Pro mist vs post tools

[Arthur Woo]

Who here prefers to bake in a look by using filtration, vs using post tools to achieve a similar effect? Will be shooting a fashion spot and unsure if I want to try to use a pro mist filter or some of the post tools I use for skin softening. I have not used filters for skin before and have one day to rent the item and no time to test, or use nothing in front of the lens and 'fix' everything in post. Would like to hear pros/cons of filtration, as well as strength. Was looking at using 1/8 for a subtle effect - any reason to jump to 1/4? Don't want the image to look too hazy. There will be closeups of the face and skin on other parts of the body, but no macro shots. Shooting Dragon at 6k so slightly concerned about too much detail on the skin.

 

[Mike Garrick]

Interesting . My believe as a director is why lock in "a look " thru the gate when we can decide in post .

I welcome someone shooting down this theory .

 

[Phil Arntz]

I've seen this question being asked a lot. My take is that I will continue using filters, even with all the room you get with R3D's in post. I find it very hard to replicate a pro-mist or a Hollywood Black Magic or Glimmer Glass for example in post.

Another reason I use filters is because it gives me a clear direction on set. I know what I want from an image and what benefits the story I am telling, so I lock that in and don't have to fiddle around in post. If all you do is post, it feels like I didn't know what I wanted when I shot it.

 

[Toby Linden]

I've seen this question being asked a lot. My take is that I will continue using filters, even with all the room you get with R3D's in post. I find it very hard to replicate a pro-mist or a Hollywood Black Magic or Glimmer Glass for example in post.

Another reason I use filters is because it gives me a clear direction on set. I know what I want from an image and what benefits the story I am telling, so I lock that in and don't have to fiddle around in post. If all you do is post, it feels like I didn't know what I wanted when I shot it.

+1

There are some valid arguments out there that it "degrades" the image or locks in the look, but if you know exactly what you want, why bother with achieving the desired effect in post production? I'm a big fan of capturing it in camera and using the DI process to enhance my vision instead of finding one. By using certain filters or in camera effects I force myself to make very deliberate choices that need to be thought through.

 

[Matthew Young]

I shot a film this February on a PMW-F3 with a 16mm Zeiss Super Speed Mk. III. I found it a little sharp in the background, (considering the film was set in the 18th C) so I used a Tiffen 1/4 fog for the whole shoot and don't regret it one bit. It gave the bokeh a lovely softening effect. It's great what can be done in post now, but based on this experience, I'll always shoot with filters if there is a specific effect I'm going for. The same goes for older lenses as opposed to adding the effect in post.

 

[Ignacio Aguilar]

A 1/8 Black Pro-Mist filter gives the image a nice but still very subtle softness that's it's worth trying if you want to get rid of today's lenses sharpness when combined with digital cameras.

The effect that you get when you do it in camera (as opposed to apply the filter in post to a clean image) always seems more filmic to me, even though you can enhance an already mild filtered image (i.e. 1/8, 1/4 Pro-Mist, Low-Con, Fog, etc) with good results in post.

My advice, if you're not familiar with diffusion, is to choose a very light filter strengh and keep it consistent through the shoot. Most problems with diffusion filters can appear when someone starts using them and mixing strenghts for wide-shots, close-ups, etc. With 1/8 or 1/4 Black Pro-Mists, for instance, the effect is so subtle you don't have to worry about changing filters for each shot size.

Also, bear in mind that the amount of visible diffusion is also affected by the lenses that you're using. Cooke Speed Panchros, Baltars, K35, etc. already look a bit diffused and low-con due to their coatings, so if you want to keep a subtle diffusion effect, you must use lighter strenghts. Ultra Primes, S4s, Master Primes, etc. will accept heavier strenghts of these filters.

 

[Arthur Woo]

I've seen this question being asked a lot. My take is that I will continue using filters, even with all the room you get with R3D's in post. I find it very hard to replicate a pro-mist or a Hollywood Black Magic or Glimmer Glass for example in post.

Another reason I use filters is because it gives me a clear direction on set. I know what I want from an image and what benefits the story I am telling, so I lock that in and don't have to fiddle around in post. If all you do is post, it feels like I didn't know what I wanted when I shot it.

Thanks for the feedback. I'm not familiar enough with filters (aside from ND) to determine whether or not it's something I'd be able to replicate in post. I saw a spot shot by someone recently using a Hollywood Black Magic and it made me think to try it. I usually have a post flow where the skin/sharpness is dialed in after the fact, just wanted to know if there was an added benefit. From what you're saying it sounds like there is something in a filter than cannot exactly be replicated in post.

 

[Arthur Woo]

+1

There are some valid arguments out there that it "degrades" the image or locks in the look, but if you know exactly what you want, why bother with achieving the desired effect in post production? I'm a big fan of capturing it in camera and using the DI process to enhance my vision instead of finding one. By using certain filters or in camera effects I force myself to make very deliberate choices that need to be thought through.

Good points. For me it has to do with flexibility if ever needed. I haven't had a client yet complain about the method I use to clean the skin, but in the event they ever do, I can just dial it back down in post if they want it to look sharper. I would like to get more familiar with filtration though as for some shoots there is less time for post than others and seeing how I can cut down on that post time.

 

[Arthur Woo]

That's a great look and argument for when to use it. It does feel a little less 'modern' if it's not extremely sharp. This look reminds me of Lomo primes.

I shot a film this February on a PMW-F3 with a 16mm Zeiss Super Speed Mk. III. I found it a little sharp in the background, (considering the film was set in the 18th C) so I used a Tiffen 1/4 fog for the whole shoot and don't regret it one bit. It gave the bokeh a lovely softening effect. It's great what can be done in post now, but based on this experience, I'll always shoot with filters if there is a specific effect I'm going for. The same goes for older lenses as opposed to adding the effect in post.

 

[Arthur Woo]

A 1/8 Black Pro-Mist filter gives the image a nice but still very subtle softness that's it's worth trying if you want to get rid of today's lenses sharpness when combined with digital cameras.

The effect that you get when you do it in camera (as opposed to apply the filter in post to a clean image) always seems more filmic to me, even though you can enhance an already mild filtered image (i.e. 1/8, 1/4 Pro-Mist, Low-Con, Fog, etc) with good results in post.

My advice, if you're not familiar with diffusion, is to choose a very light filter strengh and keep it consistent through the shoot. Most problems with diffusion filters can appear when someone starts using them and mixing strenghts for wide-shots, close-ups, etc. With 1/8 or 1/4 Black Pro-Mists, for instance, the effect is so subtle you don't have to worry about changing filters for each shot size.

Also, bear in mind that the amount of visible diffusion is also affected by the lenses that you're using. Cooke Speed Panchros, Baltars, K35, etc. already look a bit diffused and low-con due to their coatings, so if you want to keep a subtle diffusion effect, you must use lighter strenghts. Ultra Primes, S4s, Master Primes, etc. will accept heavier strenghts of these filters.

Thanks! Also great points regarding the lenses.

 

[David Mullen ASC]

Why do it in-camera instead of in post?

1) Digital diffusion and optical diffusion are not exactly the same effects.
2) Doing it in post adds time to post, especially if there are a thousand shots like in a feature. Generally budgeted time for a D.I. session is quite tight these days.
3) If you do it in post, it probably won't be there in dailies. If it's not there in dailies, everyone editing the project will get used to it not being there.
4) If you do it in post, it's not there on the monitor on the set, and if you are trying to convince an actress looking at the monitor that you are trying to make her look her best, telling her to wait until post is not always effective.
5) Being an artist means making choices and living by those choices, which is how you learn. If all of your choices are reversible in post, you never really learn whether something could have been done better next time, you just keep shooting things in a manner that is reversible in post. Why not just record takes of the actor making every possible acting choice and then creating the performance piecemeal in post? Why not do versions of the scene in different wardrobe and different lighting for different times of day and decide what you really want in post?
6) There's no guarantee that you'll be there in post to create the look you intended.

I've never felt that being able to keep changing your mind as to what you want was the best argument for doing things in post. Yes, there are times when it is more practical to save a decision for post, but there are also times when it is more practical to make a decision in prep too, or on the day of the shoot. As a filmmaker, you have to get to know your tools, and sometimes the best tool is post but sometimes it's not, it just becomes a hammer and everything looks like a nail to it. But how are you going to get to know your tools if you don't use them?

From a technical standpoint, it's hard to match the look of a 1/8 ProMist in post because when done optically, it has a very small effect on sharpness compared to the halation it creates around a bright highlight. When you do post tricks like a blurred highlights layered over a sharp background layer, you lose sharpness in the highlights in order to get the glow, and the more glow you try to create, the more you lose sharpness in the highlight. Now this actually can be nice for close-ups because of the loss of sharpness in the highlights of the skin, but it's not so desirable for wide shots where all you want is that touch of halation.

So basically what it comes down to do is that both approaches have their advantages and disadvantages.

For a native 3D movies or shots with a lot of visual compositing work, it makes more sense to do the diffusion in post.

But I certainly don't buy the argument that doing it in post is always the best method.

 

[Arthur Woo]

Wow, thanks for taking the time to write this out David. I haven't worked on sets for features and don't really plan to, but all of these points are applicable to what I shoot.

Why do it in-camera instead of in post?

1) Digital diffusion and optical diffusion are not exactly the same effects.
2) Doing it in post adds time to post, especially if there are a thousand shots like in a feature. Generally budgeted time for a D.I. session is quite tight these days.
3) If you do it in post, it probably won't be there in dailies. If it's not there in dailies, everyone editing the project will get used to it not being there.
4) If you do it in post, it's not there on the monitor on the set, and if you are trying to convince an actress looking at the monitor that you are trying to make her look her best, telling her to wait until post is not always effective.
5) Being an artist means making choices and living by those choices, which is how you learn. If all of your choices are reversible in post, you never really learn whether something could have been done better next time, you just keep shooting things in a manner that is reversible in post. Why not just record takes of the actor making every possible acting choice and then creating the performance piecemeal in post? Why not do versions of the scene in different wardrobe and different lighting for different times of day and decide what you really want in post?
6) There's no guarantee that you'll be there in post to create the look you intended.

I've never felt that being able to keep changing your mind as to what you want was the best argument for doing things in post. Yes, there are times when it is more practical to save a decision for post, but there are also times when it is more practical to make a decision in prep too, or on the day of the shoot. As a filmmaker, you have to get to know your tools, and sometimes the best tool is post but sometimes it's not, it just becomes a hammer and everything looks like a nail to it. But how are you going to get to know your tools if you don't use them?

From a technical standpoint, it's hard to match the look of a 1/8 ProMist in post because when done optically, it has a very small effect on sharpness compared to the halation it creates around a bright highlight. When you do post tricks like a blurred highlights layered over a sharp background layer, you lose sharpness in the highlights in order to get the glow, and the more glow you try to create, the more you lose sharpness in the highlight. Now this actually can be nice for close-ups because of the loss of sharpness in the highlights of the skin, but it's not so desirable for wide shots where all you want is that touch of halation.

So basically what it comes down to do is that both approaches have their advantages and disadvantages.

For a native 3D movies or shots with a lot of visual compositing work, it makes more sense to do the diffusion in post.

But I certainly don't buy the argument that doing it in post is always the best method.

 

[David Mullen ASC]

Another point is that today with high-def monitors on sets, if you carry some selection of filters of different strengths, you generally aren't going to run into surprises in post if you are careful in your use of filters on camera.

I also recommend considering a hybrid approach, do a lighter pass of diffusion in camera and then make it heavier in post if you need to. This is assuming you have decided on a diffused look for the project.

 

[Gavin Greenwalt]

#1 reason to filter on set: HDR. The optical filtration will take into account light sources that are many times above white and clipping. I could give you a nearly perfect hallation or softening effect. Like 99.999999% perfect. It would be very slow and take about 1 second per frame to render. But if you point the camera at the sun the filter is going to properly filter the entire intensity of the sun. A post effect would only filter up to white. One of the best arguments for using these filters in the first place is to soften clipped lights. An open face light will give a stronger halation than a white shirt that's just barely clipping.

Now, in the not at all distant future I think this will be a moot point. Sensors will give enough dynamic range to properly register even an open face bulb without clipping and the GPUs in a camera will give you the option to turn on and off filtration with a number of softening convolution kernels to perfectly emulate any filter you could dream of having as metadata. Filtration will become like ISO you could pick a preset or you could change a slider to dial in the intensity and size of the filtration you want per-shot.

 

[Phil Arntz]

AMEN! Thanks David.

Why do it in-camera instead of in post?

5) Being an artist means making choices and living by those choices, which is how you learn. If all of your choices are reversible in post, you never really learn whether something could have been done better next time, you just keep shooting things in a manner that is reversible in post. Why not just record takes of the actor making every possible acting choice and then creating the performance piecemeal in post? Why not do versions of the scene in different wardrobe and different lighting for different times of day and decide what you really want in post?

 

[Marc Wielage]

Another reason I use filters is because it gives me a clear direction on set. I know what I want from an image and what benefits the story I am telling, so I lock that in and don't have to fiddle around in post. If all you do is post, it feels like I didn't know what I wanted when I shot it.

I think of two good reasons to only use filters in post:

1) you can choose to specifically apply the filter effect only to certain images in the frame, like faces only and not the background (or vice versa)

2) you can choose to dial the effect back out if you realize in post that you went too far.

People see different things on a large, calibrated monitor far away from the stress and poor viewing conditions on a set.

On the other hand, I concede that there are some major-name DPs who choose to use filtration on set because they don't want the image changed after the fact, especially by other people who try to overrule his or her decisions. I would also agree that a digital filter effect is not quite the same thing as an organic piece of glass.

I've used the Tiffen DFX plug-ins in Resolve, and while they tout it as being very close to Tiffen glass filters (including Gilmmerglass), it's not exactly the same thing. It's in the ballpark, and again, we have the advantage of being able to dial the effect up or down shot to shot -- one setting for close-ups, one setting for medium shots, and one setting for wide shots -- which is much harder to do on set, especially when you're trying to make judgements from a 17" or 20" on-set monitor.

I also recommend considering a hybrid approach, do a lighter pass of diffusion in camera and then make it heavier in post if you need to. This is assuming you have decided on a diffused look for the project.

I think this is a very wise approach, provided the user has the ability to do tests prior to production to know the extent of how far they can go.