FS7: New Sony F7 XDCAM 4K Super 35mm Camcorder

Spot off.

A normalising lut is exactly what's needed as a start point - the extra data captured in a wdr file is then held in reserve for making ajustments in the grade or adding an "artistic" look of some nature without introducing banding or other artifacts.

The first thing I want to achieve with any camera is make the footage look "real" - ill make it look like fogged out of date Koda later.

If i met someone who had skin the colour of the girls on Dens film Id call an ambulance.

Of course we do want to know what the file holds in reserve too - but I'd like to see the basics covered first - no death masks :)
 
Last edited:
Don't we need both?

IF I only can see a 709 normalized image from the FS7, I think I will be highly disappointed because that's not the look
I want to achieve. Otherwise I'll pick up some 709 ENG cam anyway.
I'm interested in non-graded slog footage under controlled circumstances with dark areas, hilight areas and lots of skintones.
Why?

Because I want to know what the camera is capable of and not if Sony gives us a nice LUT for normalizing. I don't wanna see
how the "doug ****" optics respond in direct sunlight nor if the speedbooster + Canon photo lenses are sharp enough or if
somebody is capable of doing a nice gade...

I want it raw (ungraded, uncorrected) sothat I can look at it myself in my NLE or in grading tool.
This is the way I can judge colour reproduction, DR, rolloff and so on.
For cinematic appearance I'll need motion, lot's of motion. Camera motion, actors motion.

In the end, as I said before, we won't get any big surprise from the FS7 footage as we already know how the F5 or FS700/O7Q
can look like.


Just my 2 c...
 
Don't we need both?.

Of course.

Im just sayinh seen a nice normalised image woub be a nice part of that package.

One thing that scares me is honestly I've not seen too much from any of these cams that I like - or is real (except raw)

People might consider that statement to be a little behind the curve but that is how I feel.

--

BTW the dog shot optics film did show something interesting - odd highlight edges.

--

On the monitor at the test at pinewood the image looked typically sony - the highlights were not great - a little flouro.


S
 
I hear you.

But what's a "real" image?
Honestly, it's nothing I care about that much. Even I do mostly doco and corporate.
First of all, I want a certain "look" of my image not to be close to "reality".

Reality is seen different by different eyes and brains. What we can see has not much to
do with what's there. It's just our interpretation of what our eyes can recognize.
That's exactly how cinematography should work. Interprete what you see in a way that
helps the story to be told. The camera/sensor is your eye, the profiles or post pro is your
brain.
 
What is a real image - indeed a good question.

I guess a light of power x appearing one stop lower on the wfm than a light of power 2x and two stops lower than a light of 4x (aka linear)

Most DSLRs (canon/nikon) still chucked into LR/Photoshop seem 'real' to me, whereas untreated log looks to me 'unreal'

Images treated with 'film looks' also look unreal to me (however appealing)

My workflow in stills+vid over 20 years has always been to normalise an image and then optionally add a 'look'

S
 
This is why demo pieces are so damn hard. You have to know what you are shooting and why you are doing it, and then you have to know what the hell you're looking at afterward.

With product videos the other problem is that they're obviously going to show what the camera can do well and shoot stuff just within its limitations - anything that it doesn't do well or where there is a serious issue with the product...well those sorts of shots are not gonna make it into the piece! But those are actually the things we need to know - limitations and issues...any camera can shoot pretty pictures in the right hands - and especially if you're viewing on the web, you're not gonna learn much from looking at a promo.

Steve
 
Definitely like the idea of an industry-standard series of tests which each camera is put through - kind of what Alan Roberts at BBC does, various charts and measurements which give a definitive result - if camera A is lined up on this chart and shows 1000 lines in RAW and camera B on the same chart shows 800 then you can say that A has more resolution - no opinions needed, you have your answer. Same with rolling shutter - Alan used a rotating fan I seem to remember - again, use same fan, same speed and then measure the skew by the angle of the blades - you'll get an answer from 0 to 50 degrees of skew or whatever. Get the feeling that maybe this is something that we would like but the manufacturers not so much?

Steve
 
I would not expect the 'facs to show flaws.. but IMO Dens vid was horribly flawed in the skin tone department - the whole vid did not look 'real' or 'artistic'

I suggest if sony had a 'stock workflow' one should at least hit 'real'

S
 
With product videos the other problem is that they're obviously going to show what the camera can do well and shoot stuff just within its limitations - anything that it doesn't do well or where there is a serious issue with the product...well those sorts of shots are not gonna make it into the piece! But those are actually the things we need to know - limitations and issues...any camera can shoot pretty pictures in the right hands - and especially if you're viewing on the web, you're not gonna learn much from looking at a promo.

Steve

I directed quite a few demo films for Kodak. Our aim was always to stress test the new film stock so people could see how it held up in different shooting scenarios. Showing under and over exposure and comparing with older stocks via split screens was an important part of the process; it wasn't just about creating a promo film. Of course, if we screwed up, the shot wouldn't make the edit.

Den's film doesn't bother me, it's the other people who have been let loose with the camera (apart from Wingrove) that I find perplexing. It seems each region has handed the camera to someone they know and let them shoot for a few hours/days and then post the results. I don't blame the operators, but it's not great marketing, and a recipe for disaster.

As for viewing on the web. This camera will be used in a vast amount of production that will only ever be distributed via the net. It is of paramount importance that it looks good on the web. Other cameras do, so should this. Saying you can't judge it due to compression etc is missing the point.

My workflow in stills+vid over 20 years has always been to normalise an image and then optionally add a 'look'
That is standard operating procedure whether shooting film or digital. Has always been so.
 
Spot off.

A normalising lut is exactly what's needed as a start point - the extra data captured in a wdr file is then held in reserve for making ajustments in the grade or adding an "artistic" look of some nature without introducing banding or other artifacts.

The first thing I want to achieve with any camera is make the footage look "real" - ill make it look like fogged out of date Koda later.

If i met someone who had skin the colour of the girls on Dens film Id call an ambulance.

Of course we do want to know what the file holds in reserve too - but I'd like to see the basics covered first - no death masks :)

+1. If you think we can't read charts then how is is possible that anyone just twiddling around can do a decent job with log footage?

In The Color Correction Handbook, Alexis Van Hurkman has a whole chapter on this. "... when grading log encoded media... you need to add the LUT first, even though it's the second operation in the image process stack of your application." p. 124 second edition.

He goes on to "normalizing log footage manually" but you see what you can get! Death masks! Love it.

To my eye the whole soft, pale look of folks just not getting log footage back to something with at least some contrast and saturation is a bane. Skin tones and sky all over the place. Expressive? I dunno.
 
Last edited:
As for viewing on the web. This camera will be used in a vast amount of production that will only ever be distributed via the net. It is of paramount importance that it looks good on the web. Other cameras do, so should this. Saying you can't judge it due to compression etc is missing the point.

It's not missing the point at all - web use is extremely undemanding, and there are plenty of users who will have much higher aspirations for footage shot with cameras like this. Judge it on the web for web use, yes, no problem, but judge it on the web for cinema projection - that would be crazy.
Steve
 
It's not missing the point at all - web use is extremely undemanding, and there are plenty of users who will have much higher aspirations for footage shot with cameras like this. Judge it on the web for web use, yes, no problem, but judge it on the web for cinema projection - that would be crazy.
Steve

I didn't say judge it on the web for cinema use. I said, this camera will be used for a lot of web production, so it needs to look good on the web. So far, it doesn't look too hot and it should look stella. It is the target market after all.
 
With regards to normalizing S-log footage I have the option of using several different methods. In PP I can use a fast colour corrector and pull the white and black level to the edges. Add some saturation and ride mid levels to where I want the general exposure. I can also use Cineon converter to do the same thing with a different set of tools. I can use an Alexa LUT (for F3 Slog) in PP and rise black level a bit or I can use a third party plug like Filmconvert.

Depending on the tool I use the final result will look more like video or more towards film.

I admit Slog footage can be a pain to normalize if you don't know where to start but even the simple fast colour correct in PP gets the footage to normal levels so there is an easy tool at hand. Of all the methods above not all will work with all Slog footage. Sometimes I found you have to use different methods to match let's say outdoor scenes with indoor scenes. To me good looking footage comes when you use film curves that takes the edge out of video. Without such a curve you'll end up with high quality video footage that most often does not have the film look. If that's your thing.
 
It may be one of the target markets, but I think the whole design of the camera is that it will appeal to a huge range of markets.
Steve
I agree. More reason to make sure it looks pukka when it first hits the web. Not only to create a buzz, but if it doesn't look good on the web, it definitely aint going to look good on the big silver screen.
 
With regards to normalizing S-log footage I have the option of using several different methods.

Oss.

You know I think you make pretty footage.

But that is not really good enough if it is being handed off - what you (or I can do) is of little relevance. (Ive just seen a F3 job I did - a 'hand off' - and it looks terrible!)

Sony should support a one click solution (and appropriate on camera LUT) that is easy for a third party to use.

The F3 club is a small fanbase of solo ops - the C300 club and alexa club seem to be clubs that prodcos are happy to be part of.

S
 
I hear you Sam. I've also encountered frack ups by clients who does not know how to handle Slog. Sometimes I seriously consider to just go REC709 for them. I urge my clients when handing over footage to do all editing by themselves and let me do the final grading and soundmix. Sometimes they do, sometimes they sadly don't.
 
If feel the camera will make pretty footage - my bigger concern is the market perception - from Reverie on Canon have been the production house darling.

(honestly I think there is a bunch of crap C300 footy out there even at broadcast level -its not a camera I like)

But I contract to larger prodcos (often part of an international shoot) and if you say C300 it seems they are happy - say F3 and they go 'what'

It is my concern about buying this camera - really the image does not matter it is the 'brand' perception that Sony must fix.

S
 
Shot yesterday with F3/Slog and graded to my liking.

View attachment 92013

I'm no fan of the look so far from FS7 sample footage but I see that it has the quality and flexibility that I'm used to with my F3/Slog so I know it will suit me. As for clients I believe Slog3 is just as difficult as Slog so nothing gained except if the user base for Slog cameras grow substantively.
 
Last edited:
Back
Top