Scarlet vs F23

[David Rasberry]

Digital Photo Pro magazine typically has at least one article a month on some aspect or another of processing raw files. Still photographers have been dealing with raw files a lot longer obviously, and have developed more techniques. There are some interesting aftermarket plugins available for Photoshop that might be worth trying out with batch processing too. Worth a look.

 

[Daniel Browning]

Personally, I think that maybe RED needs to publish a white paper on current RAW processing practices in the industry and who is doing it right and how it can be done wrong.

It would take a library to fit all the paper necessary to describe how it can be done wrong.

Actually, there is already a library full of books, white papers, videos, etc. on raw processing practices thanks to the prevalence of DSLR still cameras.

The only difference is that R3D has audio and more frames. More frames means that raw processing techniques and software need to be fast. I guess another difference is that DSLR raw formats tend to get reverse engineered very quickly and supported by dozens of different applications, whereas the R3D format is only available indirectly through the SDK.

Other than that, I can't think of any reason why the volumes of literature about DSLR post processing can't be applied to RED.

 

[David Mullen ASC]

I'm more curious as to how post houses are converting and correcting R3D footage, not what the still industry is up to. The fact that there are a million ways to do it is not necessarily a good thing when trying to track down why some RED footage looks better than another, beyond how it was photographed. It's in RED's best interest that their footage is properly processed lest they get blamed for image problems that don't actually exist in the original files.

 

[Tom Lowe]

Learning to properly post RAW takes time. It doesn't surprise me that the guys at PlasterCity would already have some nice tricks of the trade up their sleeves for RED. The bottom line is, you either need to spend a lot of time learning RAW for Red, or pay someone who knows it well. But I do support David's idea of Red backing some tutorials or seminars or publications of some sort, trying to improve the RAW processing abilities of those who use the camera. Frankly, I have seen Red footage that looked like an embarrassment, but that could have easily been improved with a few tweaks here and there.

 

[killfilm]

I still think it's a bit of an apples to oranges comparison. There is a fundamentally different design approach here -- the Sony F23 is a broadcast video camera, it's designed to deliver on location or stage the best HDCAM-SR recording you can get, it's endlessly adjustable for an engineer, and has a ton of options. Once you're done shooting, if you used certain parameters, you could hand the tape off to a broadcaster and they could air it. That ability is partly what you're paying for, plus the whole HDCAM-SR deck, which in itself accounts for almost half or a third of the camera's costs.

A RAW camera is much simpler in some ways, the basic idea is to get the data into the post world and then decide what you want to deliver from that data. In theory, that sounds like a great idea, but in practice, we still have a post world that revolves around HD videotape and HD codecs.

Beyond that, there are other differences in manufacturing specs, materials used, etc. Plus RED's manufacturing practices are still a mystery but clearly they've found ways to spend less on certain aspects than Sony.

I got an email from a DP who asked me how to get rid of the noisy blacks in the RED. I suggested that he take the footage over the PlasterCity to see if the problem was in the original R3D files or not, at which point he admitted that he actually used PlasterCity the first time and that earlier footage came out fine, but now the producer insisted on processing RED footage himself and now the footage was coming back with noise problems. So clearly there is a weak link here, there are too many ways to process the R3D footage and not enough people with experience doing it.

Personally, I think that maybe RED needs to publish a white paper on current RAW processing practices in the industry and who is doing it right and how it can be done wrong.

with all due respect david, i dont see RED hitting the broadcast market like beta SP/digi beta unless they develop a similar workflow like P2 etc.
I think that RED's main target is obviously digital cinematography, nonetheless we need a workflow with less headacehes. I tried out the premiere workflow and it aint too bad. Needs a few tweaks here and there.

You probably will tell the difference easier than most people between the F23 and HVX200 and it was evident in cloverfield. No one in the Cinema complained at the quality, just the shaky footage though. Point being is that the HVX is a not even full raster 1080p and it did an awesome filmout.
I see your point of single vs 3chip and im wondering if there are techniques that help us by cramming the pixels a different way? or is that it, single CMOS wont even compare to three chip. let me know please.

 

[Stephen Williams]

No one in the Cinema complained at the quality, .

Then why did the HVX 200 not already kill film?

 

[GPSchnyder]

Then why did the HVX 200 not already kill film?

Because it doesn't work for any movie. Like 16mm doesn't. And like 25mm doesn't. And like the RedO... I bet you get the point.

It can work, but not everywhere.

 

[Stephen Williams]

Because it doesn't work for any movie. Like 16mm doesn't. And like 25mm doesn't. And like the RedO... I bet you get the point.

It can work, but not everywhere.

Hi,

I don't think the HVX 200 is as good as 'killfilm' is saying.

YMMV

Stephen

 

[Harry Clark]

QUOTE >Personally, I think that maybe RED needs to publish a white paper on current RAW processing practices in the industry and who is doing it right and how it can be done wrong.<

Amen to that David. Been waiting for that for about a year now. They'll say that "the technology is moving too fast for that old-fashioned model" or "we are a camera company, not a post company" or some other such thing.
Having options is great. Having NO roadmap is not. And I think that the DSLR workflow with Aperture or Photoshop is not applicable to RedCine, or going to a Scratch room somehwere.
Sorry to sound snarky but post's been a frustrating part of it for me...
Cheers,
Harry

 

[GPSchnyder]

Hi,

I don't think the HVX 200 is as good as 'killfilm' is saying.

YMMV

Stephen

And you may be right. But that doesn't change the fact that it can be used for filmouts in some cases.

But that doesn't matter, as the Scarlet will have a way better Quality than the HVX 200 and will work wonders presented digitally (Who needs filmouts anyway?)

 

[Jim McKinney]

I'm more curious as to how post houses are converting and correcting R3D footage, not what the still industry is up to. The fact that there are a million ways to do it is not necessarily a good thing when trying to track down why some RED footage looks better than another, beyond how it was photographed. It's in RED's best interest that their footage is properly processed lest they get blamed for image problems that don't actually exist in the original files.

I agree that a white paper would be extremely beneficial to everyone, especially Red Corp. I'm working with a director right now who's post-production partner almost refuses to work with Red footage. This is because nearly 18 months ago they had one bad experience! The director's on board, but his hands are tied with the post people, so our ability to use Red is jeopardized. Instead, we might have to use a lesser, but more "post-friendly" format.

 

[Lawrence Bansbach]

But that doesn't matter, as the Scarlet will have a way better Quality than the HVX 200 and will work wonders presented digitally (Who needs filmouts anyway?)

How do you know that Scarlet will be better? In theory it will be capable of technically better images (greater spatial resolution, bit depth, and dynamic range), but as has been stated elsewhere on this thread, its advantage -- tremendously manipulable RAW output -- may prove its curse. If seasoned pros are having trouble, what hope is there for serious (but less trained) hobbyists and soccer moms? To sell tens of thousands of units, there has to be a simpler, largely idiot-proof, default workflow. Don't frustrate users unnecessarily; if possible, enable decent imagery out of the box and encourage self-education and experimentation.

 

[GPSchnyder]

How do you know that Scarlet will be better? In theory it will be capable of technically better images (greater spatial resolution, bit depth, and dynamic range), but as has been stated elsewhere on this thread, its advantage -- tremendously manipulable RAW output -- may prove its curse. If seasoned pros are having trouble, what hope is there for serious (but less trained) hobbyists and soccer moms? To sell tens of thousands of units, there has to be a simpler, largely idiot-proof, default workflow. Don't frustrate users unnecessarily; if possible, enable decent imagery out of the box and encourage self-education and experimentation.

I don't think Soccer Moms are the Target (anymore). It's a Pro-Tool, so there should be a pro way of dealing with the Footage. Give a Soccer Mom a 16mm Cam and she will not be able to use it too.

And of cause do we not know anything for real about the Scarlet, but then why post here anyway?

George

 

[David Mullen ASC]

I think Jim Jannard once said that the fixed-lens 2/3" Scarlet was aimed at a larger consumer market -- the low price point is based on that.

So it's going to have to be fairly easy to use and handle the footage... or else it will become more of a camera for the Sony EX1 and Panasonic HVX200 crowd and not appeal to anyone below that.

I think it's going to have to be bundled with some sort of really simplified version of Red Alert or something, or some sort of automatic conversion software that kicks in when you plug the camera into a laptop with a simple editing software like iMovie. It's got to be made fairly idiot-proof.

Unless the whole idea behind the fixed-lens Scarlet has changed and it really is aimed at the prosumer crowd and higher, and not for general consumers.

 

[GPSchnyder]

I think Jim Jannard once said that the fixed-lens 2/3" Scarlet was aimed at a larger consumer market -- the low price point is based on that.

So it's going to have to be fairly easy to use and handle the footage... or else it will become more of a camera for the Sony EX1 and Panasonic HVX200 crowd and not appeal to anyone below that.

I think it's going to have to be bundled with some sort of really simplified version of Red Alert or something, or some sort of automatic conversion software that kicks in when you plug the camera into a laptop with a simple editing software like iMovie. It's got to be made fairly idiot-proof.

Unless the whole idea behind the fixed-lens Scarlet has changed and it really is aimed at the prosumer crowd and higher, and not for general consumers.

Yeah, but I think he also said that it's a Pro-Cam, no Consumer-Cam...

 

[David Rasberry]

I would say that any camera, still or motion, over about the $1500 price point would have to be classified "Prosumer" whatever that means. Even in the still camera market, interchangeable lens DSLR's at any price have only about 7% total market share. Cameras costing over $3k of any type are in the sub-1/2% top end of the consumer market and by definition that market segment is going to be sophisticated and serious users, pro or hobbyist.
All it will take for Scarlet is a reasonable native format workflow in Adobe Premier, AE, Photoshop, or other similar programs, to support serious hobbyists/ex-pro shooters like myself that like to play. By the time Scarlet is released, that should be in place. It should be no more difficult to deal with than similar entry level tools available to still photogs for RAW image processing.

 

[Daniel Browning]

I'm more curious as to how post houses are converting and correcting R3D footage, not what the still industry is up to.

That makes sense. If I understand you correctly, what you're saying is that the fact that all the information is out there is not enough: you need something you can point to that is from a respected authority in your own field. It requires expert knowledge to separate the wheat from the chaff.

It's in RED's best interest that their footage is properly processed lest they get blamed for image problems that don't actually exist in the original files.

I, too, would welcome a white paper from RED about raw conversion. Additional good information helps to outweigh bad information, and coming from the manufacturer adds authority in the minds of some. But I don't think it will help the blame game.

A white paper would only reduce one of the infinite reasons that RED is blamed unfairly: someone who did not have enough knowledge to process a certain image a certain way and who would have read a document from RED to increase their knowledge. It's not going to reduce blame caused by other types of images, other styles of conversion, software limitations, human mistakes, people who wont read white papers, subjectivity, communication issues, and a host of other reasons that aren't RED's fault.

In video cameras, raw conversion is the equivalent of in-camera settings. With a video camera, anyone can blame the camera manufacturer with impunity because it's hard to prove that different settings in the camera would have yielded a better image: no one can go back in time to change it and see what happens. At least with R3D it can be proven one way or the other; and you have the option of getting a second opinion. If you post the raw file, then you can get opinions on the web for free.

The still industry is a perfect example of misplaced blame. Most photographers, "expert" reviewers, and web sites are heavy-laden with misconceptions, urban legends, and raw processing myths. They blame camera manufacturers all the time for issues caused by the raw conversion or their own misunderstanding. Here are some of the most prevalent ones:

• The histogram in the converter represents the raw data (not even close).
• The raw converter is a level playing field among different camera models (it's not level at all).
• Most have no clue what raw data really looks like (it's much noisier than what their converter shows).
• Small pixels have more noise, less dynamic range, worse diffraction, and more lens flaws (they don't).
• Setting noise reduction to "off" actually turns off all noise reduction (it doesn't).
• 100% crop is a good way to compare cameras with different resolution (it's very flawed).

Most of the respected authorities in still photography get these wrong, despite the fact that they are illogical and easily disproven. So there is a lot of misinformation out there in magazines, books, and the web. The only way to cut through it is with logic and experimentation.

Anyone with enough time on their hands can become an expert in raw conversion for any raw camera. A bunch of raw files, cheap software, and voracious appetite for knowledge will do the job. There is a lot to learn, but whatever is learned will apply to all raw cameras, not just one specific model or manufacturer.

Video is easier to learn than raw. The manufacturer can't provide enough software and settings inside the video camera to match even a tiny fraction of the possible raw conversion options that are available in post production. Even if they could, they would probably choose to limit options to make it easier. Unfortunately, every manufacturer does things a little differently, so the knowledge does not transfer to other cameras the same way it does with raw cameras. Furthermore, you can't figure out what an image would have looked like with a different curve, white balance, or saturation unless you have the actual camera.

The producer obviously does not have the expertise that Plaster City does, or his blacks wouldn't be noisy. The information and tools are out there for the producer to learn and become better at raw conversion. Unless RED's whitepaper contains new information, original research, or groundbreaking new techniques, there is nothing holding him back from learning better raw conversion right now.

I think it's going to have to be bundled with some sort of really simplified version of Red Alert or something, or some sort of automatic conversion software that kicks in when you plug the camera into a laptop with a simple editing software like iMovie. It's got to be made fairly idiot-proof.

My guess: The camera will default to a new RED RGB codec. RED will distribute free decoder software for Mac and Windows so that most editing applications will be able to support it. Advanced users will configure the camera for RAW shooting and use raw conversion software that takes more time to learn. I don't think it makes sense to work on a newbie-friendly raw conversion software; anyone not interested in investing the time to learn RAW processing can just shoot RGB.

 

[Lawrence Bansbach]

My guess: The camera will default to a new RED RGB codec. RED will distribute free decoder software for Mac and Windows so that most editing applications will be able to support it. Advanced users will configure the camera for RAW shooting and use raw conversion software that takes more time to learn. I don't think it makes sense to work on a newbie-friendly raw conversion software; anyone not interested in investing the time to learn RAW processing can just shoot RGB.

Possibly, but FreshDV conducted a video interview with Ted Schilowitz at NAB. The interviewer asked Ted Schilowitz, "Will this be a RAW workflow similar to Red -- you'll go through Red Cine, same type workflow?" Schilowitz responded, "Different tools -- we're going to develop different tools and different ways to work, but they're all in the same vein as what we're doing now." From this I deduce that the two Davids may be right -- that there will be a simplified RAW workflow tailored specifically to Scarlet. I hope that means defaults that yield good results, with the ability to tweak to your heart's content.

 

[Daniel Browning]

Possibly, but FreshDV conducted a video interview with Ted Schilowitz at NAB. The interviewer asked Ted Schilowitz, "Will this be a RAW workflow similar to Red -- you'll go through Red Cine, same type workflow?" Schilowitz responded, "Different tools -- we're going to develop different tools and different ways to work, but they're all in the same vein as what we're doing now." From this I deduce that the two Davids may be right -- that there will be a simplified RAW workflow tailored specifically to Scarlet. I hope that means defaults that yield good results, with the ability to tweak to your heart's content.

I hope so; that would really great. Automatic conversion software like what David mentioned would allow for higher quality than a real-time RGB codec, and the inconvenience might be small enough to go unnoticed for most people.

 

[Chip G]

Im one to say that i'd like to see Scarlet smoke some ass. I read this article

http://news.http://www.teenink.com/Past/2002/January/CollegeEssays/IHateCows.html.net/story/861234

“The F23 has excellent dynamic range, handling highlights beautifully with a film-like quality,” said Glen MacPherson, director of photography (DP) for “Final Destination 4,” the first digital 3D action movie to be shot on location instead of primarily behind a blue or green screen. “It’s also virtually noiseless in the shadows, and the 2/3-inch image sensor is ideal for shooting 3D. The F23 is my camera of choice for all those reasons.”

The camera delivers 12.5 stops of latitude – more than any other 2/3-inch chip digital cinema camera

Can Scarlet (with what we know) come close to this?

Dunno, but by the time Scarlet comes out the Sony F65 will be coming to market. Wait tell you see the specs on it!