Home of Professional Cinematography since 1996

F5 v F55

 

I'm about to take the plunge and am trying to decide which of the following two imperfect options to
choose: F55, or F5 + R5.

For me, it the choice becomes: Narrow gamut 4K raw vs. 4K wide gamut XAVC.

It's my general feeling that, everything else being equal:

-- 4K shots down-sampled to HD appear sharper and contain fewer artifacts than HD originated shots
even when both are viewed at HD/709.
-- Extra stops of dynamic range translates to better contrast images even when viewed with a lower
DR monitor.

So by analogy, it would seem to follow that a wide-gamut image "down-colored" to 709 might produce
superior color as compared to a 709 image even when both are viewed in 709.

But others have predicted (in theory) that you won't see any difference in the color on a 709
monitor. Leaving theory aside, has anyone actually seen a difference in the two gamuts?

Cheers.

-- Noel Sterrett Admit One Pictures


 

> But others have predicted (in theory) that you won't see any difference in the color on a 709
> monitor.

my reading of your question leaves me wondering if you are combining or mixing a color gamut question with a dynamic range question

read the next sentence twice as it doesnt make sense immediately
Important to separate the difference between Slog dynamic range and 709 dynamic range from the difference between a larger color gamut from 709 color gamut. The two are two different and virtually unrelated aspects of the 709 spec.

I would argue that the theoretical (and likely practical) advantage of a wider color gamut as stuffed into 709 is most likely to show up when doing strong color correction where you are stepping outside the 709 color gamut during correction process.

The difference between a larger dynamic range used when shooting - then rendered into 709's dynamic range is a bigger and more immediate advantage - not clipping at capture and not buried in noise at capture means choices can be made when going to 709 without clipping artifacts.

For me, unless I am shooting in very controlled conditions, for 709 delivery I would take the wider dynamic range over the higher spatial resolution every time whereas for most situations I dont think the advantages of the wider color gamut are as clear...

and that comment from someone who thinks 4 perf 35 is wimpy and Vistavision is just about right with regard to spatial resolution;-)

weingartner
la
dp


 

> my reading of your question leaves me wondering if you are combining or mixing a color gamut question with a dynamic range question

Actually, no. I was just using an analogy. More pixels/gamut/dr is better even when viewed on a
lower resolution/colorspace/dynamic range monitor.

> I would argue that the theoretical (and likely practical) advantage of a wider color gamut as stuffed into 709 is most likely to show up when doing strong color correction where you are stepping outside the 709 color gamut during correction process.
>

Again, there could well be a significant difference in the way the sensor deals with all colors,
not just outside 709. All colors could be more accurate... or not.

Cheers.

-- Noel Sterrett Admit One Pictures


 

> Leaving theory aside, has anyone actually seen a difference in the two gamuts?

Yes, in fact it is quite clear to my eyes.

In fact when Andy Shipsides here at AbelCine created our basic scene files for the two cameras he had to account for the differences between how they each handled color. Take a look here:

http://blog.abelcine.com/2013/04/03/sony-f5-f55-scene-files-from-abelcine/

Given the choice of the F55 shooting 4K XAVC and the F5 w/recorder back to shoot 4K RAW, I'd choose the F55 any day. About the same price when you account for the memory needed for the RAW recorder. And is definitely splurge on the OLED viewfinder. It makes a huge difference.

Mitch Gross
Applications Specialist
AbelCine NY


 

> Yes, in fact it is quite clear to my eyes.

I wish Andy had shot something that showed how different they were instead of showing how they matched. I haven't compared them to each other yet, but the one thing I do know is that the F55 handles blue very differently to the F5: when balanced for tungsten light the F5 will see daylight as very blue, whereas the F55 will see it as a much less saturated, more pale blue--to the point where some series DPs who shoot with the F65/F55 (same CFA) don't bother using CTO on windows when shooting with tungsten light indoors. Daylight looks much the way the eye sees it.

I don't know how Sony does it but there must be a look-up table that rolls off blue saturation as brightness increases, in an Alexa-like manner, and there's something that makes it easier to do with that CFA than with the F3/F5's CFA.

There's a LOT going on under the hood of both cameras. What's most interesting is to see how a Chroma Du Monde bulges on a vectorscope between green and yellow, particularly on the F55, as the extended green gamut is compressed to fit into Rec 709. The chips smear radially, which means that each one is actually multiple colors and is a clear sign of gamut compression.

I'd think, though, that buying an extended gamut camera just to take advantage of how the extended gamut downsizes into Rec 709 isn't the right approach. I've had several colorists tell me that it's rare for a color to be saturated enough to poke dramatically outside of the Rec 709 gamut. I'd think it would be better to select a camera based on how it looks in Rec 709, knowing that this is the worst-case (and most common) scenario and that it'll almost always look better in wide gamut–if you actually take advantage of those additional colors.

-----------------------

Art Adams | DP
San Francisco Bay Area


 

> There's a LOT going on under the hood of both cameras.

As I understand it, the F5 and F55 are very, very different. The imaging chip, processing and in fact the entire signal path in the F5 are based on the F3, while those same things in the F55 are based on the F65. That alone should tell you something about the differences you're seeing. In many ways, apart from the body, lens mount, and name, they are more different than they are alike.

Personally, I'm not sure I really understand the purpose of the F5 at all. But maybe that's just me...

Mike Most
On Location Services Director
Technicolor Hollywood
Los Angeles, CA.


 

I wish Andy had shot something that showed how different they were instead of
showing how they matched.

It's a funny thing. No matter how much we try to show what range cameras are capable of delivering, what almost everyone asks for (unlike Art) is how to make them look the same. They want us to create ways to match these different cameras. An Alexa look for the C300. A Canon 5D look for the FS700. I'm shooting a doc on F3, F55, C500, AF100, HDX900 and Alexa -- give me look settings to make them all match. We get this EVERY DAY.

We used to have one base look, which Jesse Rosen established to make a nice image on the original Varicam, called JR45 (.45th gamma, inverse way of saying 2.2). It was later followed by the Video Rec and Film Rec versions and then we started doing it for all the cameras. I had a couple of MITCH files out there until it started getting complicated enough that we settled on AC files for AbelCine. We have four main variants (high saturation, low saturation, etc.) with a couple of other fun ones tossed in for various cameras.

It's what everyone wants to see. We tell people that these are base settings, starting off points for them to then work beyond to create their own looks. But I have a feeling it's one out of fifty who ever do anything other than the look we load. They just light and expose from there.

To get a sense as to how different the color rendering is between the two cameras, compare the parameters we use to try to match them and the level of offset between the two cameras on certain settings. There's a real difference.

Personally I find that the colors are far truer and more nuanced on the F55. It's akin to oversampling, which is what Noel was getting at. If the CFA can get closer to landing the correct colors to the sensor then when that gets crushed down into REC709 it has a far better chance of landing at the correct spot. More importantly and to Art's point, even when the color is already within REC709 (as most are), the rendition will be more accurate because of the saturations and purities of the CFA dyes. It's the same reason you shoot 35mm for SD video finish -- get more in there and get it as close as possible to what you want from the start. It will always end up more accurate that way. You wouldn't want a lens that could only just barely resolve the format's resolution so why have a CFA on a chip that can only just barely cover the color range of the signal path? I do overstate the point to get te idea across, and even the F5 has a CFA which is significantly wider gamut than REC7
09, but the F55's CFA is MUCH wider gamut. Wider even than film. Wider than anything is currently able to reproduce. That's a good thing -- there's always got to be a limiting factor and let it be at the very end of the chain and not the beginning. Once you get crush down there's no way back out again.

Mitch Gross
Applications Specialist
AbelCine NY


 

> ...the one thing I do know is that the F55 handles blue very differently to the F5

Shooting raw or XAVC? Before or after grading?

Cheers.

-- Noel Sterrett Admit One Pictures


 

> To get a sense as to how different the color rendering is between the two cameras, compare the parameters we use to try to match them and the level of offset between the two cameras on certain settings. There's a real difference.

No doubt they're different, but as you point out the cameras still can be graded together rather
closely.

> Personally I find that the colors are far truer and more nuanced on the F55.

That's the real question here. I'd love to take a look if you have any example shots.

> ... even the F5 has a CFA which is significantly wider gamut than REC709, but the F55's CFA is MUCH wider gamut. Wider even than film. Wider than anything is currently able to reproduce.

Perhaps wider than we can see? We work hard at blocking UV and IR, which unfortunately also impacts
the colors we can see.

So far, the color of the F5 raw shots I've done is excellent (once graded). Very natural skin tone
even under unmatched florescent lights. And with raw, the latitude in grading is simply amazing.
Almost makes it look like I know what I'm doing!

So I'm hesitant to trade raw for a much more compressed and baked-in image with a gamut I may never
be able to see.

As Mitch pointed out, the price point is similar for F55 v. F5+R5.

Cheers.

-- Noel Sterrett Admit One Pictures


 

You can dumb down the F55 to look a lot like the F5, but it's not so easy the other way around.

The F55 chromacity is tuned within the visible light spectrum. Actually it is quite close to the proposed REC2020 that is under development for Ultra HD. Guess why?

Mike Most asked what the point of the F5 was. That's a good transitional camera, and certainly Sony was able to bring it in at a more comfortable price point for many. But I'm all about the F55 myself.

Mitch Gross
Applications Specialist
AbelCine NY


 

To fully preserve the Sony F55 colour gamut I think you need to record raw to the R5.
So unless you are looking at F55 + R5 then this advantage over the F5 is reduced.

To fully preserve the Sony F5 or F55 dynamic range you also need to record raw to the R5.
This is because the Sony R5 is recording 16 bit linear data.
Meaning that there is no dynamic range compromise when compared to 10 bit Log or 709 encoding.
Grading from raw is straightforward and should translate into better results.
In addition I find that processing to other formats is faster from raw vs XAVC.

So in answer to your question you need the R5 recorder and some decent lenses.
Now see how much money you have left over and choose the F55 or F5 !

Martin Greenwood

web: yoyotta.com


 

Perceived color aside, you can always add the RAW recorder to an F55 after
you get some ROI momentum.
On the other hand, you can not add the wide-gamut CFA or Global Shutter to
the F5.

Ken Wortendyke
Producer | Director | DP | Post
Washington, DC Area


 

> To fully preserve the Sony F55 colour gamut I think you need to record raw to the R5.

While technically that's probably true, other recording formats on the F55 can use a combination of SLog2 and SGamut, which "effectively" gets you nearly all of the usable information that you could reasonably make use of with current and near future display and processing systems. So if you are expected to post and/or deliver in HD - as you might reasonably be asked to do on a television project - there are alternatives that might work quite well and allow use of lighter files and thus far less storage. And although storage might be "cheap," it is far from a non-issue when you're trying to process, deliver, and store for conform 3 plus terabytes a day for 8 days on each episode. By way of comparison, using ProRes4444 at HD resolution usually produces about 3 - 4 TB per **episode**.

I'm most definitely not arguing against the logic of recording Raw, even for television material. But the reasons for that should go beyond the color gamut and dynamic range arguments, assuming the show is being shot by a competent cameraman. And the time, budget, and workflow need to be solidly in place to accommodate the much larger data payload prior to going down that road, because the notion that it's not going to cost any more or be any more complex to handle and store is just not true, at least in the vast majority of cases.

Mike Most
On Location Services Director
Technicolor Hollywood
Los Angeles, CA.


 

> While technically that's probably true, other recording formats on the F55 can use a combination of SLog2 and SGamut, which "effectively" gets you nearly all of the usable information that you could reasonably make use of...

I assume that would be the same for the F5 as well.

> And although storage might be "cheap," it is far from a non-issue when you're trying to process, deliver, and store for conform 3 plus terabytes a day for 8 days on each episode.

The Sony F5/55 raw datarate is around 425 MB/s so 3TB would be 7 hours of footage. That may be what
is shot per day for television, but it is far more than I have ever shot or will shoot on at least a
low budget feature. I'd be amazed to get over 2.

The raw files need not be immediately processed, proxy XAVC files for editing can be recorded with
the raw, and both can be copied simultaneously with a rather simple ingest setup. The raw files
serve as a backup, help future proof the material, and can be used in a pinch to save a bad shot.

Cheers.

-- Noel Sterrett Admit One Pictures


 

> To fully preserve the Sony F55 colour gamut I think you need to record raw to the R5. So unless you are looking at F55 + R5 then this advantage over the F5 is reduced.

 

I'd love to see a direct comparison between an F55 AVC and F5 raw. But given the budget, it's really
a choice between a full F5 kit or a bare F55. No doubt Sony designed it that way.

The problem is that I'm becoming increasingly addicted to raw. Anyone know of a good rehab clinic?

Cheers.

-- Noel Sterrett Admit One Pictures


 

> other recording formats on the F55 can use a combination of SLog2 and SGamut, which "effectively" gets you nearly all of the usable information that you could reasonably make use of with current and near future display and processing systems.

I'm with Mike on this one….

Jim Houston
Starwatcher Digital
Pasadena, CA


 

> The raw files need not be immediately processed, proxy XAVC files for editing can be recorded with the raw.....

If you don't manufacture dailies from the Raw recording, it is likely that you will never know if there is a problem with those recordings until it's too late. Parallel recording paths are one of those things that sound good - especially to manufacturers - but in practice can lead to a number of very bad discoveries, such as mismatched metadata (making for a rather chaotic conform) and technical problems with the set of files that wasn't used for making dailies, as noted above. I guess I'm not the trusting type when it comes to these kind of things. But it's only because I learned the hard way. Not to mention that you still need to go through a dailies processing path if you want proper color and double system sound sync, things that are expected these days, especially in television work.

As for data payload, 5+ hours per day is not uncommon in television, where just about everything is shot with at least 2 cameras all the time these days, and where many directors seem to now have a serious aversion to the word "cut."

Mike Most
On Location Services Director
Technicolor Hollywood
Los Angeles, CA.


 

> The Sony F5/55 raw datarate is around 425 MB/s so 3TB would be 7 hours of footage.

You are correct that 3TB is about 7 hours, but the data rate is too high.

F5 and F55 4K raw is 5MB per frame or 120MB/s
4K XAVC is 1.8MB per frame or 43MB/s
HD ProRes 1.4MB per frame or 34 MB/s

Martin Greenwood

YoYotta Ltd


 

There isn't a huge difference between XAVC sLog2 and RAW latitude. It's
mostly in the extreme highlights that you get significant benefit. And I
mean extreme.

Juan Salvo
Colorist