RED comes into Avid via AMA. Just like XDCam, 5D/7D, or P2 footage. Mt editing machine is 2.5 years old now and copes just fine without any transcodes for HD finish.

Perrone all of these also work just fine with my Mac Pro, FC system which is 3 years old.

Erik not sure were you get your information but SxS solid state records 50Mb/s 4:2:2 on EX camera, I did it yesterday on my PMW-500. Also ARRI use SxS.

I didn't say it was impossible for SxS to record 50mbs 4:2:2. Until the 500 they only allowed 50mbs on cameras that had the optical drive. I was unaware the 500 was SxS only. Strange.

Furthermore, someone mentioned that the F3 can't output true 444 due to debayering but that confuses me because both the Red One and F35 can output 444 and both are single chip sensors that require debayering. Can someone shine some light on this subject?
It's not the deBayering that means no true 444, but rather the relation between output resolution and chip resolution. As David says, a Bayer chip CAN give 444 - as long there are at least 4x as many photosites as output pixels. (And with Red etc, there are more than enough for 444 1080)

It's not that you can't get 1080 444 from any Bayer chip - rather that you can't truly get it from this one. It shouldn't really be seen as bad, since it keeps the photosite area fairly large which brings other benefits. (And 444 recording will make the most of what it does do.)

Does anyone know if the F3 will be 100% content approved by broadcasters like the BBC and National Geographic as is out of the box? As a wildlife filmmaker, being able to shoot clean video in very low light will be great, however, for fast moving wildlife subjects, I need deep depth of field most of the time to stay in focus. Is there a good high power zoom lens that will work with this sensor for wildlife filming?

Thanks,

Leon Lorenz
Canadian Wildlife Productions: Grizzly Bears, Bighorm Sheep in Alberta & BC Rockies DVD Videos (http://www.wildlifevideos.ca)

Does anyone know if the F3 will be 100% content approved by broadcasters like the BBC and National Geographic as is out of the box? As a wildlife filmmaker, being able to shoot clean video in very low light will be great, however, for fast moving wildlife subjects, I need deep depth of field most of the time to stay in focus. Is there a good high power zoom lens that will work with this sensor for wildlife filming?

Thanks,

Leon Lorenz
Canadian Wildlife Productions: Grizzly Bears, Bighorm Sheep in Alberta & BC Rockies DVD Videos (http://www.wildlifevideos.ca)

Power zoom? On PL mount? Or are you talking about the Sony native mount? Deep DOF and S35 sensor are opposites. Stay with 2/3"

This is why I am hoping to see the 2/3" scarlet. For wildlife, sports, etc. Shallow DOF has it's place, but sometimes you just CAN'T chase the focus.

Yes, also with the higher frame rates, the 2/3" Scarlet makes more sense for wildlife. A video zoom with a PL adapter sounds like it would do the job with that camera.

I suspect you'd a Nanoflash or other external recorder for those HD broadcasters.

Is there a good high power zoom lens that will work with this sensor for wildlife filming?
At a recent event, I believe they said that a 15x servo lens was on the way, though not sure about timescale. It's for such reasons that the F3 has a built in grip/rocker, to give direct control. I think the mount is likely to be the F3 native, not via PL.

As far as f stop/sensitivity goes, then bear in mind the F3 is supposed to be considerably more sensitive than an EX, likely about 2 stops, all else equal. Hence, if an EX needs f2, expect the F3 to need about f4. So for parity, the F3 zoom lens doesn't need to have the same stop as one for an EX, and isn't likely to in practice for size/cost weight issues.

To take full advantage of the sensitiivity in low light, you'd then have to go to a fast prime.

RED comes into Avid via AMA. Just like XDCam, 5D/7D, or P2 footage. Mt editing machine is 2.5 years old now and copes just fine without any transcodes for HD finish. For 2K/4K finish, yes it's an offline/online process..

Red's AMA performance is the worse of the bunch. And if you want to get Red out of AMA, you have to transcode.

They are all taking guesses on the price of the Epic Light (formerly Scarlet S35) and the general opinion seems to be priced around the same as the F3. What I find interesting about that is it means RED only has the 2/3s Scarlet to compete with the S35 NXcam.

If the NXcam has 4:2:2 output through SDI or HDMI that would give Sony the advantage in this price range right? Assuming of course the 2/3s Scarlet would be in the same price range of the S35 NXcam.

Sony only have an advantage if the S35 sensor is more important than other factors. The 2/3" Scarlet has other advantageous elements to its specification, so it's more a matter of choice. Certainly, they'll need the HD SDI to match up to the AF100.

If the Epic Light really has 150fps burst and 120fps an 2/3" chip it is worth the price right there, if it is in the same ball park price as the F3. I guess we will have to wait and see on all of these cameras.

Red's AMA performance is the worse of the bunch. And if you want to get Red out of AMA, you have to transcode.

It's fine for basic cuts, and you transcode overnight or whatever.

Thanks everyone for your help to my questains. I hope Sony does make a good zoom for the F3, maybe even Canon will make a good manual one that will handle fast moving subjects like wildlife, as a deep depth of field and lots of zoom is a must. Yes, the 2/3" Scarlet might make a good wildlife camera and who knows what Canon is up to. Exciting times.

A good example of a deep depth of field footage shot with my Canon XLH1 was when I was charged by 2 grizzly bears this past summer at 25 yards. It was a close and sudden encounter and as I had seen lots of fresh bear sign I was carrying the camera already mounted on the tripod. The footage remained focused decently until the last 12 feet which by this time I had went to wide angle. Now if only the F3 will do as well. To see and read the whole story, Google BC filmmaker attacked by grizzlies.

Leon Lorenz
Canadian Wildlife Productions: Grizzly Bears, Bighorm Sheep in Alberta & BC Rockies DVD Videos (http://www.wildlifevideos.ca)

Zoom info from sony:

PL Mount: 11-16, F2.8 Due APR 2011 Price TBD

F3 Mount 18-270, F3.5-6.3 Due Fall 2011 Manual Focus and zoom only Price TBD

F3 Mount 18-270, F3.5-6.3 Due 2012 Automatic Control Price TBD

F3 Mount 17-50, F2.8 Due 2012 Automatic Control Price TBD


That is all the information for their zooms. Most likely you're going to be looking at 3rd party zooms if you need one right now, but remember they will have to be manual only.

Regarding the workflow of RED vs F3, if you just want event / corporate fast stuff record via the HDSDI out onto a nano at lower bitrates. If you want it for high end stuff record RAW.
Epic light or whatever it will be called will have much better specs, be smaller however it
a) Is still nowhere near production which the Sony is
b) Will not have anywhere near the support that the Sony has
c) Will probably have a few gliches and will need updating of builds etc etc

I still like the fact that I can keep investing in decent glass and know that it will not get superseded.

Zoom info from sony:

PL Mount: 11-16, F2.8 Due APR 2011 Price TBD

F3 Mount 18-270, F3.5-6.3 Due Fall 2011 Manual Focus and zoom only Price TBD

F3 Mount 18-270, F3.5-6.3 Due 2012 Automatic Control Price TBD

F3 Mount 17-50, F2.8 Due 2012 Automatic Control Price TBD


That is all the information for their zooms. Most likely you're going to be looking at 3rd party zooms if you need one right now, but remember they will have to be manual only.

These zooms sound interesting most certainly. 18-270 F3.5 automatic controls sounds like something for fast action on the playing field. The zoom range would cover just about anything one needed. 18-270 ... that must be a big sucker! Definitely interested but like others will see what happens when the dust settles!

I honestly hate to admit it, as much as I'm sure people who use to shoot on real 35mm film hated admitting that digital was taking over, but the fact is that today people are shooting professional productions on DSLR's for crying out loud (that's the part I hate admitting) - so now, unless you have a REALLY REALLY good reason to shoot RAW (film blow-out is the only reason I see) with the advent of DLSR's, compact lenses such as PL or 4/3rds and new cameras with sensors optimized for them, RED is DEAD.

I wouldn't agree with the "RED id dead" sentiment, but certainly there are horses for courses.

I own a 7D and find it unpleasant to work with, and the limitations to be substantial.

On the opposite end, I feel that RED pushes 4k as a standard because it suits them to, and they certainly have a bias.

Even if in 5 or 10 years 4k is the standard, I think there's plenty of time to make good money out of the F3. As others have said, the F3 suits productions that want high quality digital, but don't want the time and hassle of a RAW post workflow.

For productions that will only ever be seen in 1080p or lower quality, I think shooting in RED becomes overkill. RED talks about future-proofing, but not everyone can afford to think like that.

I certainly think RED will continue to see success for theatrical work, but they will have stiffer competition for broadcast, web, and direct to disc productions.

I sold my PS Technik mini35 setup while I could still get a pretty polished penny for it. Today I own a HVX200 I can use with 35mm adapaters if I need to. Ironically I own a 7D which I bought for still 35mm photography and to this day the switch has never seen the HD video recording mode but I ought to try it one day to prove myself wrong about DSLR.

I'm now thinking the AF100... but who knows, it's a never ending cycle.

I owned a Letus Ultimate that I used with my EX1 and HVX200. Sold it a month ago. I'm on the preorder list for one of the very first PMW-F3 cameras.

I'm also planning to sell the HVX200 - but I'll be keeping my 7D and EX1. I'm using Nikon Zeiss glass, so I can mount them on the 7D as needed, but they'll be primarily for the F3.

Can I ask how you mount your glass to the 7D? Is it only EF / EF-S that works on a 7D, or does FD mount for example work too?

What I'm using are Carl Zeiss ZF and ZF/2 lenses that have a Nikon mount. I have an adapter ring that conforms the Nikon mount to the Canon EF mount.

Not sure about an FD to EF mount, I've never looked for one.

Considering the epic or scarlet have not and prob. will not come out by the time the F3 does. Does the F3 compete with the updated RED ONE?

Just to see what people think.

Interesting question. Just finished a bunch of reading on the topic, in light of the EPIC being used on the set of the new Spiderman movie. A few high level things to seriously chew on. First, and not the most important, a usable EPIC will cost you well in excess of 30 grand, more like 40 and that does not include lenses. Secondly, and most important, as this will be the most time consuming to set up and that is modifying and adding to your hardware to deal with the increased data that the new RED cams will have and that is 3 times the amount of the old (if you can call it that) workflow/datarate. You will need bigger, faster desktops and several of them in order to keep on top of deadlines even for lowly freelance types and you will need RED's accelerator cards. Not saying it is impossible but you wlll REALLY want to go there with RED to do it. The proposition now is even more daunting in some ways that with the original RED one setup.

Not bashing RED either. The EPIC will be very appealing to many but the hardware requirements at the desktop level have to be factored in and it is not a trivial exercise.

From the people I have spoken with, the F3's advantage is its lower quality to the RED. 80% of RED work is in 2K, in bread and butter stuff - simply for workflow and the F3 holds its own in that area and takes all the RED workflow headaches out of the mix.

From the people I have spoken with, the F3's advantage is its lower quality to the RED. 80% of RED work is in 2K, in bread and butter stuff - simply for workflow and the F3 holds its own in that area and takes all the RED workflow headaches out of the mix.

The Red Epic is for high end feature film and anyone who has experienced working on 35 &16mm film projects appreciates the faster turn around on digital work flow. That said, on smaller productions, that require around 2k and less ... the F3 makes sense. It's obviously a camera for TV commercials, episodics and documentaries and can go film out with 4.4.4 if need be! That's a pretty big range ... to do film out with the F3 will require added components that will cost in excess of 10 k even if you don't get the camera with the pro-offered 3 lens. That still leaves you in and around 30 - 35k in expenses when you factor in good lens. A good investment if you have the work to cover the costs.

... to do film out with the F3 will require added components that will cost in excess of 10 k even if you don't get the camera with the pro-offered 3 lens.

I wouldn't say a film-out requires anything different to 1080p distribution. I actually find film-outs hide undesirable artifacts when compared to a 1080p monitor. I have had several film-outs made from footage recorded to XDCAM EX and the codec did not present an issue.

I can see the rational of designing a camera to use external recorders in order to keeping the base cost down and flexibility high. What bugs me is that the terrible design of the F3 means shooters will need to spend thousands in order to achieve fundamentals of making it balance on the shoulder and have a usable VF. This is acceptable with a cheap DSLR where video is an afterthought but not on a professional "cinema" camera like the F3.

The F3 is intended for digital cinematography. Why would you want to put it on your shoulder?

Since when did digital cinematography excluded shooting handheld? I've never worked with a professional operator who doesn't shoot from the shoulder for eye-level handheld. A handycam form factor is okay up to a certain size and weight but the F3, with even a light lens, is well beyond that in my opinion.

On shoulder because hand held is an aspect (if sometimes overused) of digital cinematography operating as on the sticks or camera crane. The weight of these cameras with 35mm cine lenses tends not to be as favourable for the handicam layout as say a PD150 or Z1, which also have image stabilization.

As overused as handheld is these days, the impression I'm getting
is that most D-Cinema is still being shot the correct way: mounted
to something (anything).

There a scads of shoulder-mount options; to me it makes sense to
get the one you want and put it on. The F3 isn't a run & gun ENG
camera. I just don't see a need for a built-in shoulder support.

Perhaps not the support built in, but the capability of using the camera on the shoulder by having a forward V/F option. In the past, some cameras with a flat base (like the CP16) were extremely good for hand held work.

I just don't see a need for a built-in shoulder support.

I respectfully disagree.

The camera is designed to be used with a motorized zoom (once it is released). With the hand grip and zoom rocker in the position it is in, handholding is going to be very awkward. It may not be designed as an "ENG" camera but the presence of the zoom rocker and motorized zooms suggests that it is aimed beyond just digital cinema. I can see such a camera being used extensively in EFP.

From what I have seen with the limited time I spent with the prototype, Sony have worked wonders with the sensor, that must have been expensive. Why skimp on simple ergonomics? If shoulder mount was too expensive, which I doubt because JVC managed it on far cheaper cameras, the design should have been more modular. I am looking forward to seeing what the NXCAM S35 camera brings in this regard.

The Arri Alexa includes a shoulder mount and its ergonomics are widely praised among operators.

most D-Cinema is still being shot the correct way: mounted
to something

Surely the "correct way" is the way the story (or director) dictates. Would the opening of Saving Private Ryan have been as effective if shot mounted on sticks or even Steadicam?

I think the big problem with handheld is not the amount it is used but the way it is used. The whole point is to mimic documentary. I can't stand constant, deliberate zoom adjustments and crazy long lenses shaking so much that I feel sick.

Of course I agree that SPR was an awesome use of hand-held (the
exception rather than the norm)... that was Janusz Kamiński, after all.
But the cameras he used didn't have *built-in* shoulder supports, either!

But they had forward V/Fs. The built in shoulder pad is optional, for example a Panaflex doesn't have one built in, it's heavy but still pretty good for hand held.

The F3 is intended for digital cinematography. Why would you want to put it on your shoulder?
At the presentation I went to, Sony themselves were promoting the F3 (amongst other things) as the "ideal B camera to something like an F35", and went on to give examples along the lines of "handheld, to get the POV shots, or where space is too tight to use a bigger camera".

As for supports, then Brian really says it all. You either need an external viewfinder (mounted much further forward) or accept the camera being well in front of the operator. Hardly ideal if the whole point is to have a smaller camera for use in tight spaces.

Yes, no doubt it can (and will) be adapted to get by, but why not just design the ergonomics better in the first place?

Ergonomics are everything. The point is not to only have the viewfinder in the correct (forward) position for handheld, but to also have it there so that you can 'hug' the camera for all sorts of operated shots from sticks. It is hugely beneficial to have your face near the nodel point of the camera.

It's the same reason why most operators do not use extension eyepieces when operating 35mm except out of necessity as in when they are using a geared head or are in a spot where they cannot reach the viewfinder.

Hanging your head off the back of a camera is just silly. Not only for handheld, but just about all the time.

Have a look at the Alexa, or any Aaton - these designers obviously care about the camera and take the time to consult with the users.

Ben Ruffell
DP
www.ruff.co.nz/blog

If they were both the same price, same weight, some functionality. I'd still want the F3 more.
I wonder if you ever worked with raw R3Ds, plus, it is not fair to compare 4K and 1920x1080

I'm sure if Sony used the pixel count as a measure of resolution as Red do, the F3 would be approaching 3K or more. Sony's F35 has 12.4 million pixels to achieve 1920x1080 resolution. Compare that to the 8 MP of Red One used for "4k" or the 13.8 MP that Epic use for the headline figure of 5k. Pixel count does not equal resolution with bayer sensors. I also just noticed that Epic requires a whopping 60 Watts!

I don't know the pixel count for the F3, but as it is a Bayer pattern I expect (and hope) it will be considerably higher than 1920x1080.

Perhaps not the support built in, but the capability of using the camera on the shoulder by having a forward V/F option. In the past, some cameras with a flat base (like the CP16) were extremely good for hand held work.

This is why they have threaded mounting holes on the top of the camera near where the EVF would be. This is a build it the way you want it, kind of camera. I've had the F3 on my shoulders with a standard shoulder mounting kit and it worked just fine. All I needed to do was ad an EVF or LCD and you're done.

They designed it like a handycam for a reason... If they built it like a red camera where its just a brick, it would cannibalize their upper end. DP's might take it too seriously and end up having the F3 on productions where the 9000PL might be used.

It was made to look a little consumer for a reason, this is not a mistake or poor judgment. The limitations are there so they don't ruin an entire line of cameras that cost much more.

Face it, if the F3 was the exact same specs and insides, but came in a metal housing, no ugly little view finder on the back, and just an option for a proper evf, this camera would be taken just as seriously as the Alexa. Once an F3 has the 444 option enabled its really a no brainier. This is a powerful camera, but it was put into a consumer body so they don't hurt their sales. They might also be coming out with another F4 type camera that could be just what I described above... who knows.

Interesting take Giuseppe, I believe you are onto something. Still the placement of the LCD?!!

This is why they have threaded mounting holes on the top of the camera near where the EVF would be. This is a build it the way you want it, kind of camera. I've had the F3 on my shoulders with a standard shoulder mounting kit and it worked just fine. All I needed to do was ad an EVF or LCD and you're done.

They designed it like a handycam for a reason... If they built it like a red camera where its just a brick, it would cannibalize their upper end. DP's might take it too seriously and end up having the F3 on productions where the 9000PL might be used.

It was made to look a little consumer for a reason, this is not a mistake or poor judgment. The limitations are there so they don't ruin an entire line of cameras that cost much more.

Face it, if the F3 was the exact same specs and insides, but came in a metal housing, no ugly little view finder on the back, and just an option for a proper evf, this camera would be taken just as seriously as the Alexa. Once an F3 has the 444 option enabled its really a no brainier. This is a powerful camera, but it was put into a consumer body so they don't hurt their sales. They might also be coming out with another F4 type camera that could be just what I described above... who knows.

It does little good to defend the design as most of us are concerned with having usable elements on the camera. I thought that when the EX3 came out it was a perfect type of design for the viewfinder and still do. We can't speculate on what Sony had in mind when they designed the F3 but many have commented over the years on how most backend viewfinders are useless on most prosumer or pro cameras! Now, I can see that it's possible to hide the fact that a camera is being used professionally if it's designed like a consumer handi-cam when it has a back-end viewfinder and in fact, my first documentary on China was done with Sony Hi8 cameras back in the last century for just this purpose! The F3 is a great camera and I think it's going to be highly successful. I might even buy one at some point myself but the viewfinder could have had an option to remove it and accommodate another type of viewfinder in it's stead.

Sounds like the upcoming RedRock Micro EVF is going to be a popular option if Sony enables simultaneous output of HDMI and SDI.

I'm sure if Sony used the pixel count as a measure of resolution as Red do, the F3 would be approaching 3K or more. Sony's F35 has 12.4 million pixels to achieve 1920x1080 resolution. Compare that to the 8 MP of Red One used for "4k" or the 13.8 MP that Epic use for the headline figure of 5k. Pixel count does not equal resolution with bayer sensors. I also just noticed that Epic requires a whopping 60 Watts!

I don't know the pixel count for the F3, but as it is a Bayer pattern I expect (and hope) it will be considerably higher than 1920x1080.

Could you please help me with the Red marketing of the 4k. Because it drives me nuts when I confront this from the producers.


Arriflex 35 III, Aaton A-minima, Canon 5DM2 and 7D(I feel so cheap)

The term 4k started in film when a film maker would scan the frames of the film using a single row scanner that was 4,096 pixels wide. Each line of the film was scanned 3 times, once each through a red, green and blue filter, so each line was made up of three 4K scans, a total of just under 12k per line. Then the next line would be scanned in the same manner all the way to the bottom of the frame. For a 35mm 1.33 aspect ratio film frame (4x3) that equates to roughly 4K x 3K. So the end result is that each 35mm film frame is sampled using 3 (RGB) x 4k x 3k, or 36 million samples. That is what 4k originally meant, a 4k x 3k x3 intermediate file.

Putting that into Red One perspective, it has a sensor with 8 Million pixels, so the highest possible sample size would be 8 million samples. Red Epic 13.8 million. But it doesn't stop there because Red (like the F3) use a Bayer sensor where the pixels have to sample the 3 primary colours. As the human eye is most sensitive to resolution in the middle of the colour spectrum, twice as many of these pixel are used for green compared to red and blue. So you have an array made up of blocks of 4 pixels, BG above GR.

Now all video cameras (at least all correctly designed ones) include a low pass filter in the optical path, right in front of the sensor. This is there to prevent moire that would be created by the fixed pattern of the pixels or samples. To work correctly and completely eliminate moire and aliasing you have to reduce the resolution to close to half that of the sample rate. So if you had a 4K sensor the resolution would need to be dropped to around 2K to avoid aliasing altogether. BUT a 4k bayer sensor is in effect a 2K Green sensor combined with a 1K Red and 1K Blue sensor, so where do you put the low pass cut-off? If you set it to satisfy the Green channel you will get strong aliasing in the R and B channels. If you put it so there would be no aliasing in the R and B channels the image would be very soft indeed. So camera manufacturers will put the low pass cut-off somewhere between the two leading to trade offs in resolution and aliasing. This is why with bayer cameras you often see those little coloured blue and red sparkles around edges in highly saturated parts of the image. It's aliasing in the R and B channels. This problem is governed by the laws of physics and optics and there is very little that the camera manufacturers can do about it.

In the real world this means that a 4k bayer sensor cannot resolve more than about 1.5k without aliasing issues. Compare this with a 3 chip design with separate RGB sensors. With a three 1920x1080 pixel sensors, even halving the resolution with the low pass filter to eliminate any aliasing in all the channels you should still get at least 1k. That's one reason why bayer sensors despite being around since the 70's and being cheaper to manufacture than 3 chip designs, with their own issues created by big thick prisms have struggled to make serious inroads into professional equipment. This is starting to change now as it becomes cheaper to make high quality, high pixel count sensors allowing you to add ever more pixels to get higher resolution, like the F35 with it's (non bayer) 14.4 million pixels.

This is a simplified look at whats going on with these sensors, but it highlights the fact that 4k does not mean 4k, in fact it doesn't even mean 2k, the laws of physics prevent that.

After all that, those that I have not lost yet are probably thinking: well hang on a minute, what about that film scan, why doesn't that alias as there is no low pass filter there? Well two things are going on. One is that the dynamic structure of all those particles used to create a film image, which is different from frame to frame reduces the fixed pattern effects of the sampling, which causes the aliasing to be totally different from frame to frame so they are far less noticeable. The other is that those particles are of a finite size so the film itself acts as the low pass filter, because it's resolution is typically lower than that of the 4k scanner.

Until someone actuall does some resolution tests or Sony release the data we are a bit in the dark as to the pixel count. IF it resolves around 1000TVL, which is about the limit for a 1920x1080 camcorder then it should have a 3.5k sensor or thereabouts.

Thanks Alister. That is the most cogent write up on the Bayer process I have read to date.

I have reviewed what I wrote and realised that some of it may appear incorrect as I have mixed up pixel resolution and TVL/ph resolution in the same sentence in a few areas so I have re-written it and it should make more sense.

First lets clarify a couple of terms. Resolution can be expressed two ways. It can be expressed as pixel resolution, ie how many individual pixels can I see. Or as line pairs or TVL/ph, or how many individual lines can I see. If you point a camera at a resolution chart, what you are talking about is line pairs, or at what point can I no longer discern one black line from the next. For the black lines to be separated there must be white in between, so TVL/ph is a combination of BOTH the black and white line so will always be a lot less than the "pixel" resolution. With video cameras TVL/ph is the normally quoted term, while pixel resolution is often quoted for film replacement cameras. I believe the TVL/ph term to be prefferable as it is a true measure of the visible resolution of the camera.

The term 4k started in film with the use af 4k digital intermediate files for post production and compositing. The exposed film is scanned using a single row scanner that is 4,096 pixels wide. Each line of the film is scanned 3 times, once each through a red, green and blue filter, so each line is made up of three 4K pixel scans, a total of just under 12k per line. Then the next line is scanned in the same manner all the way to the bottom of the frame. For a 35mm 1.33 aspect ratio film frame (4x3) that equates to roughly 4K x 3K. So the end result is that each 35mm film frame is sampled using 3 (RGB) x 4k x 3k, or 36 million samples. That is what 4k originally meant, a 4k x 3k x3 intermediate file.

Putting that into Red One perspective, it has a sensor with 8 Million pixels, so the highest possible sample size would be 8 million samples. Red Epic 13.8 million. But it doesn't stop there because Red (like the F3) use a Bayer sensor where the pixels have to sample the 3 primary colours. As the human eye is most sensitive to resolution in the middle of the colour spectrum, twice as many of these pixel are used for green compared to red and blue. So you have an array made up of blocks of 4 pixels, BG above GR.

Now all video cameras (at least all correctly designed ones) include a low pass filter in the optical path, right in front of the sensor. This is there to prevent moire that would be created by the fixed pattern of the pixels or samples. To work correctly and completely eliminate moire and aliasing you have to reduce the pixel resolution of the image falling on the sensor to less than that of the pixel sample rate. You don't want fine details that the sensor cannot resolve falling on to the sensor, because the missing picture information will create strange patterns called moire and aliasing.

It is impossible to produce an Optical Low Pass Filter that has an instant cut off point and we don't want any picture detail that cannot be resolved falling on the sensor, so the filter cut-off must start below the sensor resolution increasing to a total cut off at the pixel resolution. Next we have to consider that a 4k bayer sensor is in effect a 2K Green sensor combined with a 1K Red and 1K Blue sensor, so where do you put the low pass cut-off?
As information from the four pixels in the bayer patter is interpolated, left/right/up/down there is arguably some room to have the low pass cut off above the 2k of the green channel but this can lead to problems when shooting objects that contain lots of primary colours. If you set the low pass filter to satisfy the Green channel you will get strong aliasing in the R and B channels. If you put it so there would be no aliasing in the R and B channels the image would be very soft indeed. So camera manufacturers will put the low pass cut-off somewhere between a bit above green and a bit below leading to trade offs in resolution and aliasing. This is why with bayer cameras you often see those little coloured blue and red sparkles around edges in highly saturated parts of the image. It's aliasing in the R and B channels. This problem is governed by the laws of physics and optics and there is very little that the camera manufacturers can do about it.

In the real world this means that a 4k bayer sensor cannot resolve more than about 1.5k TVL/ph (3k pixels ish) without serious aliasing issues. Compare this with a 3 chip design with separate RGB sensors. With three 1920x1080 pixel sensors, even with a sharp cut-off low pass filter to eliminate any aliasing in all the channels you should still get at least 1k TVL/ph. That's one reason why bayer sensors despite being around since the 70's and being cheaper to manufacture than 3 chip designs (with their own issues created by big thick prisms) have struggled to make serious inroads into professional equipment. This is starting to change now as it becomes cheaper to make high quality, high pixel count sensors allowing you to add more pixels to get higher resolution, like the F35 with it's (non bayer) 14.4 million pixels.

This is a simplified look at whats going on with these sensors, but it highlights the fact that 4k does not mean 4k, in fact it doesn't even mean 2k TVL/ph, the laws of physics prevent that. In reality even the very best 4k pixels bayer sensor should NOT be resolving more than 3k pixels or about 1.5k TVL/ph. If it is it will have serious aliasing issues.

After all that, those that I have not lost yet are probably thinking: well hang on a minute, what about that film scan, why doesn't that alias as there is no low pass filter there? Well two things are going on. One is that the dynamic structure of all those particles used to create a film image, which is different from frame to frame reduces the fixed pattern effects of the sampling, which causes the aliasing to be totally different from frame to frame so it is far less noticeable. The other is that those particles are of a finite size so the film itself acts as the low pass filter, because it's resolution is typically lower than that of the 4k scanner.

Until someone actuall does some resolution tests or Sony release the data we are a bit in the dark as to the pixel count. IF it resolves around 1000 TVL/ph, which is about the limit for a 1920x1080 camcorder then it should have a 3k sensor or thereabouts.

There's only one extra thing that I think you may need to add to that, Alister, and it's the definition of TVL/ph - which I understand to be "TV line pairs/horizontal".

If we are talking about 1920x1080, the "pixel resolution" you talk about will be (theoretically) just that - 1920x1080. It follows that you can expect the equivalent figures expressed in line pairs to be 960 horizontally, and 540 vertically. Important thing to realise is that talking about 960lp horizontally, and 540lp vertically are both referring to lines the same distance apart on a chart - albeit at 90 degrees to each other.

Hence the introduction of lph - line pairs referenced to the horizontal. What this means is that resolving a pair of lines a given distance apart will always be given a fixed value, regardless of whether they are vertical or horizontal lines - or even diagonal. So, on the vertical axis, a resolution of 540 lp, will be exactly the same thing as 960 lph.

This all becomes especially important when charts are used with circular resolution bands, or zone plates. It means that a ring can be given a unique lph figure which is equally valid at any point around the ring.

It follows that for a 1920x1080 recording system, the maximum resolution that can be got is 960 lph. If anyone claims and sees more than that - they must be seeing aliasing.

Thanks much for this clarification.

Thanks David for adding that. One issue is that TVL/ph and Lph can be a little higher than 1/2 horizontal pixels because it is measured as the extinction point of the pair of pixels, ie the point where you can no longer see one black pixel separated from the next on the chart, so this implies that the white pixels can no longer be seen (or measured) so your actually looking at less than 2 pixels. When you measure using a scope you are looking for the point where both the white and black lines both become 50% grey That's why it is not impossible to see a measured lph resolution slightly higher than half of the pixel resolution.

In the real world this means that a 4k bayer sensor cannot resolve more than about 1.5k TVL/ph (3k pixels ish) without serious aliasing issues. Compare this with a 3 chip design with separate RGB sensors. With three 1920x1080 pixel sensors, even with a sharp cut-off low pass filter to eliminate any aliasing in all the channels you should still get at least 1k TVL/ph. That's one reason why bayer sensors despite being around since the 70's and being cheaper to manufacture than 3 chip designs (with their own issues created by big thick prisms) have struggled to make serious inroads into professional equipment. This is starting to change now as it becomes cheaper to make high quality, high pixel count sensors allowing you to add more pixels to get higher resolution, like the F35 with it's (non bayer) 14.4 million pixels

1.5k l/ph for a 16:9 array is 1500 * 16/9 = 2666 horizontal resolution. Actual measurements of 4k RED performance are around 3.2k (with negligible aliasing), which reversing the calculation would lead to ~1800l/ph. The issue of where to set the optical low pass filter "correctly" is shown to be a non-issue by the low aliasing nature and higher measured resolution of the images produced by the system, in comparison to other Bayer pattern based systems, three-chip systems and RGB stripe systems.

1kl/ph on a 3chip+prism system is, for the reasons you explain above about optical low pass filters only achievable with quite visible aliasing. 1kl/ph is 1.9k horizontal resolution, very close to the actual pixel resolution of the sensors with no-where-near enough "room" for any reasonably amount of optical filtering to work in. Good optical filtering, as you point out, is a necessary component for all types of cameras, and will necessarily reduce resolution when implemented properly. This is tough when you only have as many samples as is necessary to produce your HD image as you're now in a real battle between soft and aliasy.

Optical low pass filters don't come in sharp-cut-off varieties. I wish they did as they'd make life oh-so-much easier. The lack of control over the roll-off independent of the strength of the low pass is due to the lack of negative photons. Such darkons would make lighting so much easier too :-)

F35 is RGB stripe, with 12 million pixels used to make the image in a 1920x1080x3x2 array. Although using significantly more pixel in it's colour filter array, it manages to make for strong vertical luma aliasing and strong horizontal chroma moire rainbows due to the RGB stripe pattern coupled with low strength OLPF.

When talking about camera systems, it's vitally important to properly measure with a good high resolution circular zone plate to show resolution, aliasing and MTF performance of the system. All these factors go hand in hand and are readily visible and comparable with a single image shot on these charts. They are an invaluable tool in camera analysis.