Sony cameras and the "video look".

paulcurtis said:
I remember a conversation from, i think, cml, many moons ago about something like this and that was actually scientifically put down to the way the compression and the panning interacted to create strong contrasting strobe areas. Of course a bit of contrast adjust in post can fix that.
Click to expand...

Yes, compression techniques can change motion appearance and how motion feels, especially with insufficient bitrate. Mpeg based compression (like AVCHD in fs100) uses motion vectors and prediction to encode blocks movement around. And smaller blocks allow for more precise motion coding.
Intra frame codecs don't interfere with motion. This is one of the reasons the digital cinema standard is based on intra-frame compression.

On the other hand, I don't think that compression should be looked at as an inherent camera characteristic, especially on cameras that can output uncompressed video.
 
When you look at the demo material of Tessive Filter, the difference is striking (a least to my eyes).
So a simple fade in/fade out is doing the trick of instant real 24p look out of a normal video camera.

I wonder why no camera manufacturer is putting something like that into a camera. It could probably even double as a variable ND filter.
Maybe it´s even doable as a post filter - who knows?

The best results I ever got out of the FS100 motion wise, was recording interlaced and use RedGinat's Frames filter.


Frank
 
Frank, do you know any exmaples of the Tessive fitler compared to a film camera? All the stuff I see on their site is digital vs digital + filter.
It doesn't look very filmic to me. In fact it looks more like high frame rate, probably due to the temporal antialiasing.
 
Wow, that Tessive filter is a real snip at $7900. Ouch!

This filter is not trying to recreate film look but to overcome the limitations to a shuttered system be that mechanical or electronic. Film still suffers from the artefacts the Tessive people are helping remove.

It's really nice, gorgeous actually, but I don't think it explains the motion difference people are saying there is between the FS100 and *insert your preference* high end camera.

To my eyes having started out making films on my Dad's SVHS-C camcorder over a hundred years ago (it feels like it), everything and I mean everything looks fantastic these days.
 
Ok, I read a bit more through Tessive's site and here is what they have to say about film:
​
The difference in response between a film camera and a digital camera is much, much less than the difference between either of them and the Time Filter.
​
Click to expand...


​
More here: http://www.tessive.com/home/time-filter-faq-1
They aren't really promoting it as making digital look filmic, because it doesn't do that.
 
paulcurtis said:
I'm agreeing with cpc and Nirv here but i mention it because there's been a couple of cases where people have chosen to not use the FS (kholl for example) specifically because of the motion rendering. Yet nothing to back that up. With my conspiracy theory hat on perhaps kholl is a canon shill, but realistically i assume not - just that he has a very strong ideal especially with motion. I just would like to understand in specific cases and examples what is so offensive about it because technically i can't see that much that is innately wrong with the way it is being capture. As nirv points out - what's the playback on.

I remember a conversation from, i think, cml, many moons ago about something like this and that was actually scientifically put down to the way the compression and the panning interacted to create strong contrasting strobe areas. Of course a bit of contrast adjust in post can fix that.

Different people have different sensitivities. I can't look at most DLP systems because i can pick out the rainbow effect, unless it's spinning really fast or has lots of colours.



Best point in thread.

cheers
paul
Click to expand...

I tried to stay out of this conversation but had been reading it to see what other people think, knowing full-well that my name would come up sooner or later.

One thing I am far from is a shill for any company or any product: I have VERY little brand loyalty when it comes to camera manufacturers. You can find posts from me on every camera sub Alexa stating things that are not desirable to me. Canon is far from the company that I would even begin to Shill for, as I don't think many of their products are worth the money they charge for them. Yet, if I had to choose between an FS700 and a 5D3, there are various reasons why it would not be an obvious choice for either camera.

It's been said a few times here that it's basically about the cinematography. Okay, well, people are entitled to think what they want, but great lighting and composition doesn't change the way a sensor renders motion in any given frame. Neither does sharpness, color, so on and so forth. The first issue is that the "video look" and "cheap movie look" are two different things. One is specifically describing footage with odd motion characteristics (30P/60i) and the other is lack of proper care toward a goal that may be achieving cinematic looking images.

You can make almost ANY camera's footage look cinematic if you fight it hard enough, light it well enough, compose it like you know what you're doing. However, there are ONLY so many things you can do to make the subtle differences in the way a digital image represents twenty-four frames in motion look like what we've been used to seeing in various forms or fashions over the last number of decades.

Not all 24P is created equal, this isn't new information.

The FS100/700 are PLENTY capable cameras, but what both represent to me are still a bias toward a more "video" cadance/look and not a "film-like" cadance/look. I've been told that it's just the way Sony engineers prefer things, and it may be because of the eastern content industry. Whatever the reason, every single Sony camera I have ever used (with the exception of the F65 right now, because I haven't really had a chance to shoot anything serious with it) has this look. It's not just Sony, either. I mentioned that RED goes in and out, and I have never been able to pinpoint why. Canon cameras do a mild bit better, Alexa and Gh2 being the closest of them all to my eyes.

Nirv's point about playback is taken, I'm keen to go to every source that I can find, and have watched FS100 footage projected (I said christie before, but it's actually a Barco 2K projector), 60+ Plasma, my own 56" LCD, post color monitors, and even an iPad. I see it everywhere I go, and the likelyhood is just that I am sensitive to it. Sometimes a small shutter tweak can make it better (motion blur change) which is cool. For this reason I won't shoot above 1/45 on Canon DSLRs, or 1/40 on the GH2 (although 1/50th doesn't bother me).

No matter what I did I could not defeat it on the FS100.

Me, personally, I sincerely don't care what the audience notices or does not notice. That's not why I choose a camera. I choose a tool to be comfortable and confident in, one that matches where I consider myself to be at the current time (whether or not someone else agrees with me is neither here nor there) and fits the work.

If it weren't important then we would all still be shooting out 35mm Adapters... since the audience won't care or notice if the movie is great.

I'm STILL going to use the FS100. After shooting with the FS700 I revisited the FS100 and really liked what I saw, but it's for a very specific use to me, not one that I would choose for everyday narrative or commercial work. That shouldn't really matter to others, if you like what you like then use it, show people what you're doing with it.

By the way, Paul, I honestly couldn't show you an example that would be fitting because you'd have to literally shoot the exact same image (frame, speed, everything) to see it. I wouldn't know how to organize something so intricate, and probably wouldn't have the patience for it. That's why I end up at "if you can't see it then it doesn't matter." -- not to be offensive, but because it really doesn't matter if you can't see it.
 
I think the most sensible comment related to the carpenter and the hammer. In my hands, no matter what format or make I use, everything I produce looks like documentary. Other people using the same kit produce beautiful pictures. The same thing happens with stills. I take photographs, a good friend of mine is a photographer. He sees things I don't, that impact on his framing and choice of shutter speeds and aperture.

It's art - and all this rubbish about 'cinematic', 'aesthetics' and 'film' just set aside the combination of technique and art. We can all tweak our electronic parameters as much as we like, but unless you have the eye, it won't have the look people are talking about (yet none can actually describe it).

We've all seen excellent material on youtube, shot on God knows what that just looks stunning, and equally we've all seen material produced by amazingly expensive equipment that we've smiled at.

It's not as simple as technical specs and buttons. If you watch old movies like gone with the wind, stuck onto DVD from your old VHS recordings, they still have that special quality that lets you know it's a proper movie.
 
Kholi said:
By the way, Paul, I honestly couldn't show you an example that would be fitting because you'd have to literally shoot the exact same image (frame, speed, everything) to see it. I wouldn't know how to organize something so intricate, and probably wouldn't have the patience for it. That's why I end up at "if you can't see it then it doesn't matter." -- not to be offensive, but because it really doesn't matter if you can't see it.
Click to expand...

Firstly i just want to make sure that you know the canon shill thing was tongue firmly in cheek. I know you're not and you have an objective view of your tools.

Secondly we talk about motion cadence. And that's fine. But if we're comparing 24fps from camera to camera and viewing it at 24/48 fps with the same exposure/shutter/settings then i'm failing to understand *where* the difference is creeping in. AFAIK all sensors principally work in the same way. So when sony spoke about a certain motion look - where is that originating? The japanese 'look' to the best of my knowledge is really interlaced, sharpness and colour settings.

Aside from the numerous ways of mangling footage downstream, i can't see there being a motion difference between an alexa and AF100 if the settings are the same. Unless the physical sensor technology allows for a different light response in the silicon itself.

just interested...
paul
 
I'm not technical enough to know where the difference is coming from. I think it was you (or maybe someone else) that mentioned it could be due to Long-GOP or GOP settings. As in, the GH2 Hacks have given you the options for GOPS (1,2,3) and you can definitely see the difference in motion between them. I thought the original non-hacked GH2's motion still looked great though, Same with the GH1.

Similarly, 24F from the old XH series and 24P from the HVX200 looks completely different to my eyes, and matching up to that era in camera tech, the EX1 looked drastically different than both of those.

I've got an old old old example of 35mm Adapter footage shooting a test "horror" scene, HVX200A and EX1, and the difference is pretty pronounced although there are several variables to consider, not limited to how a rotating ground glass effected the image; those were the days.

Anyone more technical could probably shed some light on it, but I do know that I could see the difference in 24P between
 
Kholi said:
I'm not technical enough to know where the difference is coming from. I think it was you (or maybe someone else) that mentioned it could be due to Long-GOP or GOP settings. As in, the GH2 Hacks have given you the options for GOPS (1,2,3) and you can definitely see the difference in motion between them. I thought the original non-hacked GH2's motion still looked great though, Same with the GH1.

Similarly, 24F from the old XH series and 24P from the HVX200 looks completely different to my eyes, and matching up to that era in camera tech, the EX1 looked drastically different than both of those.

I've got an old old old example of 35mm Adapter footage shooting a test "horror" scene, HVX200A and EX1, and the difference is pretty pronounced although there are several variables to consider, not limited to how a rotating ground glass effected the image; those were the days.

Anyone more technical could probably shed some light on it, but I do know that I could see the difference in 24P between
Click to expand...

If that is the case, and it is a GOP thing,
then wouldn't recording to an external
recorder in nonGOP ProRes 'fix' the
problem?
 
alaskacameradude said:
If that is the case, and it is a GOP thing,
then wouldn't recording to an external
recorder in nonGOP ProRes 'fix' the
problem?
Click to expand...

Possibly. I couldn't tell a difference between the NanoFlash XDcam and INternal recording, though. Not as far as motion is concerned.

And, Frank, on the HVX200 motion... I think that's part of the mojo that everyone craved. I ACTUALLY did like 24F from canon, but tape-based?... well, okay. haha.
 
Postmaster said:
The best results I ever got out of the FS100 motion wise, was recording interlaced and use RedGinat's Frames filter.


Frank
Click to expand...
It's a shame. so many things with this camera.
Why Sony did not just took there time and got it right.
reddish skin ton, ghosting and not as convincing 24p.
Am I over reacting?
 
Nostalgic said:
It's a shame. so many things with this camera.
Why Sony did not just took there time and got it right.
reddish skin ton, ghosting and not as convincing 24p.
Am I over reacting?
Click to expand...

Yes. You are over reacting. The FS100 is a ground breaking camera, but it is not perfect. There is not a camera out there that is perfect.
 
Nostalgic said:
It's a shame. so many things with this camera.
Why Sony did not just took there time and got it right.
reddish skin ton, ghosting and not as convincing 24p.
Am I over reacting?
Click to expand...

Over reaction is an emotionally driven response, you're allowed to feel how you please, however I'd say that it's pretty much a non-issue if you can't tell. That's just me being honest.

It's a pretty sound image and I've grown to like the camera, I wouldn't tell anyone not to use it if it suits their needs.
 
Do you guys feel that this is the best Sony could do at that point for the price or those issues are a "defects" and will be corrected in the next model?
 
Wendt said:
Is there any post process that can help emulate the mechanical shutter or motion/ blur of film? I know that it is possible to write scripts within Sony Vegas, has anyone written one that helps with film motion?
Click to expand...

I would like to ask this question again. I've seen people double tracks add blur and drop the transparency of the top track to give more of a dreamy film look. What other post process may help? What about writing a script that would re-cut the footage into 24 individual frames/sec and then apply some cross-fade between the frames. This may be a real dumb idea- I know, but hopefully it will inspire the people who have better ideas to experiment and share.
 
It´s called frame blending, most NLEs can do that, but doesn´t cut it.
Best Software for the moment - that I´m aware of - is MagicBullet.
 
You must log in or register to reply here.
Back
Top