Canon 400mm F2.8 L FD (non-IS, manual)

[bojan]

It's probably a bit too late to ask the above question.. because I already bought it on Ebay for ~1k.
However, I would still like to know what exactly to expect when the lens arrives in a week or two.

I was reasonably happy with my (converted to EOS) Canon 300mm F4, but with this larger (and "L") cannon I expect much less colour fringing and better resolution, judging by Avandonk's work with his 300m mF2.8.

[sil]

have a look on Astrobin http://www.astrobin.com/search/?q=Canon%20FD%20400mm%20F/2.8%20L%20EF%20converted&search_type=0&license=0&license=1&license=2&license=3&license=4&license=5&license=6&telescope_type=any&telescope_type=0&telescope_type=1&telescope_type=2&telescope_type=3&telescope_type=4&telescope_type=5&telescope_type=6&telescope_type=7&telescope_type=8&telescope_type=9&telescope_type=10&telescope_type=11&telescope_type=12&telescope_type=13&telescope_type=14&telescope_type=15&telescope_type=16&telescope_type=17&telescope_type=18&telescope_type=19&telescope_type=20&telescope_type=21&telescope_type=22&camera_type=any&camera_type=0&camera_type=1&camera_type=2&camera_type=3&camera_type=4&camera_type=5

This link is for the Canon FD 400mm F/2.8 L EF converted, dont know how different it is for your lens, i'm a Nikon kind of guy, but search around on Astrobin to see what gear is capable of (bear in mind processing differences)

[bojan]

Hi Sil, Thanks.
Yes, this lens is quite good, however the images on astrobin are rescaled.. so not quite what I wanted to see.
Anyway, I will know soon enough ;-)

[bojan]

Today, after a lot of anticipation, the lens arrived....
But it was a disappointment.
I didn't mind the external cosmetics, neither the interface to camera (which was stuck BTW). I intended to use it for astrophotography only, so modification to EOS was on my mind from the start.
When I had a look inside against the sunlight I saw the lens have extensive damage caused by fungus and moisture on the back of focussing element. Ouch..
And then I made a biggest mistake of all: I started to dismantle it to see if this fungus can be cleaned...
And then I found out someone attempted to clean it before me and even did not properly assembled the affected element so it was loose inside. Retaining ring was loose and visible as that from the start.
The lens was not really functional when arrived, but I can't prove it now..
Focussing element (7-th from the left on attached optical schematic) needs polishing on back side, it is negative doublet, 70mm dia and 18mm thick.

So, question: where can I polish this surface, or, where can I obtain cerium oxide to do that myself? Are there any mirror-makers left around?
I did polish mirrors in the past, so.. maybe I can do it again?

The front group looks OK (just a tad of fungus on the back of the meniscus, 6-th element from the left), and this can be probably cleaned with windex and dishwashing detergent, no big deal at all.

[RobF]

Oh man, really feel for you Bojan. What a disappointment, let alone the problems from looking more closely at it. I hope there's a Canon lens guru here, or at least a recommendation for someone that could help.

If you're only doing astrophotography you may not even need the iris mechanism? Could be better to stop down externally with rings or something else, assuming you can safely reach the fungus and deal with it satisfactorily.

[Octane]

That's gutwrenching, Bojan.

Any chance Canon in Sydney would look at it for you? I had one of mine serviced there (there was a defect notice issued for the lens so it was free).

H

[bojan]

Yes.. it was a bad night :sadeyes:

The probability of finding this particular glass element is pretty slim, but I will certainly ask Canon about it, just in case..

Anyway, I think the affected surface is definitely worth re-polishing, considering the condition of the rest of the glass.
Even without SSC coating, there will be a very small amount of visible inside reflections (because the icoming light beam will be reflected back and diverged on the convex surface of the element).

[sil]

possible future refractor project with the lens elements?

[bojan]

Exactly right :shrug:

[bojan]

I tried te send email to Isaac.. but it bounced back.

Does anybody know if he is still doing what he used to do?

[bojan]

This is what I am talking about

[Stefan Buda]

Hi Bojan,

I should be able to help you if you can trust me with it. Perhaps you can drop in one day so that I can take a good look at it. My guess is that it will need to be reground before polishing. That could be done with minimum thickness loss of perhaps a few hundredths of a mm. I can't resist such a challenge so I would do it free of charge.

Stefan

[billdan]

Hi Bojan,

If you can't get Cerium Oxide you could try Rouge, or even Acetone and if all else fails toothpaste.

Bill

[bojan]

Hi Stefan, this is fantastic, thank you!

[bojan]

While the problem with the glass surface is being expertly sorted out by Stefan, I was working on modification to EOS..
I removed the iris and all mechanics around it (it will be a spare part in case someone needs it in the future), and designed the necessary parts to fit the EOS adapter..
I have two of them, one is with T-thread , another with M42x1mm.
The one with t-thread has necessary thickness to ensure the camera sensor is in the focal plane of the optical system, so I designed the fitting around it.

The printed plastic part turned out so well I decided to leave it as is.. I only need to slightly reduce the diameter of the section where thread is supposed to be.. so the metal ring of the adapter can be force-screwed on the plastic (unfortunately, my 3D printer is not capable of producing threads with such a small pitch (0.75mm.. or is it?) .. and I need to paint it black inside (or re-print it with black ABS).

[blink138]

any sort of reworking is going to cost you a multicoat bojan, you know that right?
depending upon the refractive index of the glass you are going to lose 6% off each surface, not a trivial amount!
pat

[bojan]

Pat, I am aware of that..
but it is only one surface affected.. so 6% is not a big deal. I can absorb that.

This is equivalent to reduction in aperture from 140mm down to 135.7.. negligible.

I am pretty sure the loss due to current scattering on the damaged surface is worse.. not to mention the loss of contrast.

[blink138]

it is 12%........... 6 off each surface, if both surfaces are affected of course?
pat

[blink138]

ah just re read your post, only on one surface then?
pat

[bojan]

Yep, only one..
BTW, for 2 surfaces the attenuation is 11.64%.. or transmission is 100% - 6% = 94%, and minus additional 6% of 94%, total is 88.36%..
11.64% in total ;)

[bojan]

Well, I tried to make the thread 42x0.75mm.. and guess what, it worked :thumbsup:
It doesn't look perfect, but it can be screwed in and tightened enough to support the camera weight.

[blink138]

that depends solely upon the refractive index, but that is what i said!
the higher the refractive index the more light lost due to reflection, but also the better the multicoat works!
i have seen enough and worked on many spectacle refractive indices when i was hoya lens optical technician and an uncoated THI 1.8n lens looks like a mirror!
pat

[bojan]

I've just received the message from Stefan that my lens is ready for me to pick it up :thumbsup::D

Now I have to finish the mechanical aspect of this adventure, over the next weekend.
And then, it will probably rain every night here in Melbourne, for the whole month ;)

[bojan]

Ready for star test tonight, weather permitting.

I couldn't dismantle the front group yet, so a bit of fungus is still there, but I am not worried about it right now.

[AlexN]

That looks the business mate. I'll post a pic shortly of my backyard engineering project involving an old manual focus 300/2.8. I've hobbled together an arduino/easy driver stepper focuser to the rig as well as working on implementing temp compensation. I've found through testing that the focus shifts by enough to ruin images with every degree of temperature change. I am considering two options. Temp compensation in the focuser but this will require a lot of coding that im not too sure on how I would implement and a very accurate thermal probe, or alternatively I am considering using heater bands at the front and back of the lens with a thermal probe and a PID maintaining the lens temperature at say 17 degrees all night, thus negating focus shift.

Seems there is far more involved in using a camera lens for imaging than I thought.

[AlexN]

Oh, and yes. Since having mine mechanically ready and only needing a clear night to test it's been beautiful sunshine all day with afternoon storms and cloudy nights for 7 days.

[bojan]

First test..
Just a hint of CA, when slightly out of focus.

I think the lens is as good as new :), thanks to Stefan.
This is a crop from 200%, taken at full aperture.

[Shiraz]

brilliant outcome!

[AlexN]

Sensational outcome to be sure. Looks to be a cracker of a lens

[bojan]

It seems it suffers from slight astigmatism, but I didn't expect a perfect performance at F2.8 anyway. So, with external aperture set at 3..3.5 it will be much better.

This crop is form 100% size, one frame straight out of camera (Canon 60D, full resolution).
In corners - the same.. it's as flat as pancake :)

[blink138]

hi bojan i hope it all worked out.
what are the details of stefans operation on the lens may i ask?
pat

[bojan]

He removed couple of 1/100'ds mm of glass from the damaged surface, polished and applied antireflective coating, so the lens element looked like new after that.

[AlexN]

Stefan just jumped to the top of the 10 most helpful people list I guess. Haha

[Octane]

Bojan,

That's a great outcome. I'm really happy for you! Great work, Stefan.

Would you mind posting a sample image once it's stopped down to f/3.5, please?

H

[bojan]

H,
You will have a full report.. I only need a bit of clear skies to do tests.
Probably by this weekend.

[Octane]

Awesome, cheers. :)

H

[bojan]

Comparison between F/2.8, F/3.5, f/4
Since I removed internal iris, the aperture is external (black plastic folder).
Exposure times were 5sec, 7sec 10sec, single frames, straight out of camera.
Stacks look much better - see Catalina (crop from full resolution stack 16x10sec, ISO1600 @f/4)

The difference between full aperture (143mm) and f/3.5 is drastic.. but 3.5 and 4 are almost the same (in terms of diameter, this is 114 and 100mm).

So, it seems 110mm will be just right.

[bojan]

This morning I did some processing of the frames from previous post, taken last night.

Those processed images below are crops (100% size) from the middle and all four corners of the full resolution stack (Canon 60D) 17x10sec, ISO1600, @F/4 (external aperture).
I am pretty happy with the results.

Stefan, I wish to thank you here again for your help, in my opinion this Canon monster is fully restored.

[bojan]

For comparison, below are crops from 100% sized stacks (16x) taken with f/2.8, f/3.5, f/4

Drastic difference between f2.8 and 3.5.
F/3.5 was taken with 50% increase (7sec) in exposure time.
F/4 was taken with 10sec exposure time.

[Shiraz]

just a thought, but some optics designs use the aperture placement to help control aberrations in addition to limiting the amount of light (eg a MN uses the aperture placement to control coma). I have no idea if it makes sense with your lens, but it might be just be possible that it would work better at f2.8 if a full open aperture stop was placed where the original one was.

[bojan]

I was thinking about it.. and I will try it soon.
Especially because it seems this lens should behave a tad better at full aperture, according to another owner.
However, based on experience with 200m and 300mm Canon lens (similar optical design),
(http://www.iceinspace.com.au/forum/showthread.php?t=51866&highlight=canon+200mm), (and some russian lenses like Tair 11A (135mm) and Tair 33 in particular), the external aperture is way better.. Canons 200mm f2.8 and 300mm F4 have a lot of coma in corners (APS sensor.. and lenses were designed for full frame!) with internal iris, which is not there when external aperture is applied.
I think the reason why iris is where it is in Canon large telephoto lenses is simply because it would have been mechanically very complicated (for manual design.. with motors it is easier of course) to place it near front group.. so the current position is a compromise IMO.
However, smaller lenses are different (symetrical tessar-like designs). The external aperture doesn't work with them, it actually worsens the coma and/or CA.

I converted this lens to EOS by removing everything FD from it (including iris), so I can't really tell.. and I don't want to bother the only other person I know to own this lens to do the test for me (it is not really hard to cut the aperture.. but still.. )
In order to put iris back, I need to redesign the whole iris mechanism.. and this is not easy to do.

[Stefan Buda]

Thanks Bojan for the detailed feedback. It is good thing that we were able to rescue such a nice piece of optics.

Regarding aperture stops, please note that with camera lenses the iris is an integral part of the optical design and its location is important - especially for wide angle lenses. As Ray suggested, the placement of the aperture stop at the front is only valid for image points that are near the centre of the field. In the corners some of the aberrations sneak back because the aperture stop lo longer lines up with the entrance pupil as defined by the iris set for the same f-stop.

[bojan]

Yesterday I put back the internal iris and tried again (at f/2.8).
There was no improvement at f2.8 internal iris stop compared to no iris at all.
Now I am working towards the full EOS modification (3D-printing some required parts from ABS).

[AlexN]

What is the shape of the internal iris stopped down to f/3.5 Bojan?

My tokina 300/2.8 gives fairly bloated stars at f/2.8, but tiny pinpoint stars at f/4. The only problem at f/4 due to the rough shaped internal iris I get slightly odd star shapes with some weird aberrations. I'm testing some home made aperture masks to see where best performance lies then I will get something 3D printed.

Unfortunately my lens goes directly from f/2.8 to f/4 with the internal iris. I suspect f/3.5 will give me the performance I'm after.

[bojan]

Not sure because I still have to design and build mechanism to control iris now that the m42-EOS adapter is in place.
Try with external aperture.. it is 5 minutes job to cut it from black paper or folder plastic.

[AlexN]

I've been doing tests. Just trying to nail down a method of cutting a perfect circle.

[bojan]

I am using one of those:
http://www.ebay.com.au/itm/30mm-to-120mm-Dia-Adjustable-Round-Hole-Cutter-w-Hex-Spanner-Chisel-End-Mill-/252083294010?hash=item3ab155cf3a
Also, round aperture doesn't produce diffraction spikes.

[AlexN]

That brilliant. I've been finding I can get star images better than using the internal iris but I get flares and spikes due to the holes being slightly out of round or not perfectly central.

I'll buy one of those cutters and see how I go then.

[skysurfer]

Well, this $10000 lens is indeed really good.

But I bought a 200 f/2.8L for only $500 (used) but its image quality is excellent. Pinpoint stars to the corners at full aperture.

[bojan]

Hi Alex,
The shape of internal iris at f/3.5 is not so good.. f/4 is OK.
External aperture behaves much better, as far as diffraction is concerned.
I will do comparison between internal and external and post the results as soon as possible.

[bojan]

With some delay, I thought it may be useful (for archival purpose) to add images of my conversion of this lens to EOS.