Whats the best lens for super8?

I’ve been using the RPi HQ camera with the Schneider Componon-S 50mm for a while now. I think it’s the best solution so far, and I see why it’s recommended on this site multiple times. But I’m not really satisfied for multiple reasons:
1 - When in focus it’s an overall good lens for capturing super8, but it’s like 2 megapixel good in the center, and a little less when moving out from the center. It’s not 12MP good, and I don’t see the filmgrain detail I would like to have.
2 - It seems overkill to use such a big lens, for such a small object + small sensor, this can be more efficient (cheaper). But still I would rather have a huge lens that’s great, than a tiny lens that’s blurry in the corners.
3 - It’s difficult to focus, but most are, simply becouse the super8 film is so tiny.

So I set out to test the lenses (mostly to find the best image quality) I have on the RPi HQ camera with the following 8mm wide cutout (next to a key):

First, the componon-s 50mm ‘reference’ lens, with some extension tubes:



I rate this lens:

  • sharpness center: 8
  • sharpness edges: 7
  • color fringing: 10

NOTE: I edited all images slightly so they are better comparable. I aligned, cut and white balanced them all.

Second is my old €20 m12 6-22mm zoom lens:



It performs not bad at all, but you can see some color fringing
I rate this lens:

  • sharpness center: 7
  • sharpness edges: 6
  • color fringing: 5

Third is my old m12 16mm lens (I think it’s 16mm, but there’s no description on the lens and I lost the info from where I bought multiple different tiny lenses for max €5):



Before I had the componon-s I used this lens a lot. It’s almost perfect. Technically this solution can be cheap, very small, and give the same results as big or huge lenses. But does such a lens exist, and how do I find it if it does?
Anyway, I rate this lens:

  • sharpness center: 8
  • sharpness edges: 5
  • color fringing: 9

My fourth attempt is a newly bought €16 “microscope” lens, hoping to get more sharpness. Looking at many video’s about these things they should be able to do the job:



I was disappointed and rate this lens:

  • sharpness center: 7
  • sharpness edges: 6
  • color fringing: 4

Lastly, here’s a cutout compare of the edges and center of what I think are currently my 2 best lenses for super8 capture:


A clear win for the schneider, but still not as good as I would like. Here’s the two lenses with their extension tubes, side-by-side:

And the 16mm tiny lens all alone:

Random experiment trying to use a 23MP DSLR instead of the 12MP RPi HQ. But this time the lens setup really becomes rediculous.

Another random and very simple experiment. Using a projector and the normal projector lens, you can ‘project’ the super8 frame onto the DSLR sensor. The image quality is not bad at all, and the lens setup seems really efficient.


What other lenses or setups do you use? I’m very interested so please reply, preferably with images.

Some random thoughts:

  1. With the RPi HQ camera we are doing a close to 1:1 magnification. This is basically what every macro lens does, so would using a good 50mm macro lens perhaps be the perfect solution for image quality? However, not the most efficient or cheap solution. When using a lens in its working range, we can look for reviews on how they perform. The only difference being that we use a smaller sensor so the 1:1 magnification perfectly fits the super8 usecase.
  2. When using larger sensors (APS-C of fullframe), we need a magnification of 4? This is not what any normal lens is made for (besides some exotics like made by laowa). This is more in the “microscope” range of things. Now I see there’s many cheap microscope lenses, and adapters for many DSLR mounts. So wouldn’t this be a good solution? Basically turning a DSLR into a microscope makes more senso to me than using huge lenses, with even larger extension tubes to get results.
4 Likes

Thank you for sharing this post, great information.
I have used, the Componon-S, and Nikkor EL, both 50mm f2.8 lenses. I found the Schneider performed slightly better away from the center than the Nikkor EL, with f4. In both cases, the lenses were reversed mounted with a 23.2 x 15.4 mm CMOS sensor (Nikon DSLR D3200). Both lenses definitely resolved enough to see the grain, and even minimum scratches/dust.
In general, for this use case, enlarger lenses are designed to best focus at a flat surface (paper) and from a flat subject (the film). And have the added value of being relatively inexpensive for the use case.

A few items I found extremely critical for best quality away from the center, with both lenses:

  • Select the best aperture for sharpness and resolution. This was specially critical with the larger sensor, one stop less was not as noticeable with the smaller sensor. The coinimaging site provides a great reference, and it also shows the significant loss of resolution on the corners.
  • The depth of focus is very small, and any misalignment between the target, the lens, and the lens and the sensor is highly noticeable.
  • Check the IR filter of the HQ camera for spots… there are some postings on the subject at the camera forum. I no longer have a filter in the sensor, so cannot provide a reference.

With both the Componon and Nikkor grain was nicely preserved, here is an example of the Nikkor EL on an earlier post., the Schneider is slightly better on the corners. The film was shot with the emulsion-side to the lens, using the 1/2 stop between f4 and f5.6, and D3200 DSLR.

The illuminant configuration is also a very significant part of how the grain looks, and it took multiple iterations to get to something decent (an oversized ping pong ball).

1 Like

Nice test, I wonder how I can add to this with a Canon MP-E 65mm and this cheap 300x variable microscope lens regarding decent test material. Are you using the older metal barrel version of the Componon-s 50mm f2.8? According to:closupphoto 50mm test it uses 1 lens element less and performs worse.

I was looking into lens selection this week and it was not easy to find proper dedicated macro machine vision c-mount lenses that support the pixel-size of the RPi HQ or IMX226 (1.4 and 1.84 micron). It made me wonder if going with a larger sensor like a IMX304 will always be inherently better despite requiring 2x magnification. Because the lens does not need to be as sharp.

update: Bought a film enlarger at a thrift store, it had a Soliger 50mm f3.5 enlarger lens. It’s probably not better than the Componon-s, but it seems to be good for UV photography.

As test material I plan to use a linear encoder scale, since it’s black stripes on a translucent film. Missing a few parts for my test setup though, so can’t test yet.

2 Likes

This and closeuphotography are great sites. I didn’t know these and they provide exactly the information I was hoping to find with this post. Silly I didn’t find those sites, I searched a long long time and only found a lot of info about macro photography, ending at 1:1 magnification.

Looking at the info on those 2 sites, I think it’s worth a try to get reverse mounting rings/adapters for my componon-s 50mm, and go for about 4 times magnification to an aps-c or fullframe sensor as you did.

I have the plastic version (you can see it in my first post), thank you for pointing me to that website. After reading all that new information I’m not sure how to get to 4x magnification. Extension tubes, or stacking two componon-s lenses together as the closeuphotography seems to do. I see many new interesting ideas.

Yes, I’m now using a very bright COB LED:
https://de.aliexpress.com/item/32940279993.html
And I cut up this, and used it as a diffusor:
https://de.aliexpress.com/item/1005001771536770.html

The difference with/without the diffusor is night and day.

To keep the information on this great post, here is a 2023 test comparing 50mm lenses available for scanning.

4 Likes