New Lens tryout.
Moderators: rjlittlefield, ChrisR, Chris S., Pau
New Lens tryout.
I recently won an Ebay bid for a used Leitz 42mm Summar which is designed for a macro range of 3x-15x. Having ripped open the package as soon as the postman arrived, I had to take my first image stacks with it to test it out. This single flower of a Crassula 'Morgan's Beauty' was shot with about 240mm of extension. The vertical visual field is about 5/32nds of an inch or 8mm. I calculated this to be approximately 5x lifesize, the same as my Canon MP-E at maximum. This is a stack of 35 slices. I was very pleased with the quality of the lens. I did check the individual slices. The rim around the petals are not digital artifacts, the petals were translucent right by the edge in the individual image. Hope you like the image. Let me know what you think of the sharpness of the lens.
Irwin
Irwin
- Bruce Williams
- Posts: 1120
- Joined: Mon Oct 30, 2006 1:41 pm
- Location: Northamptonshire, England
- Contact:
Irwin,
I notice that the JPG file is only 36,790 bytes which IMO is VERY small for a 533x800 image and you may well be losing important detail due to such a high level of compaction. The 2 main factors affecting JPG filesize are the amount (area) of detail* in the image and the level of jpg compaction applied. I suggest you take a look at compaction levels (pre and post stacking) and select the best quality that will still keep you within forum limits of 200Kb.
*area and type of detail makes a big difference to jpg filesize, even straight out of camera. So for example (typically) a photo of a shrub in leaf/flower will produce a much larger jpg filesize than a close-up photo of a car. So by implication a filesize of 36,790 Kb suggests a low level of detail in a subject that fills the frame.
Bruce
I notice that the JPG file is only 36,790 bytes which IMO is VERY small for a 533x800 image and you may well be losing important detail due to such a high level of compaction. The 2 main factors affecting JPG filesize are the amount (area) of detail* in the image and the level of jpg compaction applied. I suggest you take a look at compaction levels (pre and post stacking) and select the best quality that will still keep you within forum limits of 200Kb.
*area and type of detail makes a big difference to jpg filesize, even straight out of camera. So for example (typically) a photo of a shrub in leaf/flower will produce a much larger jpg filesize than a close-up photo of a car. So by implication a filesize of 36,790 Kb suggests a low level of detail in a subject that fills the frame.
Bruce
- Bruce Williams
- Posts: 1120
- Joined: Mon Oct 30, 2006 1:41 pm
- Location: Northamptonshire, England
- Contact:
Irwin,
Yes the second image looks better.
The following procedure is very unscientific, but it's quick and easy.
I loaded both images into Photoshop CS2 (as 2 separate layers). Although registration was not exact, toggling between the two images clearly showed a very noticable increase in detail in your second image.
Out of interest, you can get a good idea just how much better it is, as well as how much hidden detail is still contained within each image, by reducing (levels) gamma to around 0.35 and looking at the RGB image and at the red channels.
Bruce
Yes the second image looks better.
The following procedure is very unscientific, but it's quick and easy.
I loaded both images into Photoshop CS2 (as 2 separate layers). Although registration was not exact, toggling between the two images clearly showed a very noticable increase in detail in your second image.
Out of interest, you can get a good idea just how much better it is, as well as how much hidden detail is still contained within each image, by reducing (levels) gamma to around 0.35 and looking at the RGB image and at the red channels.
Bruce
- rjlittlefield
- Site Admin
- Posts: 24017
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Irwin,
I don't know what the small details on this flower actually look like, but if these are full-frame images, they definitely look softer than I'd expect.
Two main issues come to mind:
1. What f/stop were these shot at? At high magnification, quality can fall off quite quickly as you stop down. (See this topic, second pic, first column, for Olympus 38mm f/2.8 lens.) If you haven't already done so, I suggest running a quick series to see at what aperture your new lens is sharpest.
2. There are some large areas that look unnaturally smooth with no visible details, for example at image center, just above and below the edge of the center petal in front of the stamens. Just visually, those areas look like the same sort of "mush" that phero66 reported from Helicon Focus in this topic. Are those areas featureless in all of the individual frames too, or have low contrast details been mushed out in the stacking? If you have lost detail in the stacking, there are several suggestions in phero66's topic for how to get it back.
--Rik
I don't know what the small details on this flower actually look like, but if these are full-frame images, they definitely look softer than I'd expect.
Two main issues come to mind:
1. What f/stop were these shot at? At high magnification, quality can fall off quite quickly as you stop down. (See this topic, second pic, first column, for Olympus 38mm f/2.8 lens.) If you haven't already done so, I suggest running a quick series to see at what aperture your new lens is sharpest.
2. There are some large areas that look unnaturally smooth with no visible details, for example at image center, just above and below the edge of the center petal in front of the stamens. Just visually, those areas look like the same sort of "mush" that phero66 reported from Helicon Focus in this topic. Are those areas featureless in all of the individual frames too, or have low contrast details been mushed out in the stacking? If you have lost detail in the stacking, there are several suggestions in phero66's topic for how to get it back.
--Rik
- rjlittlefield
- Site Admin
- Posts: 24017
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
For what it's worth, the specs on your new lens are very similar to the Olympus 38mm f/2.8 bellows macro, which is my very favorite lens in the vicinity of 5X -- great sharpness and contrast, no color fringes, etc etc. See here and here for examples. A great lens -- whopping fun to use! But it's important to use a wide aperture -- stop it down and the advantages over other lenses pretty much disappear.cactuspic wrote:I suspect this is a lens that I will get my money's worth out of. I am fascinated by it's capacity for higher magnification.
--Rik
- Bruce Williams
- Posts: 1120
- Joined: Mon Oct 30, 2006 1:41 pm
- Location: Northamptonshire, England
- Contact:
Irwin,
I'm afraid I also noticed the same over-soft, almost hazy areas within the image. It's almost as if there are frames are missing from the step-focus series (I'm not suggesting this is actually the case). Rik's analysis and recommendations are excellent as always and I await the results of your next project with great interest. I also think Doug's suggestion to try different subjects is a good one - perhaps something with greater contrast and more quantifiable detail at all levels within the stack.
Bruce
I'm afraid I also noticed the same over-soft, almost hazy areas within the image. It's almost as if there are frames are missing from the step-focus series (I'm not suggesting this is actually the case). Rik's analysis and recommendations are excellent as always and I await the results of your next project with great interest. I also think Doug's suggestion to try different subjects is a good one - perhaps something with greater contrast and more quantifiable detail at all levels within the stack.
Bruce
Thank you Rik, Bruce, and John for your insightful comments. I learn more when things go sideways.
Before I posted the image, I actually had engaged in some steps to maximize detail retention. I had noticed that the image had mushed out to its full potential in the stacking. Before posting, I painted in the textures, layer by layer, in the laborious process described by John. I will try this weekend to recompile the image ( unless I just decide it is easier to retake it in light of what I wll state below.
I think that there were several separate factors that contributed to the softness: 1. the cloudy day available lighting was very soft and even to begin with and of limited contrast to begin with; 2. the series was underexposed by about a stop, cutting contrast further, 3. The reds may have oversaturated a bit; 4. There are very few hard edges,hairs or small discrete details, 5. I may not have chosen the best slice to paint or transitionsed between two painted slices as well as I should have.
Doug is right. I don't think this is a great image to test lens sharpness. I think (providing that the lens tests out) it is a good shot to use the lens, but is suspect that this subject in the chosen light would appear a tad soft in the best of circumstances. I think I needed a more graphic design, contrastier light, and a better exposure to fairly test the lens
While not a perfect test (the lighting was soft, the spines too fat), I took an image of a single spine cluster of the turbinicarpus that I have previously posted. I have not post stack processed it, so this is a raw stack. The is some haloing. But I thing it provides a truer guage of the lens.
Irwin
Before I posted the image, I actually had engaged in some steps to maximize detail retention. I had noticed that the image had mushed out to its full potential in the stacking. Before posting, I painted in the textures, layer by layer, in the laborious process described by John. I will try this weekend to recompile the image ( unless I just decide it is easier to retake it in light of what I wll state below.
I think that there were several separate factors that contributed to the softness: 1. the cloudy day available lighting was very soft and even to begin with and of limited contrast to begin with; 2. the series was underexposed by about a stop, cutting contrast further, 3. The reds may have oversaturated a bit; 4. There are very few hard edges,hairs or small discrete details, 5. I may not have chosen the best slice to paint or transitionsed between two painted slices as well as I should have.
Doug is right. I don't think this is a great image to test lens sharpness. I think (providing that the lens tests out) it is a good shot to use the lens, but is suspect that this subject in the chosen light would appear a tad soft in the best of circumstances. I think I needed a more graphic design, contrastier light, and a better exposure to fairly test the lens
While not a perfect test (the lighting was soft, the spines too fat), I took an image of a single spine cluster of the turbinicarpus that I have previously posted. I have not post stack processed it, so this is a raw stack. The is some haloing. But I thing it provides a truer guage of the lens.
Irwin