Laowa 25mm F2.8 Ultra-Macro Technical Analysis

[Macro_Cosmos]

Hi all, I'd love some feedback on this article:
https://macrocosmosblog.wordpress.com/2 ... s-ver-1-0/

First time doing this kind of test, maybe I should invest in a fixed setup.

Recently completed the article and published on my blog. This further verifies that the lens performs the best wide-open, at F2.8. However the differences between F2.8, F3.2 and F4.0 is not visible unless at a pixel level ie pixel peeping.

[Lou Jost]

Very nice, it gives the kind of information that we really need to know (and which is often missing in reviews). Great to see actual quantified resolution tests. The fact that it is sharpest in the center wjhen it is wide open is a mark of quality that is very rare among consumer lenses.

However, I'd like to argue with you when you say that it is "an error" to consider that the best aperture is slightly stopped down from f/2.8. Center image sharpness is not the only criterion for lens quality. Corner quality also matters, and this usually increases when stopping down at least a tiny bit. The "best" aperture depends on the importance of the corners, and thie best compromise between center and corner sharpness depends on the photographer's needs for a particular image. So it would be useful to look at how the corner sharpness varies with aperture. I think it improves slightly upon stopping down a thrid of a stop to a half stop, and this has little effect on central resolution, as you say. But this was only my qualitative impression; it would be nice to see real numbers for corner es vs aperture setting.

The distortion parameters you promised in the next edition will also be very useful to us.

This was a very useful test! I particularly like the standardized targets. Thanks for the work.

[Macro_Cosmos]

Very nice, it gives the kind of information that we really need to know (and which is often missing in reviews). Great to see actual quantified resolution tests. The fact that it is sharpest in the center wjhen it is wide open is a mark of quality that is very rare among consumer lenses.

However, I'd like to argue with you when you say that it is "an error" to consider that the best aperture is slightly stopped down from f/2.8. Center image sharpness is not the only criterion for lens quality. Corner quality also matters, and this usually increases when stopping down at least a tiny bit. The "best" aperture depends on the importance of the corners, and thie best compromise between center and corner sharpness depends on the photographer's needs for a particular image. So it would be useful to look at how the corner sharpness varies with aperture. I think it improves slightly upon stopping down a thrid of a stop to a half stop, and this has little effect on central resolution, as you say. But this was only my qualitative impression; it would be nice to see real numbers for corner es vs aperture setting.

The distortion parameters you promised in the next edition will also be very useful to us.

This was a very useful test! I particularly like the standardized targets. Thanks for the work.

Thanks for your suggestions Lou, I agree, corners should be looked at as well. I did place too much emphasis on centre resolution.

I'm actually not too sure how I would approach corner resolution using the target. It's going to require a whole bunch of shots just for one corner. There's 4 segments, so 4 shots each corner. If I leave 3 out, it's going to make the test less explanatory. If I include all 3, that's going to be 16 sets of photos each consisting of at least 5 shots. I guess it's not that bad actually. I need to figure out a way to perfectly align the target.

This is the size of the test target with respect to a full frame sensor. It's actually quite large.


For distortion, alignment will be crucial. It will matter with panorama stitching. I've encountered various annoying distortion issues with my silicon chip stack&stitches.

I've updated the article to elaborate on further tests.

[lothman]

Hi all, I'd love some feedback on this article:

my first impression was uh there is a new Laowa lens, because I myself think of this lens as a zoom lens, therefore I would title it:

Laowa 25mm f/2.8 2.5-5X Ultra Macro Lens

[Lou Jost]

Macrero, why not just analyze the corners of your current stack (assuming you made the stack deep enough)? No need to worry about alignment, and no need to shoot all four corners separately.

Edit, oh, I see, the target isn't big enough to put a test pattern in all corners simultaneously.

[Macro_Cosmos]

Hi all, I'd love some feedback on this article:

my first impression was uh there is a new Laowa lens, because I myself think of this lens as a zoom lens, therefore I would title it:

Laowa 25mm f/2.8 2.5-5X Ultra Macro Lens

Implemented a title change!

Macrero, why not just analyze the corners of your current stack (assuming you made the stack deep enough)? No need to worry about alignment, and no need to shoot all four corners separately.

Edit, oh, I see, the target isn't big enough to put a test pattern in all corners simultaneously.

Ideally I want to place the segments in the very edge.

Cropped into APS-C, approx.


2.8


3.2


Can't see a difference in the corners at 200%. At 500%, 3.2 is a tad sharper, however this might also be a consequence of more contrast.

[Lou Jost]

At 500%, 3.2 is a tad sharper, however this might also be a consequence of more contrast.

Thanks for checking that. The increase in contrast is also a legitimate improvement in image quality. My recommendation of about f/3.2-3.5 for this lens was based on its behavior on FF, but it looks like 3.2 is a reasonable compromise for APS too. ...Anyway this is really an excellent corner result. It is remarkably uniform across the field compared to most consumer lenses.

[Adalbert]

Hi Macro_Cosmos,
Thank you for the test!
Is it better than Mitty 5x & TL200mm & Crop at 5x?
BR ADi

[Macro_Cosmos]

Hi Macro_Cosmos,
Thank you for the test!
Is it better than Mitty 5x & TL200mm & Crop at 5x?
BR ADi

https://macrocosmosblog.wordpress.com/2 ... rm-review/

There's a comparison here in my original review. Long story short, the Mit has a bit more resolution and a lot less CA, being an apochromatic lens.

I don't have the Mit 5x which is unfortunate. Could have done a controlled comparison to quantify the differences.

[Lou Jost]

I don't have the Mitutoyo 5x either, but it should outperform the Laowa in the center. On FF I am not sure it would do as well as the Laowa in the corners. But I have not checked directly.

[lothman]

just a question, if the lens could be placed more narrow to the sensor of a mirrorless camera would it still cover full frame at smaller magnification?

If yes, with a special adaptation the gap between 1:1 from "normal" macro lenses and 2,5x (start of unmodiefied Laowa) could be closed.

Or will it start to vignette?

regards
Lothar

[Macro_Cosmos]

just a question, if the lens could be placed more narrow to the sensor of a mirrorless camera would it still cover full frame at smaller magnification?

If yes, with a special adaptation the gap between 1:1 from "normal" macro lenses and 2,5x (start of unmodiefied Laowa) could be closed.

Or will it start to vignette?

regards
Lothar

No idea, but an interesting line of thought. I now have a Z6, so I can probably slap a shorter adaptor together with my custom Z-mount adaptor and a thorlabs female F-mount adaptor.
I'll give it a go this weekend. There's too much stuff to do at the moment.

There is some vignetting already. Doing so will definitely bring more. There's always aps-c though.

[ap]

hi everyone - sorry to resurrect this topic, but i recently purchased the laowa 25mm and have a question about the rectangular rear baffle. i was getting rectangular vignetting on my FF canon and removed the baffle - this seems to have improved things, but now im concerned that doing this may increase internal reflections/haze. what do folks think about this? do i need to insert a larger (maybe circular) baffle? is there an easy way to test for reflections? the pics ive taken with it (all macro stacks) seem ok
thank you for rading
andy

[Macro_Cosmos]

hi everyone - sorry to resurrect this topic, but i recently purchased the laowa 25mm and have a question about the rectangular rear baffle. i was getting rectangular vignetting on my FF canon and removed the baffle - this seems to have improved things, but now im concerned that doing this may increase internal reflections/haze. what do folks think about this? do i need to insert a larger (maybe circular) baffle? is there an easy way to test for reflections? the pics ive taken with it (all macro stacks) seem ok
thank you for rading
andy

Sounds like a design oversight, this lens was designed before Nikon and Canon seriously joined the mirrorless train.
If it's been okay, I will personally just leave it there. You can test reflections by playing around with lights placed at various angles.

[ap]

thank you! - ill try that!
andy