FS7: Red Raven and FS7

Funny because the best looking skin tones I've shot in quite awhile were on an Epic, but I also had a great big source and the subjects were all around 18-21 yrs old. They just glowed.
I've wondered whether i could have gotten such a pretty look from my FS7, but unless they're side by side you don't really know anything.
I agree the best deal out there on a camera these days is an F3 hands down. I cried when i sold mine for around $2700 with the 444 card.

With RED before Dragon you have to be very careful to hit the sweet spot. End of the day stuff/dusk for instance, is very challenging for the Epic. The noise goes up very quickly. It also doesn't handle Tungsten/artificial light as well as natural daylight. Asian skin is also problematic, because it can rapidly become orange in post. Studio work with classical hard edged HDMI however, looks terrific with the Epic. I am of course not referring to close-ups. Close ups look good on almost any camera these days. I have seen things shot on a 5D Mark III which are absolutely breathtaking for instance. I am speaking of when a face occupies at most a quarter of the frame.
 
Last edited:
My reason for posting the Rec709a grabs alongside the full grades, was to show just how close the two are to one another. The Rec709a results are a nice neutral starting point that will give anyone nice colours to start with, that they can then easily tweak to taste.

Mark did you shoot these in XAVC or using the SR Codec on the F5? I think XAVC would be the only fair comparison here since we're talking about the FS7.


I agree with many that the closer you can get to shooting RAW on any camera the better the images will look and the more flexibility you're going to have in post to do any type of grading or getting the type of look you want.

Something definitely worth noting are the following data rates across the board that might help put things into a bit of perspective in terms of cameras and options:

HD or 2K 10-bit 422 XAVC 24p = 80Mbps
HD SR Codec RGB 444 24p = 352Mbps
HD SR Codec RGB 422 24p = 176Mbps
2K Sony 16-bit RAW 24p 3:1 = 240Mbps
2K Oddysey 12-bit RAW 24p ProRes 422 HQ = 220Mbps
2K RED RAW 24p 3:1 = 224Mbps

*There's no SR Codec on the FS7 so the option for the FS7 becomes the pricey R5 solution or Odyssey 7Q.
 
Last edited:
Will post some Red and FS7 skin tones later. Not to say which I think are best to leave it as an example of me doing my best with both cameras/workflows.
 
Absolutely. And for me the FS7 is 'almost' at that level, and can get away with it in a lot of circumstances.

Different opinions obviously, but for me the fs7 is at that level. Albeit there are some annoying bugs to work around that should have been fixed a while ago, but the image itself is at that level.

Mark did you shoot these in XAVC or using the SR Codec on the F5? I think XAVC would be the only fair comparison here since we're talking about the FS7.

Mazda stuff was shot XAVC internally.

I like yoclay's insight!
 
Mazda stuff was shot XAVC internally.

I like yoclay's insight!

Yoclay said it well indeed!

On the Mazda video if I had to place a bet, I'd say that Andrea Mancini was shot on a Sony but I've been known to be wrong before. Of course after post work it can address a lot of differences/similarities depending how much work you want to put into it.

If you really want to make any camera shine all you have to do is shoot with a much lower contrast ratio and then a lot of the differences start to blur in my opinion.


Anyone want to guess what camera this is from or isn't from?

Sequence-01_zpsmgokkfif.png
 
HD or 2K 10-bit 422 XAVC 24p = 80Mbps
HD SR Codec RGB 444 24p = 352Mbps
HD SR Codec RGB 422 24p = 176Mbps
2K Sony 16-bit RAW 24p 3:1 = 240Mbps
2K Oddysey 12-bit RAW 24p ProRes 422 HQ = 220Mbps
2K RED RAW 24p 3:1 = 224Mbps

I'd like to see a bit of a reality check on all this data rate stuff.
I'm not saying anyone is wrong about its importance, because I haven't compared them, but one of the reasons I wanted the FS7 was for XAVC-I because its client friendly. Unless you're shooting a high end commercial there are tremendous practical advantages to a lower data rate , not the least of which is getting out of the damn location without going into overtime downloading footage. Do you guys typically shoot RED at 3:1? I never have. Typically its 6 or 7 to 1 as I recall.

Data rate alone doesn't determine real world performance . I remember having an old HVX200 at 100Mbps compared to an EX-1 at XDCAMEX at 35 and the Sony blew the Panny away.
Granted it matters when you are doing heavy effects, probably for extreme grading and perhaps for extreme pixel peeping, but on regular run of the mill shooting and typical grading, will you really see a noticeable difference between XAVC and RAW ?

Anyone know what variety of ProRes the average Arri user shoots ?

Frankly 90% of this garbage just ends up on the web anyway these days.

Lenny Levy
 
On the Mazda video if I had to place a bet, I'd say that Andrea Mancini was shot on a Sony but I've been known to be wrong before. Of course after post work it can address a lot of differences/similarities depending how much work you want to put into it.

Anyone want to guess what camera this is from or isn't from?

Looks cold in there...

Looks cold in there...ha, nice one!

Mancini was Amira and Red. But guessing is albeit very hard. I do think it shows that the differences are very subtle and hard to recognize.

As far as the cold lady I have no idea. But its a nice image. Some people may rag on the blown highlights, but I like highlights blown sometimes. Maybe just a touch more eye light to help focus you on her. Nice image though.
 
I believe someone posted that the Ursa Mini looks almost like an Alexa. Which is absurd...
I said that, side by side, most would prefer Alexa. On its own, the V2 URSA looks beautiful, making the cost of the upgrade quite dubious.

Anyone know what variety of ProRes the average Arri user shoots ?
Lenny Levy
Not sure about an "average", but (if I recall correctly a two year old Rodney Charters interview) it was 422 HQ for "Dallas".
 
Lenny I hear what you are saying. It's not just one thing like bit-depth, data-rate, codec, etc.. but it's definitely a combination of sorts. It's a subjective discussion in so many ways. For example Canon DSLR's and the Sony F3 shoot internal video in 8-bit and they produce some amazing looking footage without question. Same as XAVC-L on the surface might seem inferior to XAVC-I because it's data rate is only 50Mbps when in reality because of efficiencies with XAVC-L it can actually be better than XAVC-I depending on the scene. Also if you limit the contrast of a scene to 5-7 stops you can shoot on the worst of cams and it will look good…

So yeah, lots of subjectivity when it comes to comparing image quality from camera to camera but overall I think it's fair to say that without any codec (such as uncompressed YCrBr or lightly compressed bayer-pattern sensor data) the quality goes up a fair amount, and second to that would be the amount of data bits available to represent it. More is better if you need to manipulate, re-expose, or do a heavy grade in post but you don't really need it if the image will stay "as is". Then there is Log gamma, this sort of half way point of getting more equal data distribution but it's still compressed in the end so can be limiting.

Trying to get back onto the subject of this thread (Raven vs FS7) I think on image quality alone the FS7 will produce some really amazing images but I feel RAW has some characteristics because of the lack of codec and higher bit-depth that makes it look just a little better. Since both have RAW that is fantastic.

I'm glad the FS7 has 12-bit RAW and wish Sony would add an option to shoot with the F55+R5 in 12-bit too!


Looks cold in there...

lol

Knew you guys would like this frame grab. :shocked:
 
Last edited:
Was the quote i was referring to.
Yup. I wrote that. I stand by that.

And keeping on the meme, I read a review of he latest Bond flick earlier today. The reviewer (Digitrends, I believe) panned it. And he panned Roger Deakins too. The reviewer called Rog the "greatest living cinematographer" but he did not like the cinematography of "Spectre". Which was shot by Hoyt van Hoytema. On 35mm film. I mean, these days, who can tell, right?
 
Funnily enough I watched Spectre last week and thought it gained nothing by being shot (for the most part) on film.

The whole opening Mexico sequence was a truly great bit of work but there nothing about it visually that came close to the various looks of Skyfall. Being shot on film didn't make it any more of a film.

Also while Storaro is alive that's a strong statement about Deakins. :)
 
Is this really going to descend into yet another dull 'what's the most cinematic camera' thread?'

I've been doing this for 20 years mate. You really don't need to condescend about the 180deg shutter rule to me. But if for some weird reason I could choose only one, I would take good set design and art direction over a locked 180deg shutter every time.


Cinema is an evolving language. People get used to new vocabulary gradually entering the lexicon. Plenty of films use different shutter angles to aid their storytelling.


Good. I've been doing this for almost 25 years and I am not condescending to you. But I stand by my point. You can have a full blow studio production and if you shoot it at 30 fps it's going to look like a TV show, because at this point in time the public expects 24 fps for feature work in a theater. There will always be exceptions that actually work out (End of Watch), but 48 fps for The Hobbit didn't and it wasn't just because it was a horrible movie.

All of this may change as time goes by and the market is fractured beyond the traditional TV / movie divide, but for now it is what it is.
 
Last edited:
I really don't find that big difference between 30 & 24 myself , but maybe I haven't looked that closely recently. To me its the difference between interlaced and progressive that really matters . Interlace screams video.
I remember way back about 30 years ago when we were shooting music videos on film and transferring to video that the 'New thing" became shooting at 30 fps instead of 24, because it appeared to have higher resolution. Everyone was shooting at 30 in film both 16 & 35mm and then transferring to video. Never looked anything like video though - just looked better. At the same time though the video shooters were sending their interlaced video down to a place in LA that gave it a "FilmLook" - part of which was some kind of a 24 fps conversion. It did look more like film. A while later when video cameras started to record progressive you had to shoot at 24 for the elusive "film look ." Now I find its just a subtle difference , but like I said, I haven't tried to compare closely. Most people I know shoot 24 so that probably says something.

Lenny
 
I just don't think the situation is quite as bad as so many people make it out to be with these cameras (especially since the arrival of Sgamut3.cine and SLOG3). Yes, a lot of people struggle to grade Sony footage effectively, but it truly is simpler than most people make it out to be. And I have no qualms with the results.

So here are some grabs from a recent project I shot on the F5. I'm going to post three versions of each image:

By the way Grug I like your stills.
 
So aside from all the successful cinematic movies that have used different shutter angles in their films it doesn't work? Hobbit's HFR was, whether you liked it or not, huge and the films were a smash. Avatar 2 and 3 (shudder) will have similar HFR looks too and will no doubt be the most successful films of all time. So who is to say?

Also, I think a combination of gaming with varying refresh rates and YouTube having 1080 60fps now are making this more and more a moot point. To go back to the original point though, let's you and I shoot a scene. You at 180deg with no art designer or gaffer and me at 90deg with both. I guarantee you which one will be the more 'cinematic'.

Edit: As a Brit I am shooting 25fps 95% of the time... I have shot at 23.976 a few times and then had a non 180deg shutter to avoid flickering. Nobody ever noticed.
 
... At the same time though the video shooters were sending their interlaced video down to a place in LA that gave it a "FilmLook" - part of which was some kind of a 24 fps conversion. It did look more like film..
Lenny
Sitcoms like "Cheers" and "Seinfeld" did it. Off a 1" recorder onto a 16mm film did make it look .. eh, more like film but without the hassles of using film during their "live studio audience" shows. If you look at those programs now, you can see that they did not have the same look as "Cosby" or "Alf" or "Married w/Children", which stayed 100% on video.

PS. As I had mentioned before, one can still repeat this process with digital-to-film, where the final edit is scanned onto a 35mm or sideways onto Vista Vision. Most obviously just use a film emulation software like Film Convert to get the same results at a fraction of a cost and time investment.
 
I love filmconvert, but has there ever been a time when it made it look like something was shot on film, as opposed to being shot digitally and then some film stock emulation put on top?
 
Back
Top