Diffusion Filters (Test Comparisons on Red)

[Roberto Lequeux]

I thought I'd start a small thread where people could post any test comparisons they may have of good diffusion filters on Red.

Goggling for a specific filter comparison frustrated me, and I thought if there were a good thread like this it would be such a great resource. All we need now is some generous folks willing to share the same shot/s, with and without the filters they tested.

Thanks!

 

[Andrew Wilding]

This is a good idea for a thread - it would be nice to have a resource like that. It would be unscientific, of course, with everyone who contributes shooting under different conditions, but it could be a nice broad overview of the general effects of certain filters. Of course, nothing beats renting a boatload of filters for a weekend and shooting like mad, though.

 

[Dan Hudgins]

Tiffen LC-5

Tiffen LC-5

I wonder if the Tiffen LC-5 will make things better or worse. I know on film it cuts the highlights down and pulls up the shadow detail, but also gives the image an over all "mushy" look. So testing should be on high contrast lighting, like in a forest, and with a matte box in tight, maybe with a 35mm lens?

 

[David Mullen ASC]

The Tiffen Low Con filter does not cut down on highlights unless you compensate by underexposing, it's like any other "mist" filter, it allows highlights to fog the filter and lift the shadows. The way those mist particles are made, their size and dispersion, create different looks and amounts of softening.

The Low Con is closer to a Fog and Double Fog in effect in terms of blooming (halation) but then Tiffen created the Ultra Con, which has almost no halation artifacts, just the lifting of the blacks.

Schneider added black specks to what is essentially an Ultra Con (and I think what they call a Low Con is closer to Tiffen's Ultra Con) to create something they call a Digi Con, which does darken the highlights a bit. I think Tiffen may have an older Low Con design that does this too called a Soft Con.

Technically, low-contrast filters are not diffusion filters, though they do cause some softening and it's not a big leap to calling them diffusion with things like Fogs and Double Fogs.

You can find some filter examples here:
http://www.camerafilters.co.uk/

Just remember that anything viewed on a computer screen is not going to emulate the effect of seeing that filter on the big screen. If I posted an example of a 1/8 ProMist, for example, you'd think that it wasn't doing anything yet on a big screen, you can still see the effect. For this reason, when testing filters and showing them on a computer screen or printed in a book, you often have to go quite heavy just to see the effect.

Personally, after testing both flashing and low contrast filters on film over the years, there is never more than a true 1/3-stop increase in dynamic range information at best, after which, most impressions of increased shadow detail and dynamic range are just due to the lifted blacks.

 

[A. Bastaki]

David, you are a gift to mankind that has dropped from the heavens.

Thank you, Thank you, & Thank you.

 

[Roberto Lequeux]

David, you are a gift to mankind that has dropped from the heavens.

Thank you, Thank you, & Thank you.

I completely agree. Be careful though, he doesn't like big compliments too much. :smile5:

Personally, after testing both flashing and low contrast filters on film over the years, there is never more than a true 1/3-stop increase in dynamic range information at best, after which, most impressions of increased shadow detail and dynamic range are just due to the lifted blacks.

David, is there any sharpness loss with the filters that increase dynamic range at all? If so, how would you describe it?


"The filter also has no apparent effect on the image sharpness." from http://www.camerafilters.co.uk/page48.html


Since I am mostly concerned with Blu-Ray and will be shooting 4k+ there is a lot of sharpness to spare. So would it be wise to use one of these for a LARGE portion of a movie if you were also concerned with detail?

Sorry if my questions are not great ones. I am just trying to learn about these dynamic range filters so I'd greatly appreciate you sharing anything you think is important or pertinent.

 

[Dan Hudgins]

LC

LC

The Tiffen Low Con filter does not cut down on highlights unless you compensate by underexposing..

I'm not sure if the ones we have are H&H or Tiffen, but if you put it in front of the lens without chaning the exposure the highlight density goes down (and shadow density gos up), it can be measured, it would have to since if light is moved into the shadow areas there is less for the highlights. Its easer to see on high contrast negatives I was using it on 2302.

 

[David Mullen ASC]

Are you sure it's labeled "Low Con"? Sounds like a Schneider DigiCon or Tiffen Soft Con to me...

H&H or Tiffen Low Cons don't normally have any sort of density that would lower highlight exposure (though perhaps the mist particles do create an overall extremely mild drop in exposure). Take a look at the filter itself -- do you see any sort of dark specks or dots in it? Or just the usual misty specks?

The old rule with mist filters, low cons, was to slightly underexpose to counteract some of the washing-out effect in the bright highlights, which creates the illusion of increased exposure similar to how smoke can do that. Not that I followed that rule...

Again, I don't think these filters do much to improve dynamic range once you reset the blacks in post, not enough to be worth one more filter in front of the lens. When I use Low Cons, it's when I want a pastel Low Con look -- for example, "Barry Lyndon" was mostly shot with a Low Con #3:

However, sometimes I feel that any of these filters that have black specks or dots to darken the highlights (Black ProMist, Black Frost, Black GlimmerGlass, Classic Black, etc.) make a digital image feel a bit more like film, perhaps it's the flattening/darkening in the highlights combined with the mist particles adding a noise/grain texture. Sometimes I don't feel that way and I think people use them too much as a crutch to get a "film look".

Traditional Low Cons do slightly soften and cause halation. Ultra Cons do much less of either, but since contrast and sharpness are tied together, you can't really lower the contrast without also lowering the apparent sharpness. Also, one problem with UltraCons is that they flare easily, like when panning across a lake when the sun suddenly glances off of the water and hits the lens, the entire filter milks out momentarily.

Though it's hard to tell here, it's pretty subtle, I used a filter similar to an Ultra Con called a Smoque filter for a scene in "Big Love":

 

[Stephen Pruitt]

Hi there, David. . . beautiful images all. . .

If you were forced to select just one of the Smoque filters, which strength would it be? I guess I'm going for a noticeable effect here, but I sure don't want to over do it.

What do you think of the Schneider Classic Soft filters? Are they useful?

Thanks much.

Stephen

 

[David Mullen ASC]

Smoque filters are meant to simulate a haze around light sources. I used a #2 Smoque here for this scene in "The Astronaut Farmer" (the actors were getting sick of the smoke...):

The trouble with the Smoque filter is that it reacts to lights in the frame, so when an actor momentarily blocks that light or the camera pans off the light, the effect disappears.

Classic Softs are diffusion filters, they soften definition by using lenslets to throw small areas of the image out of focus around the sharp portions. They are not really contrast-lowering filters, they are more like using nets, for example. "War of the Worlds" used nets and Classic Softs, on the heavier side. You see here:

NET

CLASSIC SOFT

 

[Dan Hudgins]

LC set

LC set

Are you sure it's labeled "Low Con"? Sounds like a Schneider DigiCon or Tiffen Soft Con to me...

They are a set of LC-1 to LC-5 in 4x4, and I have some 2x2 that look about the same. Those marks are etched into the corner of the filter. I also have some D-1 to D-5 filters which I guess means diffuser, those are H&H. The D-5 is so strong I cannot think of a use for it, but D-1 looks nice a soft glow around bright points. I also have F- filters which I guess means Fog filter. The only ones I use much are the LC- filters, I don't like the Fog filters much, maybe for use with bleach bypass ECN or something like in Quintet (1979).

And yes they cut the highlight density, mid tone density probably stays the same, and shadow density comes up, as does a slight veil.

These filters are uncoated, so that in itself will reduce transmission to between 95% and 90%. Also they look somewhat milky so some of the light is being reflected as well as being scattered. So maybe they transmit 90% to 85% so there will be a loss of highlight density in the negative. The glass use in some of these filters is made out of is a little green so that can cut another 5% off the red end to make the filter transmit maybe 80% to 75%, so the highlight loss is about 1/6 to 1/4 stop plus the effect of the scatter.

We had a custom glass CC50M filter made up, but the green glass used added to the magenta losses so the cut was more than maybe 1/4 stop over using a CC50M gel filter.

The point of the low con is to reduce highlight density and rase shadow density, which they did in the tests, but the results on black and white look a little more than misty, its not a good look for some subjects I thought. 2302 is a high contrast film, so the density changes were many times what you could measure on ECN.

Any filter will cut highlight density, even a UV filter, because it does not transmit 100% of the light.

With the RED ONE you get some other issues, since the de-mosaic software will probably auto-adjust the highlight clip points and maybe also the black levels. DSLR probably also compensate in the de-mosaic, at least for their JPG mode if not the RAW mode.

Because the white clip will adjust the top of the histogram, you may not see the loss from the filter except that the noise may come up a little.

There are also some effects on what lens (gauss or retrofocus) is used and how close the filter is to the lens, and the f/ stop used, and the matte in the matte box, and how bright lights are that are in the frame, and what filter number is used, so you cannot say from one shot to another how much up or down a gray card in the shot will increase or decrease in exposure, the more bright lights in the frame the more the shadow areas will come up. Also you have to keep direct light off the filter since it will show shadows of the edge of the matte, or flare if it gets hit with strong direct light from out of the frame.

 

[Roberto Lequeux]

Thanks for your answers David.

I really love the look of that Astronaut Farmer frame David. I wonder what it would look like used out in a sea of highlight on an exterior with snow all around. But I am thinking that I might be able to get most of that in post quite well. I would love to see it on the monitor on set though, and for dailies... must... get... great... D... P...

 

[David Mullen ASC]

From a cinematography standpoint, Low Cons basically just fog the image slightly and you do not treat them as having any filter factor. I can't say how they are considered from a post standpoint though.

They are related to the Fogs, just that the mist particles are finer, and with Ultra Cons, even finer and thus throw less of the image out of focus.

The Harrison Diffusions (D1, D2, etc.) aren't used much anymore -- the last major movie I can recall was "Malcolm X", the early scenes. It's rather foggy for a diffusion filter.

Harrison Double Fogs are basically Low Cons combined with Fogs, they aren't twice as heavy Fogs. "McCabe and Mrs. Miller" is a good example of the use of Double Fogs (plus flashing.)

As I've said, none of these contrast filters do much in terms of improving dynamic range, but they help a little. I'm just not sure they are worth the extra glass on the lens unless you are going for a visible effect.

 

[m.chang]

I'm trying to decide between a Schneider Black Frost 2 and a Classic Soft 2.
Does anyone have any examples they could share of those two filters? or something comparable?

I want the diffusion to be noticeable, but not to the point where everyone is walking thru fog.

Thanks for the help!

 

[Steve Sherrick]

I recommend Tiffen DFX software if you are trying to get an idea of what these filters do. They are not exact replicas of their glass counterparts, but they can give you an approximation of the effect. I would agree with David about the differences between viewing these on a small screen vs a big screen.

Since I'm mentioning these filters, David have you done any comparisons between the glass versions and the plug-in versions? I am often telling people to filter at the lens for a variety of reasons, but the plugs are interesting. Would love to hear your thoughts on this.

 

[paulherrin]

if you're doing post work, and you can do it in post just as well and dial it more easily... there are some reasons why you might not, but not very many for me.

depends on you, and the situation.

 

[David Mullen ASC]

I actually haven't been in many D.I. suites around town that have ready-made diffusion effects at the touch of a button, most have to spend some time doing blur overlays, etc. It's been tedious enough, and time-consuming enough, in expensive D.I. sessions, for me to believe that if you want a filtered look, then just shoot with a filter on the camera. I actually think that you retain a bit more sharpness with optical diffusion than some post tricks to create an equal amount of halation and whatnot.

I'd save post diffusion for when it makes sense, like the occasional shot where you decide you need it when you didn't while shooting, or some efx composite where everything was shot clean but now it needs some diffusion to match the rest of the filtered footage.

My one caveat with lens diffusion is to always underdo it when in doubt as to the strength because you can always add more in post if needed, but you can't get rid of it easily. After I evaluate diffusion filter tests, I generally pick the filter that is one strength lighter than the one I liked the most.

Now if you are just planning on using diffusion now and then for cosmetic reasons, like when an actor is having a bad day, then it's probably just as easy, and more controllable, to do the softening in post unless you are worried about an important actor seeing themselves in dailies sans the diffusion. But if every shot in a movie is supposed to have a mild filtered effect, let's say a 1/8 White Frost on everything, then it's simpler to do it in-camera rather than process every shot in the movie with a diffusion effect in post.

The thing is that a D.I. session is expensive, and even if adding a diffusion effect to each shot is relatively easy, you know you are going to spend a few extra minutes looking at the effect after the colorist has applied it and then monkeying around a bit to perfect it before you move on to the next shot -- and some movies have well over 1000 cuts in them, so a few minutes extra on every shot adds up if every shot needs that diffusion effect. It can be just enough of a time suck to cause you to change your mind and live without the effect.

But people should do whatever they feel more comfortable doing, if they aren't savvy and experienced with camera filtering, then they may be better off saving it for post.

 

[paulherrin]

lol, that's what custom presets are for... unless you're in a session and want to make some cash.

very good points though, david. there are too many factors for one answer, like so many things.

i'm generally for anything that is purposeful.

 

[Steve Sherrick]

And in the case of the Tiffen DFX plugins, the rendering time can be quite lengthy, especially at 16bit. But as David mentioned, in times when the effect was not done in camera but becomes necessary in post, makes for a great option. Although I don't think some of them nail the effects exactly as their glass counterparts, they can come close and with time and tweaking, probably very close. So a nice option indeed.

One of the reasons I have advocated to filter at the lens, even in the world of RED and RAW, is that it gives the DP some control over image. In other words, it's baked in because it's supposed to be, much harder for it to be messed with later on. As David mentions, experience is a big factor here. I would say most pro DPs feel pretty comfortable with basic filtering decisions. It's when you get into extreme filtering that the comfort level changes.

 

[paulherrin]

exactly.