Question about BAFFLES to improve sharpness

Started Oct 17, 2009 | Discussions
Antoine Palade Contributing Member • Posts: 589
Question about BAFFLES to improve sharpness

Context: this is a question about the use of cardboard baffles at the back of some legacy lenses, to reign-in stray lighting that results in hazy, David Hamilton-like pictures when used wide-open. The baffle is a round cardboard cut-out whose diameter matches that of the rear of the lens. At the middle of the baffle, a hole has been cut out, whose diameter varies depending on the lens. The resulting "focusing" of light reaching the sensor helps improve sharpness dramatically when used wide open. The idea has been developed by a french guy ( http://www.esperado.fr/fr/objectif_manuels/les_baffles_pour_les_optiques_manuelles_24x36_ou_moyen_format.html ) and some on this forum have also mentioned this solution when using legacy lenses (e.g. old lenses from film times) at full aperture (e.g. f1.4).

Question for those of you that might have tried baffles: What do you glue the baffle to?

I'm aware that the use of rubber cement (that sticky plastic found at the back of new credit cards) is mentioned. But what do you glue the baffle to? Do you glue it directly on the exposed glass? I'm not particularly keen to do that, no matter how harmless that rubber cement looks like. I have two lenses I want to try that solution on: a Zuiko 50mm f1.4 "Silvernose" and a Canon FL 55mm f1.2. I haven't looked at the Zuiko yet, but on the Canon 55mm I would have either to glue the baffle on the lens (not!) or on the metal edge just outside the lens. However, that metal edge looks like it would be in contact with the m4/3 adapter (I don't know, I haven't got it yet).

The french guy just mentions putting the baffle on, and I think it is implied that the baffle goes straight onto the glass (I'm fluent in french, so that's my take on it). But the baffle being cardboard/paper, I'm sure it would scratch the lens, no doubt about it in my opinion.

Any ideas or suggestions?

OP Antoine Palade Contributing Member • Posts: 589
Re: Question about BAFFLES to improve sharpness

No one?...

theunz Forum Member • Posts: 61
Re: Question about BAFFLES to improve sharpness

can't help much. i asked that question myself. I'm about to try to make one this weekend. From what i have looked at..i think you can apply some tiny tiny tape or cement to the side of the lens barrel...touching the metal a little. Again i have not done anything yet. Just got the black paper today.

Also...maybe you can just try to put it on the adapter.... that would be my second test.

good luck.

-- hide signature --

theunz

thinkfat Senior Member • Posts: 1,185
Re: Question about BAFFLES to improve sharpness

I'm going to give it a try soon. I'll probably put it inside the adapter, not onto the lens itself. Maybe I'll have to make the hole a different shape, though. If I put it very near to the sensor it should probably also have the shape of the sensor.

-- hide signature --

Everybody loves gadgets, until they try to make them
http://www.flickr.com/photos/thinkfat
http://thinkfat.blogspot.com

 thinkfat's gear list:thinkfat's gear list
Panasonic Lumix DMC-GH1 Canon EOS 5D Mark III Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Canon EF 50mm F1.4 USM Canon EF 85mm F1.8 USM +12 more
OP Antoine Palade Contributing Member • Posts: 589
Re: Question about BAFFLES to improve sharpness

theunz wrote:

Also...maybe you can just try to put it on the adapter.... that would be my second test.

Yeah, maybe...Although I think that the idea is to stop the light going diffusing around the back of the lens; I'm not sure to what extent putting the baffle away from the lens element might not defeat the purpose by letting stray light in. I'm about to send an email to the french guy who developed the idea, see what he thinks. Better go to the source. I'll post any answer I receive in this post, so keep an eye on it if you're interested too.

Cheers

-- hide signature --

theunz

Ehrik Veteran Member • Posts: 8,014
Search Oly SLR talk

I know people have experimented with this on the Oly SLR talk forum. So a search there might bring up something useful.

My impression was that people didn't see that much difference, if at all. But the 4/3 mirror box and the µ4/3 sensor box are different geometrically with different potentially reflective surfaces so may behave differently. It shouldn't hurt if done properly.

If it does have a visible effect, it should be there also stopped down since the image circle is always too big. Perhaps even more so since the illumination is more even.

Just my two oere,
Erik from Sweden

OP Antoine Palade Contributing Member • Posts: 589
Important Update on "how-to"

I sent an email to Christophe - the french person who developed the baffle idea. I asked him where to out the baffle? On the lens itself? And how to fix it in place?

Christophe's answer was to never put anything on the lens itself. He suggested fixing the baffle into the adapter itself, on the camera side (you were right, thinkfat!). He simply uses some carpet scotch tape adhesive to hold the baffle in place, but anything will do. The main idea, though, was for the baffle not to touch the lens.

There you go, that should put that one to rest!

Ehrik: thank you for your suggestion; I'll try the Olympus SLR forum too.

christan Regular Member • Posts: 233
Re: Important Update on "how-to"

Thanks for sharing the info ... appreciate it ...

Antoine Palade wrote:

I sent an email to Christophe - the french person who developed the baffle idea. I asked him where to out the baffle? On the lens itself? And how to fix it in place?

Christophe's answer was to never put anything on the lens itself. He suggested fixing the baffle into the adapter itself, on the camera side (you were right, thinkfat!). He simply uses some carpet scotch tape adhesive to hold the baffle in place, but anything will do. The main idea, though, was for the baffle not to touch the lens.

There you go, that should put that one to rest!

Ehrik: thank you for your suggestion; I'll try the Olympus SLR forum too.

-- hide signature --
Ehrik Veteran Member • Posts: 8,014
... to improve contrast

Ehrik wrote:

My impression was that people didn't see that much difference, if at all.

I also think whatever improvement one gets will be more related to contrast (including flare) than sharpness. Stray light from the unbaffled lens could find its way to the sensor and there brighten the darkest areas the most, relatively.

So to test the effect I would shoot targets with high dynamic range. Bright areas located at various places in the frame and also very dark areas (light traps) and then the relative difference between the light and dark areas with and without baffles could be measured.

Shooting with the sun just outside the frame could be very telling.

theunz Forum Member • Posts: 61
Re: Important Update on "how-to"

thanks for the update.
--
theunz

OP Antoine Palade Contributing Member • Posts: 589
Solution: Variable baffle

Form all this discussion, two things become clear:

1. A baffle is useful to improve sharpness and CA when using older lenses on 4/3 and m4/3 cameras.

2. Since there is a need to experiment with various hole diameters in the middle of the baffle depending on the type of lens used, it follows that a practical solution would be to build an adapter with in-built, variable baffle . In other words, an adapter with an in-built diaphragm, preferably 9-blades to achieve a roundness of the 'hole'. You can visualise this as a normal adapter, used to mount, say, an older FD Canon to your m4/3 camera; however, the adapter has also a ring that controls a diaphragm based on the diameter of the resulting 'hole', i.e. the adapter is indented to allow you to open the diaphragm at 12mm, 13mm, 14mm and so on, up until, say, 30mm. That way, you could vary the size of the 'hole' by rotating the ring on the adapter until you achieve the desired sharpness, much in the same way you control your aperture manually via ring on an older lens . This would eliminate the need to cut out a variety of baffles with different diameter holes in the middle, it would alleviate the problem how to fix the baffle, and it would make the whole process much more intuitive.

The issue is, of course, who will build it? It doesn't look like something I could do at home, that's for sure... Second problem - and potentially the biggest problem - is price. A simple adapter to mount various lenses onto your m4/3 camera is already costly if you're getting a good one. Add to that a diaphragm with 9-blades and you might end-up with something that makes the whole process undesirable because it might just cost too much.

slaine New Member • Posts: 24
My quick test

Hi all, I just went out to the arts and crafts shop and did a very quick and dirty before and after test with my Hexanon 40 mm. I am not sure how fair this test is as I do not have a tripod and therefore the two shots have very slightly different framing. Everything else was left the same though.

You can see the 100% percent crop out of the centre of the image here:

With baffle:

http://picasaweb.google.com/lh/photo/VRNglBHGQEE3OxEL6_vakQ?feat=directlink

Without baffle:

http://picasaweb.google.com/lh/photo/fJnUiq-KNuva3j8Kb1vsow?feat=directlink

duartix Veteran Member • Posts: 3,612
Re: Important Update on "how-to"

Antoine Palade wrote:

I sent an email to Christophe - the french person who developed the baffle idea. (...) He suggested fixing the baffle into the adapter itself, on the camera side (you were right, thinkfat!). He simply uses some carpet scotch tape adhesive to hold the baffle in place, but anything will do.

Thanks for clearing that. I've placed that question on the original thread a few days ago but it went unanswered. I was also thinking about baffling the adapter instead as it's less work, but for my Konica lenses I would need two adapters since the optimum diameters are very different for each one and still I would need to compromise a few...

I guess I'll baffle the lenses and put them on their original state as soon as I commute definitely from 4/3 to m4/3...
--
Duarte Bruno

 duartix's gear list:duartix's gear list
Fujifilm FinePix F40fd Olympus PEN E-P1 Panasonic Lumix DMC-GH2 Olympus M.Zuiko Digital ED 14-42mm 1:3.5-5.6 Panasonic Lumix G Vario 14-42mm F3.5-5.6 ASPH OIS +4 more
Tim F 101 Senior Member • Posts: 1,020
Clever idea

Hi Antoine,

It sounds to me like you will partially reproduce the effect that microscope engineers use to maximize sharpness in confocal microscopy. I use one of those for my day job, a Nikon, so maybe I can help.

A 'baffle' sounds like it should significantly reduce the effect of stray light inside the camera. Thus, the lower quality the lens (and adapter) the more improvement you should see. However, there is a chance that even a well-made baffle may reduce the apparent brightness of your image and necessitate a longer exposure, although I cannot predict by how much.

So why didn't camera makers try it before? I don't know, but I would guess that the diversity of focal lengths (see below) and the interfering presence of a flip mirror would make it impractical for most SLRs and any fixed-lens camera with a zoom.

The suggestion that older lenses will benefit more also strikes me as correct. 35 mm film lenses admit quite a bit of light that will not strike the 4/3 sensor and therefore can bounce unpredictably inside the camera.

Another commenter is also correct that you should attach the 'baffle' (microscopists call it a pinhole) at a variable distance from the sensor depending on the lens. To work best the baffle should be at the focal plane, which is exactly the distance in mm from the sensor as the focal length at which you are shooting. If the focal length is too long (e.g., 300 mm) then the focal plane will be inside the lens rather than the camera body. You can still try, but the hole in the baffle needs to be BIG to avoid vignetting and you won't get as much benefit.

The baffle should be matte black. A white baffle will bounce incident light around inside the camera.

The hole should be perfectly circular to avoid fringing and optical artifacts.

Test a number of different hole sizes, and don't be afraid to go small. At the focal plane light you want to detect should be focused into a very small space, so a small and perfectly centered hole should do it.

Good luck, and feel free to email me (portusjacksonii@yahoo.com) with any questions.

OP Antoine Palade Contributing Member • Posts: 589
Re: Clever idea

Tim F 101 wrote:

Hi Antoine,

...So why didn't camera makers try it before?...

I think they have done that for lenses that are supposedly being used within the same sensor format. Here, we're talking about using older lenses that were made to be used on a 35mm format, on the smaller sensor of m4/3. I don't think that lens manufacturers have taken this possibility into account.Thus, the issue of stray light bouncing around at the back of the lens. I know this is not a very scientific explanation, but I think I got the jist of it in principle.

Your point has some validity, though. Panasonic and Olympus - the makers of the m4/3 format - surely would have thought that people would be using other manufacturer's lenses with this flexible m4/3 format. Therefore, they could have devised some sort of way to "baffle" the sensor in a way similar to what you're saying. But I guess no manufacturer is really interested in promoting third-party lenses when they've got their own to sell. And theirs don't require a baffle...

Your comment about using a baffle at a precise distance from the focal plane is great info, but I don't know how precise one can make this happen in situations where we're talking about using bits of cardboard cut with a knife! The whole process of using baffles here seems very..ummm..."agricultural" in comparison to what you're saying. How this can be improved upon to actually adhere more rigorously to some laws of physics, I don't know. But for sure, it involves a higher degree of engineering than what cardboard baffling entails!

I also noticed some talk on the internet about baffles within the sky-watching community. It looks like telescopes are prime candidates for refraction and, maybe, solutions exist in their domain that could be transferred to the m4/3/ format - something a bit more advanced than cardboard and second-guessing. The same goes for the field of microscopy, as you mentioned. Maybe there is already some sort of ready-made pinholes, as you call them, that could be used/adapted for the m4/3.

Thank you anyway for you offer to help - let's see where this is going.

Franka T.L.
Franka T.L. Veteran Member • Posts: 8,161
Re: Question about BAFFLES to improve sharpness

Its very old and nothing more than a behind the lens aperture , and by reducing the peripheral ray ( that usually are lesser quality imaging vise ) anything that constraint the light this way can be used and had been used in many imaginative manner from way past .. its the same as stopping down your lens ( contrast, sharpness vise )

The most famous is the Lotus Root front insert of the Rodenstock Imagon , or in reverse the Silver center spot of the Leica Thambar . One can do that behind the lens as said, but its not convenient. The front filter with reverse grad effect or simply a very deep hood ( again cutting peripheral rays ) is much more practical ..

-- hide signature --

  • Franka -

Michael J Davis
Michael J Davis Veteran Member • Posts: 3,669
Re: Question about BAFFLES to improve sharpness

Antoine Palade wrote:

Context: this is a question about the use of cardboard baffles at the back of some legacy lenses, to reign-in stray lighting that results in hazy, David Hamilton-like pictures when used wide-open. The baffle is a round cardboard cut-out whose diameter matches that of the rear of the lens. At the middle of the baffle, a hole has been cut out, whose diameter varies depending on the lens. The resulting "focusing" of light reaching the sensor helps improve sharpness dramatically when used wide open. The idea has been developed by a french guy ( http://www.esperado.fr/fr/objectif_manuels/les_baffles_pour_les_optiques_manuelles_24x36_ou_moyen_format.html ) and some on this forum have also mentioned this solution when using legacy lenses (e.g. old lenses from film times) at full aperture (e.g. f1.4).

Sorry can't help with the glue question, but...

I'm bothered at the idea that sharpness can be improved by baffles. The optimum place for increasing sharpness is at the point of the diaphram (stop) - just stopping down will increase sharpness.

From my experience of lenses over the years I can tell you that a baffle may increase contrast and hence apparent sharpness. The light that is not directed towards the sensor can scatter around the camera interior and lighten the shadows and cause the highlights to blow. That's what the baffle is for. Therefore the shape of the baffle should be more in line with the shape of the sensor, not a circle.

You might like to conside the possibility of using a baffled lens hood (ie have the baffle on the outside - play with it until you just find that you are getting no additional vignetting. Again a mask the shape of the sensor would be optimum. (Think about it - that is what a lens hood is doing, plus preventing light from outside sources falling on the lens surface.

Mike
--
Mike Davis
Photographing the public for over 50 years
http://www.flickr.com/photos/watchman
G1 FZ50

 Michael J Davis's gear list:Michael J Davis's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Panasonic Lumix DMC-GX85 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +5 more
Ehrik Veteran Member • Posts: 8,014
Sharpness shouldn't change

Antoine,

Antoine Palade wrote:

Form all this discussion, two things become clear:

1. A baffle is useful to improve sharpness and CA when using older lenses on 4/3 and m4/3 cameras.

Sharpness and CA - this is unlikely if done properly (see below).

2. Since there is a need to experiment with various hole diameters in the middle of the baffle depending on the type of lens used, it follows that a practical solution would be to build an adapter with in-built, variable baffle . In other words, an adapter with an in-built diaphragm, preferably 9-blades to achieve a roundness of the 'hole'.

Ideally the baffle would be more rectangular to make a closer fit the sensor.

It's true that its size should vary with lens. A lens that has its rear pupil (where the light seems to emanate from when looking from the rear, not the same as rear element) near the sensor would need a smaller baffle than a lens that is more telecentric and has its rear pupil farther from the sensor. And a faster lens would need a bigger one.

That way, you could vary the size of the 'hole' by rotating the ring on the adapter until you achieve the desired sharpness,

But the baffle is supposed to stop the light that doesn't hit the sensor straight on, but misses it and bounces around the sensor box and eventually hits the sensor. This light should be diffused and would only affect large-scale contrast, not sharpness in general (once the global contrast is restored) and definitely not CA.

IF you are seeing an improvement in sharpness or CA, then you are blocking out the light that's reaching the sensor directly . This is not the normal intention of a baffle. You are using it as an aperture, effectively stopping the lens down (but not at the optimal place in the light path). Use the lens' built-in aperture instead.

When you see a change in sharpness or CA, or a loss of brightness (relative to a fixed exposure) also in the brighter tones in the image, then that is a sign that the baffle is too small.

I'd suggest unmounting the lens, open the aperture fully, point it to a bright, frame-filling scene and let the light shine on a flat white paper, orthogonal to the lens axis and the exact same distance from the lens as the sensor would be. Mark a rectangle on the paper, the size of the sensor and centre it in the image circle. Now you'll see the size of the image circle and can try out different baffles until you arrive at one that matches the rectangle (with a little margin of light outside). Test at infinity focus and closest focus to make sure. (And for the unlikely event it's a zoom, at the extreme zoom settings.)

Just my two oere,
Erik from Sweden

duartix Veteran Member • Posts: 3,612
Re: Clever idea

Tim F 101 wrote:

A 'baffle' sounds like it should significantly reduce the effect of stray light inside the camera. Thus, the lower quality the lens (and adapter) the more improvement you should see. However, there is a chance that even a well-made baffle may reduce the apparent brightness of your image and necessitate a longer exposure, although I cannot predict by how much.

Lets get this clear right away as there is no mention of it in this thread .

The baffle WILL slightly reduce the lens brightness wide open. If you are interested in knowing how much, here are some figures: http://forums.dpreview.com/forums/read.asp?forum=1041&message=32809706

-- hide signature --

Duarte Bruno

 duartix's gear list:duartix's gear list
Fujifilm FinePix F40fd Olympus PEN E-P1 Panasonic Lumix DMC-GH2 Olympus M.Zuiko Digital ED 14-42mm 1:3.5-5.6 Panasonic Lumix G Vario 14-42mm F3.5-5.6 ASPH OIS +4 more
Deep_water New Member • Posts: 6
Re: ... to improve contrast

Hi,

I also think that the baffle can't increase the sharpness unless it reduces the incoming light.

But I would still like to know, what really causes this "ghosting". Reflections of the light that doesn't hit the sensor can be ruled out, because then it wouldn't get better with stopping down.

For me it seems to have something to do with the lens-construction. For example on the biofos-page: http://biofos.com/cornucop/omz_ep1.html#SECTION3

you can see that the 85/2.0 does show a lot of ghosting, whereas the 90/2.0 doesn't. Both have almost the same specs...

If this ghosting didn't appear on film-cameras, but only on digital ones, it could also be some kind of deflection on the micro-lenses on the sensor, if the lightrays don't hit it at right angles. Just my thoughts.

Best Regards
Eduard Schmidt

Keyboard shortcuts:
FForum MMy threads