Jump to content


Photo

Tiffen "Diffusion filter" and "Soft FX" filter


  • Please log in to reply
14 replies to this topic

#1 Bryce Lansing

Bryce Lansing
  • Basic Members
  • PipPip
  • 73 posts
  • Cinematographer

Posted 01 November 2009 - 03:03 PM

I found a filter on eBay listed as "Diffusion #2 filter", but I'm confused, is this the same filter as a "Soft FX #2 filter"? If not, what is different?
  • 0

#2 Chris Keth

Chris Keth
  • Sustaining Members
  • 4427 posts
  • 1st Assistant Camera
  • Los Angeles

Posted 01 November 2009 - 03:54 PM

I found a filter on eBay listed as "Diffusion #2 filter", but I'm confused, is this the same filter as a "Soft FX #2 filter"? If not, what is different?


It's not the same. Tiffen currently lists gold diffusion and black diffusion. No plain old diffusion. My guess is that it would look something like a white pro-mist. You can call tiffen and ask. They're a very approachable company.
  • 0

#3 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19765 posts
  • Cinematographer
  • Los Angeles

Posted 01 November 2009 - 04:53 PM

It's not the same. Tiffen currently lists gold diffusion and black diffusion. No plain old diffusion. My guess is that it would look something like a white pro-mist. You can call tiffen and ask. They're a very approachable company.


There was an old Harrison "Diffusion" filter that was rather misty, but the Tiffen Gold and Black Diffusion-FX are very different and are modern designs, newer than their Soft-FX filter.

Basically a diffusion filter works on the principal of throwing small areas of the image out of focus while allowing sharp areas to pass through unaffected, thus overlaying an out of focus image over a sharp image (as opposed to just blurring the entire image, which is basically just an out of focus image.)

The classic example is a fine net stretched over the lens -- the threads of the net cause light rays to be diffracted, defocused, whereas the gaps in the net allow light to pass through unaffected.

Glass diffusion filters use some sort of defect, indent, bubble, whatever, in the glass to blur that part of the image and then leave clear areas that allow sharp detail to pass through.

Now this is similar to how a "mist" or fog/low con filter works, which has some sort of mist particle suspended over the filter that causes light to spread, and also somewhat softens the image (some more than others.) It's just that with diffusion filters, instead of a light scattering particle, you have an optical indentation or something to cause that area to become defocused.

The trouble is that in some ways, this pattern of indents or bubbles or "lenslets" also act to scatter light and thus also produce some of the effect of a mist filter -- halation around lights plus lowering of contrast.

Ira Tiffen worked on creating a diffusion filter that would merely reduce fine detail without much other visible artifacts, i.e. minimal to no halation, no loss of contrast, just softening. At the time, two common glass diffusers on the market were the Tiffen Soft-FX, the older Harrison Diffusion, and the older Mitchell Diffusion. I think Ira may have also designed the Soft-FX. There was also the Harrision Black Dot Texture Screens at the time.

Anyway, the Harrison Diffusion (seen to good effect in the early scenes of "Malcolm X") has a particularly foggy look. The Mitchells do not. And both the Harrison Diffusion and the Mitchell tends to throw too much of the image out of focus, there is a feeling that nothing is quite in focus.

The Mitchells have a sort of trapezoid pattern indented into the glass, and Ira told me that he felt that the straight edges of the pattern were part of the cause of the blurriness issue. But that pattern is probably also why Mitchells don't create a misty look.

The Schneider Classic Softs, which I think came out around the same time as the Black Diffusion-FX, use a round lenslet in a regular pattern. The lenslets are rather large as well. This large lenslet size and regular grid pattern is a bit of a problem if you stop down too much or use a format with a lot of depth of field because the pattern starts to become visible. Schneider later solved this with their "HD" version, which uses smaller lenslets (and they may be in an irregular pattern).

Ira Tiffen basically looked at the Tiffen Soft-FX, which did a lot of nice things with diffusing the image (it uses a sort of kidney-shaped lenslet in an irregular pattern) -- basically an attempt to improve over the Mitchell diffusers I believe, or was a replacement for an earlier Tiffen Diffusion filter -- and refined the idea further to reduce classic diffusion filter artifacts. He added a pattern of black dots to restore the contrast / blacks (similar to how black specks work in other filters like Black ProMist, etc.) He made the lenslet patterns small and random.

So the final effect is a filter that reduces fine detail without making the picture look blurry, and without halation nor loss of contrast.

Personally I think the Tiffen Black Diffusion-FX is the best filter for softening a face in an otherwise unflitered movie, i.e. the shot does not end up looking filtered. Next closest thing is probably a Schneider HD Classic Soft or the lightest Mitchell Diffuser ("A" I think.)

However, I also think this makes the filter a bit "boring" and technical -- when I want some level of diffusion artifacts, like halation, I tend to pick another filter. Lately I've been trying the Schneider Classic Soft Blacks (which is a regular Classic Soft + a 1/8 Black Frost) or their Hollywood Black Magic (which is an HD Classic Soft + 1/8 Black Frost.) The little amount of Black Frost gives the image a mild misty halation, plus with the Classic Soft Black, you get that Classic Soft artifact, the blurry bubble effect around points of light.

The Tiffen Soft-FX is very nice as well. My main problem is simply that the levels are too spaced apart, the jumps between the 1/2 (the lightest), the 1, and the 2 are too obvious. I like to use no filter on the wide shots, the lightest on the medium shots, and the next heaviest on the close-ups, so the diffusion sort of sneaks up on you. You can do that with the Soft-FX but I wish the strengths were slightly closer together.

The Harrison Black Dot Texture Screens, which came out in the late 1970's and were used famously on "Excalibur" and many other movies shot by Alex Thomson around that time, were probably the first to use black specks to restore contrast and blacks in a diffusion filter. The trouble was that the black speck material was rather randomly applied, making it hard to be consistent, so you had to buy the whole series of strengths to get the right progression. I'm not sure what the diffuser pattern technique was, though it was some sort of rippled glass pattern or something, you can see it when the camera pans in some shots in "Raw Deal", shot by Alex Thomson.
  • 2

#4 Juan Siquot

Juan Siquot

    New

  • Basic Members
  • Pip
  • 2 posts
  • Cinematographer
  • Paris, France

Posted 05 November 2009 - 04:20 AM

Thank you David for this exhaustive answer about filters. It was very useful.
  • 0

#5 Miguel Bunster

Miguel Bunster
  • Basic Members
  • PipPipPipPip
  • 301 posts
  • Cinematographer

Posted 08 November 2009 - 03:44 PM

Thank you David for this exhaustive answer about filters. It was very useful.


Do you feel the Soft-FX 1 lowers "sharpness" or "focus" to much in wide shots and mediums? I am using it on the F900R on a Fujinon HD lens for a really low budget movie and I like how it feel but haven't been able to see it on the big screen. I tested it in different frame Sizes and I see the effect but felt it wasn't a to much of a compromise in the image vs diffusing in a nice way.

Just curios to see how you feel about it, if it will make the image overall feel a tad out of focus the #1 Soft FX.
thanks!
M
  • 0

#6 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19765 posts
  • Cinematographer
  • Los Angeles

Posted 08 November 2009 - 05:15 PM

The softening of a Soft-FX is similar to a Classic Soft or black net in the degree of fuzziness it adds - I would use no filter for extreme long shots, a #1/2 Soft-FX (the lightest) for wide and medium shots, and a #1 for medium and close shots. It partly depends on shot contrast... a backlit shot with a lot of edge highlights increasing the feeling of sharpness, or a hard-lit shot, can take more diffusion and still feel sharp compared to a low-contrast soft-lit shot with no strong lines or edges. And since backlight also creates a nice halation with diffusion filters, I tend to feel that it's a good combination. So I would use a #1 Soft-FX on a medium shot that is backlit or has strong edges, or a hard sidelight, etc. but I would use a #1/2 Soft-FX if the shot was low-contrast and soft-lit with no strong lines or edges.

A lot of this is personal taste and gut feelings, I try to not get too technical about diffusion filtration, it's sort of a feeling you get about a scene, etc. Just remember that you can always soften a shot more in post if necessary but it's harder to undo too much diffusion. On the other hand, no guts, no glory...
  • 0

#7 Miguel Bunster

Miguel Bunster
  • Basic Members
  • PipPipPipPip
  • 301 posts
  • Cinematographer

Posted 09 November 2009 - 12:59 AM

David,
Thank you for your response!

I agree with everything you say but I came to a situation where I cant really do it in post or hope (I know how to do it and did post myself for some time) because is a small movie and the "DI" will be limited and I will be lucky to color correct properly. That said, my concern came from the right term you used "degree of fuzziness" the Soft-FX #1 would do. I don't have a 1/2 so my only choice is #1 which i like a lot and have been able to judge only from the 17inch HD monitor. I been using it in all the shots because as mentioned I am pretty sure doing it in post wont be an option and I want to achieve a certain softness and halation on highlights which are in a lot of over expose windows mainly and as well soften the skin.

I am shooting on the F900R with a Fujinon HD Eng lens wide open, and with a low contrast setting and detail to "0". Lighting is high contrast but mainly done with big soft sources from windows, no hard light on people at all, on the surroundings yes. I feel when I bring back the contrast and a tad of sharpness the fuzziness will go away (I do not see much fuzziness beside one shot that rised my concerns). I feel the #1 is really subtle overall.

In your experience do you feel a #1 overall would create to much "fuzziness" that would compromise the image (in wide shots and mediums shots) and make it feel too soft that looks out of focus? I don't feel is the case and most is pretty contrasty between key and fill but done through soft sources.

As you said no guts no glory and I feel is worth using the filter in the case of this movie and I preferred it of BPM's in this particular film. But a more seasoned experience recommendation would be great.

Thanks you and best.
Miguel
  • 0

#8 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19765 posts
  • Cinematographer
  • Los Angeles

Posted 09 November 2009 - 10:52 AM

I'm sure the #1 Soft-FX will be fine, HD tends to be edgy and cut through a certain amount of diffusion, especially if you aren't shooting in LOG but in Rec 709. But why can't you get ahold or buy a #1/2 Soft-FX?
  • 0

#9 Miguel Bunster

Miguel Bunster
  • Basic Members
  • PipPipPipPip
  • 301 posts
  • Cinematographer

Posted 13 November 2009 - 01:56 AM

I'm sure the #1 Soft-FX will be fine, HD tends to be edgy and cut through a certain amount of diffusion, especially if you aren't shooting in LOG but in Rec 709. But why can't you get ahold or buy a #1/2 Soft-FX?


David,
cause its low budget......so low and i got this filters that don fauntleroy lent me as a favor and really like whatthe #1 is doing but cause i am lighting low con i was concerned. But thank you for your input and time. Things are looking good and made some mini tests where i increase the contrast (which was planned) and helps so all good.

On another note I am curios but i tested the ASA of the camera by exposing to a gray card (quick way) with my low ocn setting in hyper gamma 3 and was getting a 160 asa. In other cameras I was always getting more around 320. Have you came acroos this rating on a f900r?

thank you
m
  • 0

#10 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19765 posts
  • Cinematographer
  • Los Angeles

Posted 13 November 2009 - 03:21 AM

On another note I am curios but i tested the ASA of the camera by exposing to a gray card (quick way) with my low ocn setting in hyper gamma 3 and was getting a 160 asa. In other cameras I was always getting more around 320. Have you came acroos this rating on a f900r?

thank you
m


I haven't used the F900R or shot in HyperGamma 3 but if HyperGamma is similar to LOG, they tend to increase highlight detail by placing "white" (Zone 10 I guess) at 70 IRE instead of 100 IRE in order to record "super whites" (overexposure detail.) Therefore the image seems a little underexposed because the highlights aren't as high up in IRE level as they would in normal Rec 709. You can compensate by overexposing and thus rating the camera slower, but you lose some of that advantage in overexposure information being recorded.
  • 0

#11 Miguel Bunster

Miguel Bunster
  • Basic Members
  • PipPipPipPip
  • 301 posts
  • Cinematographer

Posted 13 November 2009 - 12:48 PM

I haven't used the F900R or shot in HyperGamma 3 but if HyperGamma is similar to LOG, they tend to increase highlight detail by placing "white" (Zone 10 I guess) at 70 IRE instead of 100 IRE in order to record "super whites" (overexposure detail.) Therefore the image seems a little underexposed because the highlights aren't as high up in IRE level as they would in normal Rec 709. You can compensate by overexposing and thus rating the camera slower, but you lose some of that advantage in overexposure information being recorded.


David,
\thanks you for the reply. Yes what you say makes sense and the asa has been changing a bit but is Between that range. About the diffusoin in wide shots you are right is not the best to drop it there but what really kills it is the quality of the fujinon HD eng lens we are using is really bad...looks so soft in wide shtos wide open so trying to work around it.

The othter thing i am trying to figure out (this movie i couldn't test the camera but a day before) is how grainy it is in the lower range of the latituted. Way more grainier than other f900's i used. I am using the Hyper gamma 3 plus the BBC settings (like 0.45 gamma) to get more info on the recording to nano flash (1920x1080 4:2:2) but for example on well exposed scenes at day there is grain in the hair of the actors! even though i am plannin to bring the blacks back and this reduces the grain i am impressed how grainy this is. I am curios if this is a producto fo the low con setting where i am not recording true blacks (flat liner black in the wave form).

Best!
M
  • 0

#12 Tony Brown

Tony Brown
  • Sustaining Members
  • 689 posts
  • Cinematographer
  • London

Posted 17 November 2009 - 01:01 PM

Be careful with Mitchell diffusers. There were two types (they latterly became Panchro diffs)

The early diffs had an irregular pattern etched into the surface of the glass, not unlike a mutton cloth fabric type pattern. The later ones were a very regular diamond etch

The early ones were wonderful, they are very rare, I've been trying to find a set for 20 years. Movietech in London have a set of 3x3 which I have used but I would kill for a set ( or even just the A & B ) of larger ones

The diamond etch are very ordinary

Then of course there were Mitchell sliding diffusers, literally a long variable diffuser that you slid through the matte box until the desired amount of diffusion was achieved. We had them at Panavision in London around 1976.....
  • 0

#13 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19765 posts
  • Cinematographer
  • Los Angeles

Posted 17 November 2009 - 01:06 PM

Be careful with Mitchell diffusers. There were two types (they latterly became Panchro diffs)

The early diffs had an irregular pattern etched into the surface of the glass, not unlike a mutton cloth fabric type pattern. The later ones were a very regular diamond etch


Yes, Ira Tiffen told me about the diamond pattern in the newer Mitchells, he felt that they led to more of the image looking out of focus than it should, which was my impression on the only feature where I used Mitchell diffusion, it just tended to look like I missed the focus.
  • 0

#14 David Rakoczy

David Rakoczy
  • Basic Members
  • PipPipPipPip
  • 1579 posts
  • Cinematographer
  • USA

Posted 17 November 2009 - 04:56 PM

I worked with an awesome 1st AC (Chris Taylor) for years and years and he always referred to 'Soft FX' Filters as 'Soft Focus' Filters...

I tried them once and found their ability to defocus an image (way) too powerful.. I did not like the effect at all! Never used them again.. but that doesn't mean they don't have their 'place'... it's just not in my particular Filter Case. ;)
  • 0

#15 Rob Vogt

Rob Vogt
  • Basic Members
  • PipPipPipPip
  • 437 posts
  • Other
  • New York

Posted 17 November 2009 - 09:17 PM

Hey David, Glad to see you back! Also Miguel, just wondering your reason for not wanting to use BPM in your situation is?
  • 0


FJS International, LLC

CineLab

Wooden Camera

Paralinx LLC

Visual Products

CineTape

Opal

The Slider

Abel Cine

Gamma Ray Digital Inc

Aerial Filmworks

Broadcast Solutions Inc

Technodolly

Media Blackout - Custom Cables and AKS

Glidecam

Willys Widgets

Metropolis Post

Tai Audio

Ritter Battery

rebotnix Technologies

Rig Wheels Passport

Wooden Camera

Aerial Filmworks

Tai Audio

Opal

CineLab

CineTape

Ritter Battery

Broadcast Solutions Inc

Abel Cine

Metropolis Post

Rig Wheels Passport

Visual Products

rebotnix Technologies

Media Blackout - Custom Cables and AKS

Paralinx LLC

FJS International, LLC

Glidecam

The Slider

Gamma Ray Digital Inc

Willys Widgets

Technodolly