What’s wrong with electronic variable NDs

B

bob'slaststand

Member
I was watching a presentation on the Sony Venice camera and the 8 stops of servo controlled ND inside the body and was wondering why they did not put the electron ND system in there from their smaller prosumer cameras like the F5 and F7.

An electric ND that can be adjusted in a stepless manner seems like an incredible idea and other manufacturers like Panavision have also been rolling out mounts and adapters with this new technology.

I’ve never got my hands on one so can someone explain why this technology might not be considered suitable for a high-end camera like the Venice? Does it have problems with color cast or polarising the image like a variable?
 
After reading a post in the Sony forum regarding, I believe an fs5(?), there may be an issue with possible IR pollution, since as the "ND" level increases(blocking more visible light) nothing is blocking the corresponding/relatively increasing IR. With real, fixed level ND filters, they can build in corresponding IR filtration in each individual level of ND. With the variable system, are they just making the built-in IR filters set to the minimum ND level, a mid-level, or the max ND level? Are there negative consequences to blocking IR that corresponds to 8 stops of ND and you're only shooting with 2 stops of ND?

That being said, I do really like the idea of a continuously variable system system(ideally 0-10 stops).
 
After some research, Sony claims in their own information online that the Electronic ND does not suffer from any color shift and it does not polarise the image.

It uses an LCD layer as do the other new electronic NDs entering the market at the moment like the new Panavision design.

I did find discussions on how many had struggled with color shift during the design but of course claim to have solved these problems on release.

If Panavision put their name on a finished product and put it put on the market then I’ll trust that they will have nailed it but theirs is still in testing.

Sony’s reluctance to put it in the Venice is saying something but to be fair the cine rental market is notoriously suspicious and slow on the uptake of new technology and it may simply be a marketing decision. Not wanting to put anything that could be seen as gimmicky or of an unknown quantity in there.

Whatever the case for me personally the idea of a clean, un-polarised, seamless transition variable ND is a very big deal and I’m surprised how little interest there has been in this new technology. I found no critical testing of any of the new LCD NDs anywhere, Sony or otherwise. If anyone knows of something I’ve missed or has there own experience I’d be interested to hear about it.
 
bob'slaststand said:
I’m surprised how little interest there has been in this new technology. I found no critical testing of any of the new LCD NDs anywhere, Sony or otherwise. If anyone knows of something I’ve missed or has there own experience I’d be interested to hear about it.
Click to expand...

Funny, I was just looking at this video today. Looks pretty good to me--
 
bob'slaststand said:
After some research, Sony claims in their own information online that the Electronic ND does not suffer from any color shift and it does not polarise the image.
Click to expand...

How does one use liquid crystal to control light without polarization? Do you have the link?

Maybe that's why it's not in Sony's higher end cameras (i.e. it somehow degrades the image. An LCD is lot cheaper than the complex mechanical glass filter system in the Venice).
 
Good point jcs. I didn't know how LCD’s worked. I look it up after seeing your post and after reading up on it a bit I can see what you mean, it seems that an LCD will always polarise. You never know, maybe they found a way around it but from my limited understanding, it looks to be fundamentally tied to how an LCD functions.

This is a good explanation why it is not migrating to the higher end cameras and why there has been so little fuss about it in general.

However, I cant believe that Panavision is developing “cinema’s first” LCND to sit in front of their new Millennium DXL2 camera and Primo lenses if it polarises the image. It also seems strange that they are R and D’ing a product, expected release date 2019, that Sony and many other smaller companies have already brought to the market. It’s not even stepless. During the interview, they talk about removing color cast but never mention polarisation however the last comment thread on the youtube video asked:

“I wonder what's the advantage over polarizer-based variable ND's"
“you could still use a polarizer in addition to it I guess”

and cinema5D answers:
“Yes this should enable you to use polarizers on top and you don't have the polarizing effect like with circular manual screw-on variable ND filters. Nino"

Dose the nu-polarised color neutral LCND exist?

Dose anyone have one of these Sony cameras and know if it’s LCND polarises the image?
 
I'm not sure how useful this is because I have no notes about how much ND I was using because I was rolling it on the FS5 so it's continuously variable but it was my original post that Run & Gun is referring to. It was quite heavy ND though.
This is the video where it happened:
https://vimeo.com/287688718 Its just the first 1:10 secs. In the end I didn't really attempt to get rid of the IR pollution because I don't have the skill to be honest. The client was very happy but I do intend to avoid the problem in the future but I haven't had time to do any testing yet.
 
This is pretty cool (adding an additional polarizer makes the combined filter brighter):

A depolarizer solution seems unlikely: https://en.wikipedia.org/wiki/Depolarizer_(optics)

A typical circular polarizer (linear polarizer (1/2 wave plate) + 1/4 wave plate (https://en.wikipedia.org/wiki/Waveplate)) won't block more light when rotated, but it will change the color (e.g. B+W KSM C-Pol MRC).

In sunlight I've had decent luck with variable NDs (polarizer based) and skin tones (perhaps models wearing makeup aren't as affected as natural skin re: light sub-scattering in the skin surface: https://www.univaasa.fi/materiaali/pdf/isbn_978-952-476-348-6.pdf ).

To check the Sony: place a linear polarizer (or polarized sunglasses, 3D shutter glasses, etc.) in front of the camera and rotate it. If the recorded (or live-viewed) image maintains constant brightness during rotation, it's not polarizing the light.

It's possible that there's new LCD tech that can neutrally block light without relying on polarization, e.g. https://www.youtube.com/watch?v=d_S4P_y018k
 
The Sony Venice is aimed at productions where they will have on-hand a full set of ND filters, an AC to perform necessary filter management, and- most of all- the time to perform the filter addition / subtraction. The amount of time needed to swap out ND filters isn't significant in the pace of feature filmmaking ' "high-end" productions. There is typically plenty of time for swapping lenses and management of filters.

That said, electronic variable ND has benefits that "high-end" productions might find beneficial, but I suspect that the timing just wasn't right for Sony to incorporate it into the Venice. Maybe the Venice was too far along already in development when electronic ND was sorted out fully and / or Sony didn't want to potentially put any DP's off to the Venice by incorporating something so young, so unfamiliar to DP's, and possibly seen as "gimmicky" or still associated with "low-end" shooting. The DP's the Venice aimed at weren't likely using variable ND and likely also considered variable ND a crudity that no self-respecting DP would ever use.

I've asked some well-known features DP's about electronic ND and they'd never heard of it. I think I've brought it up with five different DP's and not a one of them had heard of it. That wasn't surprising to me. Last week I mentioned a speedbooster to an experienced First AC on a big-budget film and he had no idea what a speedbooster is. And he wasn't much interested in what it is. You might be surprised at the extent to which members of Camera Departments only know the gear that they use. Variable ND is something that many of the DP's the Venice was targeting have no experience with and if they are familiar with it have a negative association with.
 
Thanks Patrick. I’m a little surprised the FS5 using it’s own in built electronic ND left the image with IR pollution. I presumed that they would have compensated for it. As I understood you do have a button to engage or clear the Sony's internal NDs. I thought that would include a IR filer of some description.

If LCND systems all produce IR pollution maybe the advantage of internal mechanical NDs is simply that they can include bespoke IR filtration to each ND.

Jcs, this is getting to the point where I don’t feel comfortable trying to predict whether its likely or not, I will certainly have a very different impression of Panavision’s R and D department if they do manage it.

Yes that’s pretty crazy, I didn't expect to break reality when I first posted this question.

I’m not sure about the study linked, maybe I missed the reference to polarised light in the title or preface, you’ll have clarify what your trying to say with that.

Personally I think the makeup departments over do it for my taste and have had my struggles trying to get them to do less to the talent. That said I’ve also found that the effect of a polariser on skin can be subtle though of course best avoided. My greater concern with polarised ND systems is the lack of control of polarisation in bright environments.

I should point out that I am not a professional camera man and am here looking for better informed opinions than mine.
 
Yes, the traditional the camera department. I agree with you there, in-fact I think we used almost the same wording though it’s more convincing to hear it from someone who lives in that world.

But as I posted before Panavision is building one for their Millennium DXL2 ecosystem and Sony has gone to all the trouble of building internal mechanical NDs into the Venice, why get beaten to the post by Panavision if they already have the technology in-house, worried about brand image? Maybe.

I suppose the question I’m really interested in is dose this technology exist? Or is there some fundamental failing to these LCNDs that prevent them from ever replacing a set of ND filters?
 
JPNola said:
The Sony Venice is aimed at productions where they will have on-hand a full set of ND filters, an AC to perform necessary filter management, and- most of all- the time to perform the filter addition / subtraction. The amount of time needed to swap out ND filters isn't significant in the pace of feature filmmaking ' "high-end" productions. There is typically plenty of time for swapping lenses and management of filters.

That said, electronic variable ND has benefits that "high-end" productions might find beneficial, but I suspect that the timing just wasn't right for Sony to incorporate it into the Venice. Maybe the Venice was too far along already in development when electronic ND was sorted out fully and / or Sony didn't want to potentially put any DP's off to the Venice by incorporating something so young, so unfamiliar to DP's, and possibly seen as "gimmicky" or still associated with "low-end" shooting. The DP's the Venice aimed at weren't likely using variable ND and likely also considered variable ND a crudity that no self-respecting DP would ever use.

I've asked some well-known features DP's about electronic ND and they'd never heard of it. I think I've brought it up with five different DP's and not a one of them had heard of it. That wasn't surprising to me. Last week I mentioned a speedbooster to an experienced First AC on a big-budget film and he had no idea what a speedbooster is. And he wasn't much interested in what it is. You might be surprised at the extent to which members of Camera Departments only know the gear that they use. Variable ND is something that many of the DP's the Venice was targeting have no experience with and if they are familiar with it have a negative association with.
Click to expand...

This is so true. I brought my DJI Osmo onto the set of Prison Break in Morocco a few years ago and caused quite a stir with the DP, Ops and ACs, they all freaked out over it and wanted to play with it. The Osmo had been out for a while but they had never heard or seen it. Their world was simply Arri and Red, they had no knowledge of much else as far as camera technology.
 
bob'slaststand said:
I suppose the question I’m really interested in is dose this technology exist? Or is there some fundamental failing to these LCNDs that prevent them from ever replacing a set of ND filters?
Click to expand...

Yes, the technology exists. As far as LCND fully replacing a set of traditional ND filters, I think that will happen.

As best as I can glean from information available online, there as of yet is no system that has zero impact upon color but the engineers are getting the impact upon color to a point where it is so minimal as not to prevent LCND from being widely adopted, even in more critical production.

This might be informative:

"A variable neutral-density filter using guest-host liquid crystal devices was fabricated for color video image sensing. The liquid crystal devices contain black dichroic dyes whose absorption is controlled by voltage applied to the device. The filter keeps the spectral balance of transmitted light almost constant for three primary color wavelength regions, even when its transmittance is changed. The fabricated filter exhibited a high extinction ratio of 70:1 for an applied voltage of 20 Vrms, and the color temperature shift of the transmitted light was less than 200 K.

Video-Sensing Effects with Liquid Crystal Neutral-Density Filter. | Request PDF. Available from: https://www.researchgate.net/public...ts_with_Liquid_Crystal_Neutral-Density_Filter [accessed Sep 29 2018]."

https://www.researchgate.net/publication/274794443_Video-Sensing_Effects_with_Liquid_Crystal_Neutral-Density_Filter


_________________________

And here is some information as to how it will be employed, from Tony Ngai, a member of the Society of Motion Picture & Television Engineers, currently with Salon Films, Hong Kong:


"the current prototype can do steps and can be step less.
With the final product we are planning to develop it so you can select specific stops and will be able transition smoothly with a user interface on the side of the filter. We are hoping to put in a small OLED screen so the use can see the ND value. We are also thinking of a way to remotely control the ND through a popular remote lens control such as the Preston or Cmotion."


https://groups.io/g/CCML/topic/panavision_liquid_crystal_nd/21562663?p=,,,20,0,0,0::recentpostdate%2Fsticky,,,20,2,0,21562663
 
puredrifting said:
This is so true. I brought my DJI Osmo onto the set of Prison Break in Morocco a few years ago and caused quite a stir with the DP, Ops and ACs, they all freaked out over it and wanted to play with it. The Osmo had been out for a while but they had never heard or seen it. Their world was simply Arri and Red, they had no knowledge of much else as far as camera technology.
Click to expand...

Hah! I had the exact same experience on the set of Underwater. The crew were amazed and amused. The Director of the film even asked if he could try it out and walked with it in hand over to his DP saying "Check this out!". The DP was also intrigued and wanted to play with it. My having the Osmo and the crew's interest in the Osmo was a great "in" with the crew that helped me gain access for BTS. The Director allowed me access I might not have been given otherwise.
 
I've actually heard of some Steadicam guys who have been using Osmo's(and the like) for a few years for rehearsals/planning shots out on-set.
 
I linked the paper on skin sub-surface light scattering as a possible physics explanation to common comments "polarizers: suck the life out of skin / make skin look dead". Early CGI skin rendering certainly looks fake / dead, because they didn't model sub-surface light scattering (and certain aspects of specular reflections).

Recent CGI skin modeling looks much more alive / realistic, here's a recent example which adds skin microstructure to the model which significantly changes specular light behavior to improve realism: http://gl.ict.usc.edu/Research/SkinStretch/

Polarizers certainly affect specular reflections, so perhaps sub-surface scattering and specular behavior can explain why polarizers make real skin look bad, and could also explain why models with full foundation and make up where little or no skin is directly visible can still look OK when using a polarizer (or variable ND based on polarizers).

Glass-based NDs bulit into cameras (such as the C300 Mark II) are not color neutral, causing a white balance shift. My C300 II appears to shift towards green when the NDs are activated. Canon's docs state to re-check WB when using the NDs.
Page 69 of the manual:
Depending on the scene, the color may change when turning the ND filter on/off. Setting a custom white balance (p 74) may be effective in such case.
Click to expand...
 
jcs said:
Glass-based NDs bulit into cameras (such as the C300 Mark II) are not color neutral, causing a white balance shift. My C300 II appears to shift towards green when the NDs are activated. Canon's docs state to re-check WB when using the NDs.
Page 69 of the manual:
Click to expand...

That's a blanket statement and needs to be evaluated on a camera-by-camera basis. I have no issues with the internal ND in my VariCam's or my F55. I'd also wager that the internal ND in an AMIRA and VENICE are all extremely high-quality.
 
Run&Gun said:
That's a blanket statement and needs to be evaluated on a camera-by-camera basis. I have no issues with the internal ND in my VariCam's or my F55. I'd also wager that the internal ND in an AMIRA and VENICE are all extremely high-quality.
Click to expand...

Plural (more than one camera) vs. blanket (all cameras). For example, the Alexa Studio does not have a full spectrum ND and is stated as having color shifts: http://cinematography.net/ir-grabs/ir-grabs.html (F55 and 65 stated as full spectrum; note color shifts on these NDs). Quick search didn't find F55 internal ND examples- have you specifically tested yours (might not be noticeable for everyday shooting)?

I only found this out by doing a deep dive color analysis for skin tones (google: Filmic Skin C300 II, for more info). It's not an issue, just something I observed from real-world testing; perhaps a few hundred Kelvin- easily fixed with WB or in post. It appears lower cost NDs shift light at different frequencies (colors) non-linearly, so it's not possible to fully correct with a simple WB adjustment. Apparently it's a complex problem, otherwise Canon could build in auto-compensation when the internal NDs are engaged (or perhaps Sony did a better job with software correction). Colors are jumping all over the place with these NDs as well: https://nofilmschool.com/2013/08/true-nd-filter-color-shift-test

Given the lack of true color calibration for current cameras systems (all the way to the top), everything is ultimately adjusted in post (getting computer and monitor systems accurately calibrated is another issue).
 
You must log in or register to reply here.
Back
Top