F3: Sony F3 vs F35 vs Red One Camera Test Disaster

eddavid

eddavid

Carbonite Member
https://vimeo.com/73827487

Shot with f3, f35 both in 4444 mode to pix 240
and red one outputted redlogfilm.

shot with cooke s4, superspeed, and ultraprimes

check it out let me know what you think.

I'm a big fan of the f35 with cooke s4 lenses for a nice natural look
 
Thanks for posting. I was surprised how the RED handled the highlights - not very graceful along with a magenta color cast. I agree, the F35 looked the best.
 
Thanks
Using yes glimmerglass 1 - didn't have 1/2 with me

Using 50mm ultraprime, superspeed mk3, cooke s4
 
I do like the F35 footage a lot. Am I mistaken in that it seemed the F3 had slightly the widest DR? The F3 and F35 could have been intercut.

The Red was really struggling with the lighting.
 
F3 does have the wider dynamic range, about two stops more than the F35 by every test/comparison I've seen. However most of those two stops are down in the shadows, so the two Sony cams are pretty evenly matched for highlight latitude.
 
Grug said:
F3 does have the wider dynamic range, about two stops more than the F35 by every test/comparison I've seen. However most of those two stops are down in the shadows, so the two Sony cams are pretty evenly matched for highlight latitude.
Click to expand...

Two stops is a lot.

So for a very similar image, two stops of DR, better low light, lighter weight, easier repairs I think I'll stay with the F3 for awhile longer. It seems as though the F3 is going to remain under valued (especially with S-Log and a 10 bit recorder) for the forseeable future because it's actually great.
 
starcentral said:
F35 clear winner to me. But F3 footage seemed milky, or lacking in contrast.
Click to expand...

Contrast isn't something to judge footage by. You know that contrast is a simple adjustment in camera or in post. With more contrast the F3 would look sharper too. Rolloff, DR, are the type of things you can judge there.

I'm not saying the F3 footage is better, but close enough that I don't want to give up other things like the ISO.
 
I was looking forward to viewing this but couldn't help feeling there were way to many variables to know what I was looking at. Comparative tests are really hard and unless you control the variables really well I think you learn more about the testers than the cameras.
Some questions:

1- Was the F3 set to Slog or what?
2- It looks like the exterior light was changing constantly as well as your exposures, so how can you judge DR or highlight handling. Looks like it got a lot hotter after the F35 test.
3 - I thought it was odd that the blacks looked so milky in the F3 and that just made me wonder what the set-up was . Likewise I didn't like the skin tone on the F3 and know it can be prettier not so magenta.
4 - The magenta high;lights on the RED were just bizarre. I don't know what caused it but something was wrong there. Nobody would be shootng Hollywood features on a camera that looked that rotten.
5 - How were these things handled in post and was the colorist experienced in all 3 workflows?

So why did the F35 look better?

- Was it being set-up by someone who knew that ca,meta better than the others?
- Is it simply so much better that even with a set-up as crappy as the others it makes great pictures?

I don't know
 
1. F3 was set to s log 444 rbg into the pix 240
2. i was looking at blooming from the lenses - how it handled the roll off.
3. its possible i didn't add contrast to the f3 as much as the f35. this was a very rushed and quick test.
4. yes that highlight surprised me. i never had that happen before.
5. it was me the colorist - and i've shot many years on all three systems - but i did the grade in 3 way color correct very quickly and very similarly.

I've shot jobs with the f35 and f3 intercut and its been fine. But I was more interested in motion, highlight handling, baseline color correction, sharpness - that's what I was looking for.

it was definitely not a perfect test but i learned a lot from it.
 
In my tests, The FCP 3 way color corrector simply will not grade Slog correctly unless you are using a curve designed for Log and go through a few additional steps.
It can really look bad, unnatural and highlights just won't translate right.
I've had some luck with a tripack of plug ins including the Red GianT LUT buddy, Arri LUTs and then adding the 3 way and a free plug in called shadow highlight, but without all three I couldn't get anything to look nearly as good as old fashioned 709. Still not perfect even then, but close.
Glad you got something from it though.

i would imagine the same for Red footage but that's an entirely different can of worms.
 
I own both the Sony F3, and a Sony F35. I have the SR-R1 recorder, so I can get the max settings out of both cameras. My lenses are Arri Ultra Primes. I've shot RED, but very minimally.

I don't think this test is terribly informative though - not for those of use who weren't onsite, and aware of the variables that aren't specified. Stress testing a camera is a good thing to do as an operator, but it's easy for people who don't understand the context of the test to misconstrue the results.

Speaking as someone familiar with F35, F3, and S-Log:

* The F3 does have a wider dynamic range. The F35 is rated at 11.3 stops, and the F3 with S-Log is close to 13.5 I think.
* The F3 has a better lowlight range.
* The F35 is "grainier", but the grain, even when visible in stills, melts away in motion. Much more like film grain, feels unique to me among all the grains I've seen in the digital cameras.
*The F3 assigns more stops to shadows, the F35 puts a lot in the highlights, has a roll-off much more like film.
*With the F3 it makes sense to shoot under, and pull up in post. With the F35 you really want to nail the exposure, so you're not pulling up from the grain.
* The F35 just does something different with flesh tones than the F3.
* The F35 is higher resolution, don't let the fact that they both record to a 1080p container fool you - the F35 is pulling from more photosites, the images turn out crisper. And while both the F3 and F35 can record to a 1080p 444 RGB container, only the F35 has photosites RGB for each pixel. The F3 "cheats."

Both cameras are great - but the F35 is it's own animal.

*You have to be super careful when using simple color correction tools - even if the footage is 10 or 12 bit, some tools and programs only process the color in 8-bit.
 
A quick question for both Ed and James, how do you guys think the 10-bit 4:4:4 output from the F3 compares to the 12-bit 4:4:4 from the F35? Sharpness of the F35's supersampled sensor aside, does the 12-bit make much of a difference?
 
Well, keep in mind the F35 can do 10 bit RGB 444. 12bit is an option.

So, with 8 bit, you have 256 brightness levels.
10 bit you have 1024.
12 bit you have 4096.

So, it's color precise color information. Really, the value depends on exactly how heavy you plan to push the color grading, or fx compositing.

That said, the two projects I shot this weekend I used the 10bit RGB mode on the F35. Less data, and the 12bit was overkill for the client.

A professional colorist might have a more educated opinion... I'm glad to have it as an option, but I haven't pushed it to the limits yet.
 
all good comments - i agree my test was inconclusive but useful to me and some other people at least.

The 12-bit gives wonderful options I found for pushing dynamic range naturally
 
Just for those of you interested. I just shot and edited this small preview clip for a larger documentary. Interview shots were all shot on an slogged F3 to Samurai, all other shots on the F35. I find the F35's image so organic and creamy looking with amazing detail too. This is really noticable when shooting objects with great detail such as churches etc. Best Sparky
 
I love the F35 and its unique look, but you obviously do not know how to shoot /process Red - that's all this test proves. My shots on red, including my high-contrast shots - look nothing like this. F3 vs F35 aspect in informative.
 
You must log in or register to reply here.
Back
Top