Very Poor IQ with AmScope 4x

[Tim Boomer]

Hello, fellow photomacrographers! This is my first post, so please let me know if I'm doing something horribly wrong or annoying. Apologies if this has been covered elsewhere, but I could not find the same issue elsewhere in the forums.

I bought an AmScope 4x finite microscope objective (model PA4XK-V300) directly from amscope.com last year, and I have to say I am not impressed with the IQ. I suppose it's about what I'd normally expect for $22.99, but it's not at all in line with what I was expecting based on glowing reviews by Robert O'Toole, Allan Walls, and others.

I'm using it with ~103mm worth of flocked extension tubes and adapters, so the total tube length is very close to 150mm, once the 46.5mm flange-to-sensor distance on my Nikon D810 is factored in. I also tried adding in another 10mm of extension in case I was mistaken that I should compensate for not using an eyepiece, but doing so did not improve IQ at all. I'm shooting with diffused flash, 1/250 sec shutter speed, ISO 64, mirror-up, EFCS, cable release, manual rail on a ball head mounted to a sturdy table, and stacked so that field curvature is not an issue. I am able to get excellent results with my Mitty 10x using the exact same setup (except for the 200mm tube lens instead of hollow tubes), so I do not see how the issue could be with shooting technique, though I'm always happy to hear kind suggestions. =)

Hopefully these photos will demonstrate what I am seeing. I compared with my Laowa 25mm set to 4x with all the same parameters just to have some frame of reference. The test subject is an oak leaf with a moldy whitefly pupa - not very pretty, but it should get the point across.

There is a slight magnification discrepancy and a color shift, but I'm not too concerned about those aspects.

AmScope 4x vs. Laowa 25mm @ 4x Overall Comparison

Of course, downsizing to 512px covers a multitude of sins, so let's look at some 100% crops. First the center:

Center crop comparison

Then off-center:

Off center crop comparison

And finally the corners:

Corner crop comparison

I didn't bother to scale the images to compensate for the discrepancy in magnification, but I think it's clear that the AmScope is mediocre at best in the center and downright unusable by the time we get to the full-frame corners. There's just no way that an objective this bad could earn such a good reputation.

I thought maybe I got a lemon, so I eventually ordered another one, this time from Amazon (where it also has glowing reviews). The objectives look the same and, sadly, perform the same. Here's one with the nosecone removed:

AmScope 4x with nosecone removed

The actual optics are deeply recessed by about 16mm, not the 2mm that Robert mentioned in his review. This leads me to believe that I received the wrong version (twice!), or maybe AmScope has simply changed the design. That would beg the question, where can one buy the "good" version of this objective in 2022?

Or if I do have the right version, what could I do differently to extract higher IQ? Here it is attached to my camera:

AmScope 4x mounted on D810

Thank you all very much for your time!

[Scarodactyl]

There are at least two lookalike models out there (viewtopic.php?f=25&t=44363) but interestingly yours looks like the good one. It may still be from a different factory or with different manufacturing specs than the good ones, or perhaps just bad QC.

[ray_parkhurst]

I've gone through several different versions of these Chinese-made 4x objectives, and they vary widely in IQ. I see that Amscope has two versions available right now. I recently purchased one and it was "OK", didn't cover FF well but was good on APS-C. The other version was not available so I didn't try it, but I do have 3 other earlier versions that look similar. Only one version was exceptional, and I have not been able to get any more of them. So...caveat emptor, and I would not expect the kind of quality shown in Robert's review as it does not seem likely you'll get the same objective.

[Tim Boomer]

This is very good information, Ray and Scarodactyl. Thank you so much for taking the time to reply! Going back to Robert's review, I can see that his version appears to be different from mine (his nosecone unscrews in the middle instead of the shoulder, and the internal optics are not nearly as recessed). Robert's copy actually looks like the generic version to which Scarodactyl referred, so it indeed seems very hit or miss with these cheap Chinese 4x objectives. Thanks again for the help!

[Chris S.]

Agreed that the Amscope looks poor. And it's clearly not designed to reach the corners of a full-frame sensor.

This said, can we expect even a good 4x/0.10 objective to hold up against a Laowa 25mm f/2.8? By my calculation, the Laowa has a numerical aperture of over 0.17. Using only the NA as a metric, one could expect 70 percent more resolution out of the NA 0.17 Loawa than the NA 0.10 objective.

Granted, a dedicated 4x objective may be better corrected at 4x than a generalist 2-5x lens. And I haven't used the Laowa--maybe it's not best wide open? (If one has to close down the Laowa to avoid aberrations, the NA goes down, too.) Am I missing anything else?

If not, we'd expect to need something like the (excellent) Nikon apo 4x/0.20 /160 to beat the Laowa. (This said, I've only used my Nikon apo 4x/0.20 on an APS-C Nikon sensor--not a full-frame sensor such as the Nikon D 810 body has. I doubt this objective would hold up well in the corners of full frame.)

--Chris S.

[ray_parkhurst]

Agreed that the Amscope looks poor. And it's clearly not designed to reach the corners of a full-frame sensor.

This said, can we expect even a good 4x/0.10 objective to hold up against a Laowa 25mm f/2.8? By my calculation, the Laowa has a numerical aperture of over 0.17. Using only the NA as a metric, one could expect 70 percent more resolution out of the NA 0.17 Loawa than the NA 0.10 objective.

Granted, a dedicated 4x objective may be better corrected at 4x than a generalist 2-5x lens. And I haven't used the Laowa--maybe it's not best wide open? (If one has to close down the Laowa to avoid aberrations, the NA goes down, too.) Am I missing anything else?

If not, we'd expect to need something like the (excellent) Nikon apo 4x/0.20 /160 to beat the Laowa. (This said, I've only used my Nikon apo 4x/0.20 on an APS-C Nikon sensor--not a full-frame sensor such as the Nikon D 810 body has. I doubt this objective would hold up well in the corners of full frame.)

--Chris S.

I was mostly concerned with coverage in my comparisons of the different versions. Only one version I tested had good FF coverage, and most had issues even at corners of APS-C, but this is not surprising given the intended use. The objective in Robert's test covered FF, but it does not look like the one I have which covers FF.

The surprise for me in this thread was the good performance of the Laowa 25mm. I had not considered this lens but I will now look into it.

Also forgot to say "Welcome" to @Tim Boomer! Excellent first post BTW.

[Tim Boomer]

The surprise for me in this thread was the good performance of the Laowa 25mm. I had not considered this lens but I will now look into it.

Overall, I'm happy with the Laowa 25mm. It has some issues for sure (can't think of a lens that doesn't), but it gets the job done. Maybe I will do a full post on it at some point. Also, thanks for the welcome! =)

[Tim Boomer]

Agreed that the Amscope looks poor. And it's clearly not designed to reach the corners of a full-frame sensor.

This said, can we expect even a good 4x/0.10 objective to hold up against a Laowa 25mm f/2.8? By my calculation, the Laowa has a numerical aperture of over 0.17. Using only the NA as a metric, one could expect 70 percent more resolution out of the NA 0.17 Loawa than the NA 0.10 objective.

Granted, a dedicated 4x objective may be better corrected at 4x than a generalist 2-5x lens. And I haven't used the Laowa--maybe it's not best wide open? (If one has to close down the Laowa to avoid aberrations, the NA goes down, too.) Am I missing anything else?

If not, we'd expect to need something like the (excellent) Nikon apo 4x/0.20 /160 to beat the Laowa. (This said, I've only used my Nikon apo 4x/0.20 on an APS-C Nikon sensor--not a full-frame sensor such as the Nikon D 810 body has. I doubt this objective would hold up well in the corners of full frame.)

--Chris S.

I prefer to use the Laowa around f/3.2, as I can't see a difference in sharpness from f/2.8 (at least with my copy on a D810), but CA is noticeably (albeit slightly) improved. IQ is good enough at 4x-5x, though I wouldn't call it excellent, which is why I was looking at other options. But now I can see that I was asking too much of this AmScope or any 4x 0.10 objective. Thank you so much for the explanation and thoughts!

[Chris S.]

I prefer to use the Laowa around f/3.2, as I can't see a difference in sharpness from f/2.8 (at least with my copy on a D810), but CA is noticeably (albeit slightly) improved. IQ is good enough at 4x-5x, though I wouldn't call it excellent, which is why I was looking at other options. But now I can see that I was asking too much of this AmScope or any 4x 0.10 objective. Thank you so much for the explanation and thoughts!

Tim,

If you're stopping the Laowa down to f/3.2 for CA reduction, its NA is something over 0.15. Conversion of f/# and NA calculator

There are lots of used Nikon 4x/0.13 Plan /160 objectives out there, and of course a few of the Nikon 4x/0.20 Plan apo /160 objectives.

NA 0.13 is close enough to NA 0.15 that if tested head to head, it would be an interesting horse race; the higher NA still has the theoretical edge, but other considerations (i.e. control of aberrations, sensor coverage) might make one or the other better in practice. I'd be willing to bet that the NA 0.20 apo would handily edge out the Laowa at 4x, though--at least within the APS-C portion of your full-frame sensor. If you can entertain infinite objectives, the Mitutoyo 5x/0.14 seems to punch above its weight, and is quite good into the corners on APS-C--and probably out to FF corners as well.

I echo Ray's welcome! From the style of your post, you could have been around here for years; I forgot you were a newcomer.

--Chris S.

[rjlittlefield]

If you're stopping the Laowa down to f/3.2 for CA reduction, its NA is something over 0.15. Conversion of f/# and NA calculator

It's not hugely wrong, but that calculator is not really appropriate here.

The calculator converts between NA and f-number as alternative descriptions for the same ray cone.

But when a finite lens like the Laowa is set on f/3.2, it won't actually be giving an f/3.2 cone because the lens has to be positioned more than 1 focal length away from the subject in order to focus. To get the right NA, you have to correct the f-number for that increased distance.

Under reasonable assumptions, the correction is a factor of (1+1/m) for magnification m.

Using that formula, at 4X, you'd be looking at a working f-number of 3.2*(1+1/4) = f/4 (subject side), equal to NA 0.125 .

I recently wrote a new FAQ on the closely related issue, FAQ: How can I calculate effective aperture?

Using the formulas in that FAQ for this application, you would calculate sensor-side effective aperture Feff = 3.2*(4+1) = f/16, then convert from Feff to NA as NA = m/(2*Feff) = 4/(2*16) = 0.125.

--Rik

[Tim Boomer]

Tim,

If you're stopping the Laowa down to f/3.2 for CA reduction, its NA is something over 0.15. Conversion of f/# and NA calculator

There are lots of used Nikon 4x/0.13 Plan /160 objectives out there, and of course a few of the Nikon 4x/0.20 Plan apo /160 objectives.

NA 0.13 is close enough to NA 0.15 that if tested head to head, it would be an interesting horse race; the higher NA still has the theoretical edge, but other considerations (i.e. control of aberrations, sensor coverage) might make one or the other better in practice. I'd be willing to bet that the NA 0.20 apo would handily edge out the Laowa at 4x, though--at least within the APS-C portion of your full-frame sensor. If you can entertain infinite objectives, the Mitutoyo 5x/0.14 seems to punch above its weight, and is quite good into the corners on APS-C--and probably out to FF corners as well.

I echo Ray's welcome! From the style of your post, you could have been around here for years; I forgot you were a newcomer.

--Chris S.

Thanks a bunch, Chris! I read through many posts here before finally joining. There is not only a huge wealth of knowledge, but also such a kind and helpful community of macro and micro enthusiasts/experts! I have found my people!

And as for infinite objectives, I actually prefer them, since I always bring my 200mm macro with me. Its integrated tripod collar makes it much easier to use than some amalgamation of tubes and adapters. It's what I use for my Mitty 10x/0.28, and it works out great. Of course, 10x is too much magnification for some subjects. I tried pushing it down to 6.25x with a raynox 250 tube lens, but the edges were not good, even within the APS-C portion of the frame. I had to crop so much to make it usable, I was back at 10x equivalent.

I do eye that Mitty 5x/0.14 from time to time. I'm sure it would beat out the Laowa 25mm at 5x (at least in the center), but I just don't think I can justify the cost (even used) when the Laowa gets the job done well enough. And it is really nice to have variable magnification and aperture.

[Tim Boomer]

It's not hugely wrong, but that calculator is not really appropriate here.

The calculator converts between NA and f-number as alternative descriptions for the same ray cone.

But when a finite lens like the Laowa is set on f/3.2, it won't actually be giving an f/3.2 cone because the lens has to be positioned more than 1 focal length away from the subject in order to focus. To get the right NA, you have to correct the f-number for that increased distance.

Under reasonable assumptions, the correction is a factor of (1+1/m) for magnification m.

Using that formula, at 4X, you'd be looking at a working f-number of 3.2*(1+1/4) = f/4 (subject side), equal to NA 0.125 .

I recently wrote a new FAQ on the closely related issue, FAQ: How can I calculate effective aperture?

Using the formulas in that FAQ for this application, you would calculate sensor-side effective aperture Feff = 3.2*(4+1) = f/16, then convert from Feff to NA as NA = m/(2*Feff) = 4/(2*16) = 0.125.

--Rik

Thanks a bunch for the detailed info! As a math major once upon a time, I always appreciate numbers and formulas. Though at the end of the day, I have to admit I usually end up just making sure the photos look good.