Actually I based my statement more on the results i get from my 10x objective, rather than theoretical consideration. Perhaps I got one that is not quite up to specs (still works well enough for me, though, so I am not planning to buy another one), but what I see is that with pixel-shift I get little or no improvement in sharpness with this objective.rjlittlefield wrote: ↑Sun Mar 05, 2023 10:43 am[...]
I expect that you're using some different calculation.
What is it?
--Rik
New olympus 90mm 2x macro
Moderators: rjlittlefield, ChrisR, Chris S., Pau
- enricosavazzi
- Posts: 1478
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
Re: Review now online
--ES
- enricosavazzi
- Posts: 1478
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
Added MTF to review
I added a few MTF graphs and other materials to the review. What the MTF graphs tell me is that when switching to the S Macro range, the lens does something that improves image quality, in spite of the penalty in nominal aperture (F/5 vs. f/3.5). This is the opposite of what I expected (and originally said in my review), but nonetheless those are the numbers.
--ES
- rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Review now online
enricosavazzi wrote: ↑Sun Mar 05, 2023 3:44 amMy reasoning is that an M Plan Apo 10x with 200 mm tube lens already fully utilizes the resolution of a 20 Mpixel Micro 4/3 sensor. By pushing down the magnification with a 90 mm tube lens, you do get more than twice the lp/mm in the optical image (the total number of lp across the field of view of the objective remains the same, but the field of view shrinks in absolute size). The optics project an image containing much finer detail on the Micro 4/3 sensor, but the sensor was already at its resolution limit with a 200 mm tube lens and cannot record any more detail.
OK, now I'm confused.enricosavazzi wrote: ↑Sun Mar 05, 2023 2:21 pmActually I based my statement more on the results i get from my 10x objective, rather than theoretical consideration. Perhaps I got one that is not quite up to specs (still works well enough for me, though, so I am not planning to buy another one), but what I see is that with pixel-shift I get little or no improvement in sharpness with this objective.rjlittlefield wrote: ↑Sun Mar 05, 2023 10:43 am[...]
I expect that you're using some different calculation.
What is it?
--Rik
It seems to me that if pixel-shift gives no improvement, then the conclusion should be that the objective + 200 mm tube lens is not providing enough detail to fully utilize the sensor.
But it sounds like you're saying the opposite, hence my confusion.
Am I reading something wrong?
--Rik
- enricosavazzi
- Posts: 1478
- Joined: Sat Nov 21, 2009 2:41 pm
- Location: Västerås, Sweden
- Contact:
Re: New olympus 90mm 2x macro
If we use the Nyquist limit, I can retract the part of my statement where I say that "an M Plan Apo 10x with 200 mm tube lens already fully utilizes the resolution of a 20 Mpixel Micro 4/3 sensor", since the Nyquist limit of 2 pixels (3.49 * 2 = 6.98 μm) is below the computed diffraction cutoff at sensor of 9.8 μm (i.e. sensor outresolves lens).
I computed a theoretical 50% MTF for an ideal 20 Mpixel Micro 4/3 sensor at 115 lp/mm ( https://www.savazzi.net/photography/mic ... s.html#mtf ), which yields a sensor resolution per cycle of 1000 / 115 = 8.7 μm. This is closer to the diffraction cutoff of 9.8 μm but still slightly lower. If we go up to 80% MTF as often used in lens tests, sensor resolution per cycle drops to 1000 / 67 = 14.9 μm (i.e. lens outresolves sensor). The Nyquist limit happens to be (in this case, not as a general rule) at 30% MTF.
Pushed down at 4.5x, I think we agree the objective clearly outresolves the sensor, which I regard as the main point of my original statement.
I computed a theoretical 50% MTF for an ideal 20 Mpixel Micro 4/3 sensor at 115 lp/mm ( https://www.savazzi.net/photography/mic ... s.html#mtf ), which yields a sensor resolution per cycle of 1000 / 115 = 8.7 μm. This is closer to the diffraction cutoff of 9.8 μm but still slightly lower. If we go up to 80% MTF as often used in lens tests, sensor resolution per cycle drops to 1000 / 67 = 14.9 μm (i.e. lens outresolves sensor). The Nyquist limit happens to be (in this case, not as a general rule) at 30% MTF.
Pushed down at 4.5x, I think we agree the objective clearly outresolves the sensor, which I regard as the main point of my original statement.
--ES
Re: New olympus 90mm 2x macro
I think that's a very good reason for pushing down, especially since we can use pixel shift to get sub-pixel resolution.Pushed down at 4.5x, I think we agree the objective clearly outresolves the sensor, which I regard as the main point of my original statement.
- rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: New olympus 90mm 2x macro
Enrico, thanks for the clarification. At this point I agree with both you and Lou.
--Rik
--Rik
-
- Posts: 15
- Joined: Fri Mar 17, 2023 1:14 am
- Location: España, Miranda de Ebro
Re: New olympus 90mm 2x macro
A small comparison with 60 mm macro Olympus would be very clarifying.
Thanks
Carlos
Thanks
Carlos
Re: New olympus 90mm 2x macro
Yes...the 60 is quite good at f/3.5 so it might have a resolution advantage.CarlosHermosilla wrote: ↑Thu Mar 30, 2023 7:02 amA small comparison with 60 mm macro Olympus would be very clarifying.
Thanks
Carlos
-
- Posts: 15
- Joined: Fri Mar 17, 2023 1:14 am
- Location: España, Miranda de Ebro
Re: New olympus 90mm 2x macro
Hi, just tested this lens as a tube lens with a Mitu 5x. If you focus at medium distance, not infinity, the lens does not produce any vignetting and is perfect for M4/3 focus bracketing. The focus range is so wide you can focus at a distance that is not minimum focus and it fits More than enough travel to do the job perfectly. Grettings to all
Carlos
Carlos
-
- Posts: 15
- Joined: Fri Mar 17, 2023 1:14 am
- Location: España, Miranda de Ebro
Re: New olympus 90mm 2x macro
I apologize for the joy of precipitation. I have mounted a series and the result is not good outside the center of the image. The quality degrades towards the edges and a lot. I have to try a duplicator to see if in this way the entire frame is usable, in the individual shots I did not appreciate the final quality well. Therefore, as a tube lens, it does not vignetting but it does present a quality drop problem towards the edges that I will see if it can be resolved a bit.
Re: New olympus 90mm 2x macro
That's not really surprising to be honest.CarlosHermosilla wrote: ↑Wed Apr 19, 2023 3:08 amI apologize for the joy of precipitation. I have mounted a series and the result is not good outside the center of the image. The quality degrades towards the edges and a lot. I have to try a duplicator to see if in this way the entire frame is usable, in the individual shots I did not appreciate the final quality well. Therefore, as a tube lens, it does not vignetting but it does present a quality drop problem towards the edges that I will see if it can be resolved a bit.
With a tube lens of 90mm you'd be looking at a total magnification of around 2.25x (at infinity) - which is pretty far from the optimal range for the Mitutoyo in the first place. And if you're then using the tube lens in a "macro"-setting it gets even more complicated with focal length reduction etc. - and if you use the internal focussing to focus stack/bracket you'll end up with changing focal lengths/distortions and the background will also change...so a bit of a nightmare.
Re: New olympus 90mm 2x macro
I am not so pessimistic. I do not think that your first point is correct; infinity-corrected objectives can work just fine with any focal length tube lens focused at infinity. The only thing special about 200mm is that the magnification is the nominal magnification. There is nothing else special. The MFT tube length equivalent to 200mm (in FOV) is 100mm, quite close to 90mm. An objective that covers FF with a 200mm tube lens can also cover an MFT sensor (with exactly the same quality image) on a 100mm tube lens.FotoChris wrote: ↑Wed Apr 19, 2023 5:56 amThat's not really surprising to be honest.CarlosHermosilla wrote: ↑Wed Apr 19, 2023 3:08 amI apologize for the joy of precipitation. I have mounted a series and the result is not good outside the center of the image. The quality degrades towards the edges and a lot. I have to try a duplicator to see if in this way the entire frame is usable, in the individual shots I did not appreciate the final quality well. Therefore, as a tube lens, it does not vignetting but it does present a quality drop problem towards the edges that I will see if it can be resolved a bit.
With a tube lens of 90mm you'd be looking at a total magnification of around 2.25x (at infinity) - which is pretty far from the optimal range for the Mitutoyo in the first place. And if you're then using the tube lens in a "macro"-setting it gets even more complicated with focal length reduction etc. - and if you use the internal focusing to focus stack/bracket you'll end up with changing focal lengths/distortions and the background will also change...so a bit of a nightmare.
The changing focal length during focusing, as you mention, is real problem. And certainly other factors can also wreck the lens's utility as a tube lens. But deviation from 200m is not by itself an issue.
Re: New olympus 90mm 2x macro
I have to agree with Lou. The departure from infinity focus is the issue, not the focal length. As far as sensor coverage, the physical size of the tube lens aperture will restrict the coverage, i.e. for a shorter focal length tube lens, a larger aperture setting will be required to cover the same area (and variable aperture lenses are not typically optimized at their extremes).
Re: New olympus 90mm 2x macro
That's basically what I was talking about. The magnification is directly tired to the (usable) image circle and below 3x it's getting very difficult even on small(ish) sensors.
I've yet to see an image of decent quality from a Mitutoyo 5x below 3x - so far. But I may be wrong and on M43 it's doable.
However - my primary concern was that it doesn't make a lot of sense to use a Mitutoyo close to 2x stacked on a native 2x lens.
I've yet to see an image of decent quality from a Mitutoyo 5x below 3x - so far. But I may be wrong and on M43 it's doable.
However - my primary concern was that it doesn't make a lot of sense to use a Mitutoyo close to 2x stacked on a native 2x lens.
Re: New olympus 90mm 2x macro
"However - my primary concern was that it doesn't make a lot of sense to use a Mitutoyo close to 2x stacked on a native 2x lens."
That's a very good point.
But this:
"I've yet to see an image of decent quality from a Mitutoyo 5x below 3x"
still seems be laboring under a misconception about the importance of magnification in infinity-corrected optics. As I said, the aerial image projected by a Mitu 5x + 100m m tube lens on an MFT sensor (m=2.5) is exactly the same as the image projected by the same objective on an equally good 200mm tube lens on FF. If you are happy with one, you must be happy with the other; they are the same image, just scaled differently. (There can be differences in sensor quality between these two formats, but that is a different story, and has nothing to do with the lens.)
That's a very good point.
But this:
"I've yet to see an image of decent quality from a Mitutoyo 5x below 3x"
still seems be laboring under a misconception about the importance of magnification in infinity-corrected optics. As I said, the aerial image projected by a Mitu 5x + 100m m tube lens on an MFT sensor (m=2.5) is exactly the same as the image projected by the same objective on an equally good 200mm tube lens on FF. If you are happy with one, you must be happy with the other; they are the same image, just scaled differently. (There can be differences in sensor quality between these two formats, but that is a different story, and has nothing to do with the lens.)