.
There isn't very much documentation around Olympus' implementation of focus bracketing. There are 10 levels of DOF, but nothing to say how wide each level is or how much it changes for each aperture setting. I need to do some more rigorous testing to find the best settings for a particular image. The focus bracketing doesn't work well with extensions as the individual steps are too small. This limits the usage to 1:1.
The first image was stacked and stitched from jpegs and the others from RAW, so yes, the first is sRGB and the others Prophoto.
One of the issues with in camera focus bracketing is the exposure changes the same as if you were using bellows draw to change the focus point. Each focus bracketing sequence uses the same exposure settings as the first image. For really deep stacks this exposure difference can be quite large. I used aperture priority mode in these images and set the number of exposures to 20 or 30. The camera remembers the last focus point, so continuing the stack is just a matter of triggering the shutter again. This also resets the exposure, so overall the exposure is fairly consistent.
The main reason for using the Prophoto color space is that it is more forgiving when editing. Most monitors and printers can't use the full gamut, so outputing in Prophoto isn't as important. I didn't try to manipulate the colors in these images, so there are out of gamut areas. The little prints I've made of these look quite nice in spite of the out of gamut issues.
Rik:
It's hard to describe the magnification for a focus stack. These images all start at 1x or 1:1 (at the sensor), but I don't know what the actual magnification is at the rear of the stack. In retrospect, angle of view suffers from the same problem in that the angle of view of the first image is not the same as the last.
There isn't very much documentation around Olympus' implementation of focus bracketing. There are 10 levels of DOF, but nothing to say how wide each level is or how much it changes for each aperture setting. I need to do some more rigorous testing to find the best settings for a particular image. The focus bracketing doesn't work well with extensions as the individual steps are too small. This limits the usage to 1:1.
The first image was stacked and stitched from jpegs and the others from RAW, so yes, the first is sRGB and the others Prophoto.
One of the issues with in camera focus bracketing is the exposure changes the same as if you were using bellows draw to change the focus point. Each focus bracketing sequence uses the same exposure settings as the first image. For really deep stacks this exposure difference can be quite large. I used aperture priority mode in these images and set the number of exposures to 20 or 30. The camera remembers the last focus point, so continuing the stack is just a matter of triggering the shutter again. This also resets the exposure, so overall the exposure is fairly consistent.
- rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
That's a good question. I'm not sure I can answer it as I've only either printed small (like 8"x10") or large (like 30"x60"). Actually even the large prints were still only 1/4th what the full size print would have been.
An interesting thing is most people tend to stand back to view the large prints even when encouraged to get close. If you can control the viewing distance, then single image captures can be printed quite large and shown effectively.
One of the benefits of shooting a pano, then downsizing to print, is very little, if any, sharpening is needed.
An interesting thing is most people tend to stand back to view the large prints even when encouraged to get close. If you can control the viewing distance, then single image captures can be printed quite large and shown effectively.
One of the benefits of shooting a pano, then downsizing to print, is very little, if any, sharpening is needed.
- rjlittlefield
- Site Admin
- Posts: 23625
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
I get this as being 300 pixels per inch, so I guess your "full size print" would be 150 ppi?Begonia
103mp
1850 images
Print size: 36"x33"
That sounds consistent with my experience about what's needed for a big print to look good up close.
That is interesting.An interesting thing is most people tend to stand back to view the large prints even when encouraged to get close.
Max Lyons, who made the first gigapixel panorama (of Bryce Canyon), reported exactly the opposite. His experience, at the PMA Spring 2004 show where a suitably huge print was displayed, was that people would approach as needed to see the finest details.
I can think of a lot of aspects that might explain the difference. Gigapixel images were brand new at the time, so viewers might have simply wanted to experience the unique capability. That might be amplified because of the audience, PMA tending to attract tool-and-technique people.
But I wonder if the nature of the image is also part of the difference.
Bryce Canyon is a big place, so even the finest details appearing in Lyons' image show things that are human scale or larger. One of the actual-pixel crops in the above link shows an entire person only about 50 pixels high! As a result, no matter how close you looked at the Bryce Canyon print, you would still be seeing things that you're familiar with.
In contrast, the begonia is a small subject, and the details shown in a 36x33 inch print would be things that you would normally see only through a 10X-20X microscope. I imagine that those of us who are used to looking through microscopes would find much to catch and keep our attention at that scale. But I would not be surprised to see that most other people would be more attracted by the composition than by the fine details, and that those people would just back up to a "normal" viewing distance.
Any thoughts along those lines? Have you asked any of your viewers about it?
--Rik
Max's gigapixel image was what started me on panoramas and I think you're right about the subject matter making a difference in how people view the images. On the other hand a 36"x36" image is pretty small compared to Max's image, so that could also be part of it. I've never asked anyone why they didn't get closer. When I encouraged them to do so, they would only spend a short time there before moving away.
I think I'll have to start looking for some subject matter for a gigapixel macro. Now that my camera has EFSC, one image won't wear out the shutter
I think I'll have to start looking for some subject matter for a gigapixel macro. Now that my camera has EFSC, one image won't wear out the shutter