An interesting solution to eliminating moire and aliasing from your DSLR?

DSLRs

Moire hell!

Video grab with the Softar 1…without this would be moire hell
Still photo without softar

Spencer Voss contacted me today to tell me about the success he has had with eliminating the dreaded aliasing and moire so prevalent in Canon DSLRs. He uses a softening filter made by Carl Zeiss called Softar 1,2 & 3. These differ from traditional softening filters as they retain their sharpness and are not plastic!

This from the Zeiss website:

Expressing Emotion: The Zeiss Softar®

Motion picture directors of photography have been aware of it since a long time: for some applications, Zeiss lenses can be just too sharp! Close-ups of female beauties, for instance, do not really ask for maximum sharpness of every detail, do they? Instead, Hollywood’s goddesses expect to be given an aura of immaculate, radiant beauty… Introducing the Zeiss Softar®:

”Hang on to a Dream”
In the morning’s transition from a dream towards reality, when you look at the bright daylight outside the window in your darkened room, such a bright object will, for a few moments, appear with a radiant aura, a romantic glow of light around it. Like candlelight! The same visual appearance takes place, that same romantic glow shows when we look deeply into the eyes of a loved one…

Can this visual experience be expressed photographically? Today’s advanced camera lenses with their documentary, clinical, unemotional high-resolution rendering of object details do not lend themselves to supporting such emotional concepts like auras and romance and dreams. However, there has been a strong desire among photographers and cinematographers alike to add a sense of emotional warmth to some of their images.

As paradoxically as it may seem, no one has been more aware of this than Carl Zeiss. This is because Zeiss lenses are purposely designed to deliver the extreme opposite: perfect sharpness and total lack of romance. Since users of Zeiss lenses appreciate this typical characteristic for much of their work, but occasionally want to add ”dream-like romance” in a well-controlled way, Zeiss invented what we believe is the perfect solution: Zeiss Softar attachments.

Zeiss Softar attachments consist of a plane-parallel optical precision plate with tiny lenslets of varying size, randomly distributed over the surface of the plate. The optical surface is manufactured at the Zeiss Oberkochen plant with the same perfection applied to the optical surfaces of Zeiss lens elements. The quality of every single Zeiss Softar is verified on an interferometer, an optical precision measuring instrument capable of higher accuracy than any mechanical measuring device.

Unlike most softeners and diffusors the Zeiss Softar attachments are not mass produced pieces of plastic. This is why genuine Zeiss Softar attachments do not lose focus and do not produce unsharp out-of-focus images. Zeiss Softar attachments also do not lose important fine detail like eyelashes. Neither do they lose overall contrast and color saturation, which would result in dull images like the ones produced by so many softeners, ”soft lenses”, and diffusors. However, Zeiss Softar attachments reduce, lower, ”soften” the visibility of skin blemishes and freckles thus easing the life of the portrait and beauty photographer, saving retouching efforts, supporting the acceptance of his or her work and increasing the economic success.

Summing it up: Zeiss Softar attachments are designed and manufactured to the same exacting standards that apply for Zeiss camera lenses. They are not toys or substitutes, but professional tools for demanding photographers and cinematographers alike. In the hands of talented photographers Zeiss Softar attachments can lend an ethereal aura to portraits, bridals, wedding photography, and can even visually add emotion and romance to images of products, landscapes, and other subjects.

Zeiss Softar attachments don’t:
· produce fogged, dull photos with a total lack of brilliance, like most softeners, diffusers, and many soft-focus lenses do,
· lose eyelashes and other important detail in a portrait, like almost all other softeners do,
· lose overall contrast and color saturation, like most diffusers and other softeners do,
· create unwanted artificial optical patterns in the pictures, like softeners with concentric rings or textile fabrics do,
· limit the angle of view to ”portrait lens” angle only, like soft lenses do,
· limit depth of field by forcing wide and medium apertures to be used only, like soft lenses do.

Zeiss Softar attachments do:
· add a dream-like glow of romance, a sense of emotional warmth to an image without sacrificing sharpness or color saturation,
· produce a silky smooth appearance of human skin, with blemishes and freckles nicely ”retouched”,
· produce a softening effect that can be seen, evaluated and controlled on the • • viewfinder screen – at the full brightness of the wide open aperture,
· produce a softening effect that is the same anywhere in the frame,
· produce a stable softening effect over a very wide range of f-stops; depth-of-field can be used as usual,
· allow to focus the lens as easily as usual, be it with or without Softar attached at the full brightness of the wide open aperture,
· work with a variety of focal lenghts, including many wide angles,
· come in two different grades I and II, which can be easily combined into ”III”,
· come in a wide variety of sizes from 49 x 0,75 to 105 x 1, both Hasselblad bayonets 60 and 70, and Rollei bayonet VI.

Zeiss Softar attachments can be sourced from Hasselblad, Kyocera/Contax, Rollei, and from well-known filter manufacturers B & W, and Heliopan.

NOW, reading that blurb all sounds good and all, but I don’t want to make my subjects look younger, I want my moire to be minimised…that’s all!

The thing is they do give it a slightly different look and Spencer recommends the Softar 1 as it’s droplets are even and small and give the most pleasant image.

Now i haven’t tried these yet, but I have ordered a couple of second ones off of ebay. I will report back asap…i know there are anti-moire filters out there but I have to try them, be great to compare the Zeiss ones with these to see how they stack up if someone can point me in the right direction. Of course the perfect solution would be for Canon to work out a better way to downscale to avoid this. Using a softening filter is not ideal but these are very different to the softening filters I have used in the past due to their design…but hey, it’s pretty than nasty moire! Still am not certain I like the subtle promist look, it’s a baked in look that you are stuck with…anyway I am rambling and I will shoot something with one soon.

A comment has been added below that you can do this yourself with some clear filters and nail varnish…be interested to see if this works! I have ordered two clear Tiffen filters and will get my nail varnish out of the cupboard!

Here are some frame grabs from Spencer with and without the Softar.

With the Softar 1, notice the highlights on the car
With the Softar 1

58 comments

  2. Would have to see some video first, but I don’t really like the look of the stills … wondering if that footage was overexposed or if the blown highlights on the car (and the sky) were caused by the Softar …

    Using older lenses on a DSLR seems to be a better trick. I use mainly M42 lenses from different manufacturers (Mamiya Sekor, Zeiss Jena, Vivitar) which have little to no moiré while still retaining sharpness … and when I do have moiré, moving the focus ring just a tiny bit completely removes it without losing focus …

    I’ll try to do some tests later this week to show what I’m talking about.

  6. Interesting info and great to have a fix. But the price to pay for losing the moire seems to be that your footage may take on the ‘Emanuelle’ look.

  7. So by now it is known that the sharper the lens the more extreme the moire will be (if the conditions that will trigger moire are met of course) .There are too many filter options out there to disperse (“soften”) the light that enters the lens and each one of them can be used for various aesthetics.These filters are not like UV’s for moire,they are like special effects,you can not use them in every situation.I’d still like to test them but im pretty positive they will be useless in most if not all occasions.

  8. That’s is great news. I’m sort of hoping these issue will be sorted in the next 5D. In the meantime I’m loving the Panasonic GH1 with no trace of moire or aliasing with the 13 hack.

  10. Interesting. I wonder what the visible difference is between these and say a Tiffin Black Pro-Mist 1/4 which is what I’ve been using up until now?

  11. interesting, although could you not get the same effect by using magic bullet or some other post effects to add this ‘glow’ and softening later?

    applying this in camera cannot be undone later. I have been using old Nikon glass and like Pascal have had good success in eliminating moiré without the use of additional filters.

    I too would like to see some video.

  14. Ok, can I save you a LOT of money ?

    20 years ago we made filters like this when I was just starting out.

    Once you get past all the ziess BS, its nothing more then a peice of glass with dimples on it. you can make the exact same filter for $10. get any UV / clear filter that fits your lens, then using CLEAR nail polish make little dots. keep the center 1/4 of the filter more or less clear. thats it. super simple.

    I’ve been using some 138mm filters from the 70’s with great results too. promists are really the end of the line in the evolution of diffusers… go back to the earlier days and you’ll find some great filters out there that will eliminate or greatly reduce moire.

  16. Call me a skeptic but I would think that the nail polish would not be able to achieve the “Don’ts” listed above. You can get pretty cheap soft filters if you like already made but sharpness goes out the window. For somethings this is fine but there is a different in quality of glass.

  19. Hi Philip
    thanks again for letting the cat out of the bag. I’m going to try these out. Do you think they might work in combo with a polarising filter? I’m guessing some nice sharp technicolor feeling might materialise?

    Hope you’re doing well.

    Julian

  21. I’ve been using a layer of cling wrap on UV filters and/or the lens rear element to help reduce moire. This does reduce your contrast and sharpness overall obviously but not as much as you might think. Costco Kirkland Signature brand seems to have the least amount of blooming/streaking on hard highlights. Frustrating to apply but it does help like any other form of diffusion. I have the Caprock 4×4 filters and rarely find they help unfortunately – even stacked. Thanks for the tips Steve and Philip!

  22. I respect Philip’s work very much, and I appreciate that he shares his knowledge and experiences with us here, but really? I think the results from the stills are highly inconclusive, and this post reads like an advertisement (though I don’t think Zeiss is a sponsor of the blog, are they?). Especially when he hasn’t even tested it himself (yet)! Not to he mention the fact that existing “anti-moire” (not softening) filters exist, but they don’t work worth a damn.

    All I can say is, I hope this DOES work! I look forward to his results.

  23. This almost looks like the magic bullet looks filter “unbloom.” I have to buy one of these filters. Not to get rid of the moire effect, but so older ladies stop complaining about how bad their wrinkles look in HD! ; )

  25. Is there not a filter in FCP/Premiere that will do the same thing? You can just create the same effect in your workflow

    i may have missed the point 🙂

  26. Bloomy filters, tested by Philip Bloom! 😀

    Seriously, I’m looking forward to an extensive test of yours!
    I hope it doesn’t turn out very ”dreamy” of ”romantic”… (some like that, but I rather see sharp clear images)

  27. i got this tip from…Philip Bloom!

    If you’re getting moire, try moving further/closer if you can. Doesn’t always work, can’t do it all the time, but it’s free!

    Thanks again Philip!

  28. Haha, sorry but I kinda laughed reading this. What about… buying a fantastic vintage lens for less money than the actual FILTER? (Oh wait, don’t btw, prices will go up).

  30. Hi Philip,
    I have been using a secondary OLPF I had made in china the specs are:

    Quartz crystal OLPF (AAFilter), 2 layer 0° & 45°cut size 40mm (w) x 25mm (h) x 3.38 with AR coating.

    It sits behind the lens in the camera body on a plastic sledge. Installed when the mirror is up. Takes only a couple of seconds to install, works on any lens as its filters the optical image before it hits the rear elements.

    You don’t get the blooming softness effects, its not 100% but It knocks the edge off and gives a nice cinematic colour.

    I’ll bring this along next Saturday in the Londonia meet.

    Here is a film made with this secondary filter.
    http://www.vimeo.com/13706907

    BR, James

    2. James,

      Any additional information you can provide on that OLPF filter would be very helpful. Is it an already-available commercial product, or did you have to also have the plastic sledge custom-made?

      An APS-C-sized OLPF that blurs the image enough for DSLR video sounds like it’d be expensive.

  31. This filter creates an interesting look. It would be great if we could see some video of you using this filter. Specifically portraits of someone, so we can see the before/after effects on skin and clothes.

      1. Well another filter that’s able to do similar diffusion while maintaining sharpness are Mitchell Diffusion filters. I have been using them for 30 years and they have been around a lot longer than that. Think Mitchell Camera. The construction is really similar, a pattern in the glass not between the glass. The pattern isn’t random but regular but still works and makes older women gorgeous.

  36. Dear Philip,

    i saw the post of my tests you’d put on your website
    nice!

    i’d made bad example of a video still from a 1920-1080HD made by the Canon 5DMKII without any filter on it

    it was just a day in the city when the rain stopped, 
    the sky cleared and a rainbow came up

    i knew that using no filter whatsoever would result in at least 4 moire patterns on tiles and bricks

    I found this link about another way of getting rid of moire with test results

    http://www.dvxuser.com/V6/showthread.php?t=200980
    http://www.dvinfo.net/forum/canon-eos-7d-hd/480602-controlling-moire-7d.html

    Indeed the Caprock filters

    I did a test today with ND4 and also changing of the sharpness inside the canon 5dmkii
    will send you the results stills from the video using wetransfer.com

    findings

    my situation:
    i live in the tiny country of holland where all is made out of bricks and tiles
    most unfortunatly these are everwhere except in nature
    i use the canon 5dmkii for hdvideo commercials, virals and the likes

    the canon gives me nasty moire patterns when filming masters and wide shots (of the bricks and tiles which appear in all cities)

    I searched and found some solutions to get rid of it,
    as posted by you on this blog

    the first is a softening filter called softar 1 from zeiss, they are in 3 grades: softar 1,2 and 3, they soften but also get rid of the moire which makes my material un useable.

    softar 1 is ok
    softar 2 a bit softer
    softar 3 extreme results

    the second i found is to use a ND4 filter,
    this gives me the possibilty to get my f stop lower (so shallow dof)  and still film in the timing of 180 degrees. so on t45.

    the third solution is to lower the sharpness settings inside the camera, in canon  5dmkii that is in live view, choose picture styles, and change the sharpness

    ofcourse changing the contrast as well could do but for me that was not part of my test

    so my results are as such:
    indeed the moire as far as I have seen occurs at a specific resolution in real world environment,
    with that i mean : in the test i did with the canon 5dmkii yesterday 
    from the same point filmed, there was a moire in just 1 bit of the brickwalls in the right side of the frame

    when i pointed the focus of the camera on a point closer to the camera:
    instead of lets say 50metres, to a  point of lets say 15 metres
    and using a ND4 to lower the DOF (in the same iso100, timing 45 (180 degrees))
    so the depth of field did not reach the point of moire, the result was imminent, 
    no moire: at that specific point of moire at 50 metres

    so using an even more heavy ND, lets say ND4 + ND2= ND6 would bring the f stop all the way to 1.8 or 1.4 
    resulting in no moire 

    but when i put the focus on that specific point of moire
    the results were dissatisfied, very hard to get rid of the moire at 50metres on the brick walls, 

    it seems that that specific point on that wall, on that specific distance is kind of the same resolution as the 1920-1080 or of the scanning of the 22MP chip
    when turning the focus a bit further or a bit closer (with ND4) the moire disappears
    so the moire is, in my opinion, indeed a case of two resolutions: the brick walls at 50metres and the scanning matching very close to each other

    softening the frame a bit works 
    changing the sharpness of the canon works at sharpness 4-5

    zooming inside the frame before into recording did not give me a workable result of checking for moire
    so even perhaps writing the data down to cf using the compressor may result in a bit of extra moire

    indeed a soft or vintage lens, as old lenses are, can make a whole world of difference
    in film we would always choose Cooke S4 lenses when Arri/Zeiss were too hard

    what i also tested is the order of the filters
    when using a softar the best is to have the filter as close to the front lens as possible
    the tests i did where with a standard lens : canon 50mm f1.8
    this lens has about 15 mm from the filter screw side on the front of the lens to the glass front end of the actual lens itselve

    another lens i have is the 50mm carl zeiss 1.4
    this lens has only 5mm distance measured from the front of the filter screw side to the front of the lens

    the softar will be much closer to the carl zeiss lens than it is right now to the canon lens

    what i did was
    changing the softar in order
    so 

    1; lens, 
    2nd:ND4
    3th: softar
    this is a big difference 
    too soft and much to sofatr effect

    changing the order to
    1: lens
    2:softar
    3: nd4 

    is a much better result
    while the difference in this case is only the screw of the filter itselve,lets say 5mm

    changing the filter all the way to the fron of the lens as in case with the carl zeiss would make the filter even less visible
    while maintaining the positive results

    new canon usm lenses also have their front lens as close as possible to the screw side in front of the lens

    spencer

  37. Not to attempt to comment on what is happening technically with the Softar, but at first glance, the grabs are, to me, reminiscent of the DOF adapter look as it’s shot through the ground glass… not a look I’m after.

  38. Philip,

    Just tested comparing softar vs caprock,

    Softar is basically useless with wides ( 11-16, 20-40 ) also useless with 50mm only worked with my 80mm mamiya f1.9 but barely.

    Caprock 1.0 works with 11-16mm at 15mm and 20-40mm f2.8. Hazy with 50mm.

    I’d conclude this softar is not a good investment.

    Ted

  39. Any update on this?

    Has anyone noticed a obvious improvement using these filters? I’ve got a big shoot in the next couple of days and I’m deciding whether or not I should pick some up.

Leave a Reply