Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

RED narrative production issues- R&D priorities!

"Change" is the only constant word in this chaning world.

Dear morpheusfilm,
Though they are partiall valid points, you don't have to so avert to them, learn how to handle them.

RED is a company that openly accepts the constructive criticism, infact many of the features of RED ONE were suggested by members like you and me.

After few years, unless you are Academy Winner/ASC member, you no choice except to use RED, because it was already decided by your employer not you.

Ever heard of joke "Death by Bunga Bunga" ..? choice is yours.

A quote from 300 "choose your words carefully, they may be your last words as 'cinematographer'".
 
Reiteration, Yes I still like RED

Reiteration, Yes I still like RED

Your post sounds like typical DP that got used to the old type of ...., forgive me, the LEGACY equipment, and has trouble to figure out the new ways the industry is going to operate.
I've been shooting HD for years but I wasn't posting to make some personal attack on the RED as much as evaluating the camera from a real production standpoint, and attempting to explain what I believe are the highest fix priorities. I actually tech for this camera and go out with it from a rental house. We have the latest version firmware build and the playback crashes the camera as many times as it works. There still is no way for the propriatory onboard LCD to work while simultaneously you send a signal to the director- Yes there is a work around that involves mounting a second external on board connected to the HD SDI port, and then you can send the other hdsdi or preview signal to the director but what about the focus assist? You can't use it if you want to send a signal anywhere else. You end up having to switch back and forth in the menu from RED LCD to the HDSDI ports constantly, and carry two on boards if you want to do it fast, which you will need to do. And quiet mode is still bad if you are under 8 ft from your subject as far as sound goes, plus when the camera heats up and you are in a warm environment, the camera shuts down. This has happened several times on set and involved another 5 minute delay while the camera re-boots, with everyone standing around waiting for camera. But hey, take it as me just grinding an axe if you wish but I certainly know that these limitations are real problems with high stress production situations.
 
Lense breathing

Lense breathing

And yes it was the 300mm and the zoom.
 
film latitute vs. digital capture

film latitute vs. digital capture

Graeme Nattress said:
Let's just disagree on that one small point here. None of the published curves for any film stock that I've seen show remotely near 16 stops, never mind more.

Graeme
Click to expand...
Try doing the grey card test with Kodak stock 50D #5201 and will find just how many more stops film has. And again I would point out that each stop is DOUBLE the amount of light so even the difference between 11.3 and the 14 which you suggest is significant enough to make a difference, especially when the RED currently has no way to skew the gamma curve to have more lattitude in the highlights than in the shadows. Sure you can underexpose but then overall contrast is effected and you have to tweek all the client/directors monitors so that they won't be worried by your purposeful underexposure. RAW is pretty nice and simple though. It would just be nice to move the gamma midpoint to either favor the shadows or favor the highlights. This is how you would get the best from DAY EXT. and NIGHT INT.
 
morpheusfilm said:
Try doing the grey card test with Kodak stock 50D #5201 and will find just how many more stops film has. And again I would point out that each stop is DOUBLE the amount of light so even the difference between 11.3 and the 14 which you suggest is significant enough to make a difference...
Click to expand...

I don't think you need to teach Graeme what a stop exactly is.
Anyway, he didn't say 14 stops, he said 13.3 of which 8.3 is linear.
A lot of stock would have considerably less DR than that though.

Kodak themselves show only 11 stops of DR in their graph of 50D 5201.
http://www.kodak.com/global/images/en/motion/products/negative/5201_char2.gif
 
i guess some cameras work better than others, mine has never ever crashed on playback and i have used it constantly since 9/7/07 and playback since 11/2.

camera reboots in 90 seconds...not sure where you get 5 minutes. and the newest firmware yet to be released, or so i have heard, cuts that time to a lower number.

and high stress comes from poor planning, no it is not your fault, but read my post to new owners...it is imperative that anyone whom rents the camera or owns the camera to understand the nature of the camera at this time and along the way, nobody promised anyone a finished product, in fact jim clearly states that it may never be finished.

good luck and use those carriage returns it helps break down individual ideas into digestible chunks.
 
We use stops because that adequate reflects the right proportion of how the light sensing dynamic range feels to our eyes. To say each stop is double the light is right, but our perception to light is non linear, so talking in stops with 14 just being one more stop than 13 gives the right feel to the numbers. To say that extra stop is DOUBLE (and it must be in caps) the light, is misleading as every stop is double. One stop is not different to the rest.

To say: "RED currently has no way to skew the gamma curve to have more lattitude in the highlights than in the shadows" is not just false and misleading, it shows that you really don't understand dynamic range, image sensors or how the RED is working. And of course, because the sensor is linear, unlike with film, under or over exposure doesn't effect contrast as you're not putting different parts of the tonal range on a curve.

As pointed out, the film curves don't show that much dynamic range. You'd think that if that stock was oh-so-superior, it's creators would show greater range in the curves that they use to advertise it with.....

Graeme
 
Sometimes frustration and pointing fingers will help learning and I think I am.
 
Graeme Nattress said:
To say: "RED currently has no way to skew the gamma curve to have more lattitude in the highlights than in the shadows" is not just false and misleading, it shows that you really don't understand dynamic range, image sensors or how the RED is working.
Click to expand...

Very true, Graeme!

Graeme Nattress said:
As pointed out, the film curves don't show that much dynamic range. You'd think that if that stock was oh-so-superior, it's creators would show greater range in the curves that they use to advertise it with.....
Click to expand...

I think that the film stock folks might be more conservative in their graphs?

The only way to tell is for someone to do a side-by-side with film and Red.

Problem is, once you get hold a Red for a day you don't exactly feel like sitting around and doing test charts! What a waste of such a beautiful thing! And especially if you had to get a film camera, pay for scanning, etc. That's quite a packet of cash that I'd much rather spend shooting something ;)

But if someone does do it, let's hope they do it properly. Eg not just monochrome stouffer wedges and resolution charts... but put colored gels on it, test the limits of color resolution, etc. similar to what they did in this test of still cameras:
http://www.ddisoftware.com/sd14-5d/

If I have to consider shooting bluescreen instead of greenscreen, I'd want to know how much resolution would be lost, etc.

I think the Red will win in the price/performance category hands-down, and will very likely be superior to film for shooting greenscreen plates, etc, and not far behind film for general shooting at standard speed. For something like high-speed photography of blondes against bluescreen (where Red would be 2K windowed and not have the Bayer green advantage), _maybe_ film or a F23 would make more sense, but those have their blue channel issues too.

Anyone want to do a decent test? Maybe in a few months when Red has released their new lower-noise circuits and they've had a bit of breathing-room to tweak their algorithms?

Then there would be no need for all of these weird discussions. Although I think everyone secretly enjoys them ;)

Bruce Allen
www.boacinema.com
 
There's not perfect camera or technology. All we can do, is use our best thinking to come up with the best compromises and balance of features.

Monochrome tests are good as they can show you what's going on, in general. I've used colour tests too, to get more of a picture of what's happening. Resolution isn't all that makes up the picture - you could have high resolution of nasty looking pixels and not be happy. I also like zone plates to look for aliasing issues and for a much better way to measure resolution than a standard resolution chart too.

But really, what it comes down to, is the "whole" picture. That means looking at the whole thing, through a full pipeline and output as a finished product. The geeky techy stuff can help you get the best out of the finished product, but the proof of the pudding is in the eating. And to me, to have 4 to 5 mins of lovely looking 4k on a CF card I can get quickly onto my computer and begin working with is as important as the image itself.

Graeme
 
Floris Liesker said:
I don't think you need to teach Graeme what a stop exactly is.
Anyway, he didn't say 14 stops, he said 13.3 of which 8.3 is linear.
A lot of stock would have considerably less DR than that though.

Kodak themselves show only 11 stops of DR in their graph of 50D 5201.
http://www.kodak.com/global/images/en/motion/products/negative/5201_char2.gif
Click to expand...
Kodak shows -7 stops in underexposure and plus 6 in overexposure. 7+6 = 13 stops total but you will notice that in the overexposure section, the graph simply stops and is cut off, not showing the actual gradual tapering of how the film actually continues to react. It has a curved edge just like the shadows but they don't publish it because they don't want to guarantee it as filtration and lenses affect this out point, as well as processing. Overexposure and underdevelopment (zone system techniques) can actually alter this entire curve gaining you several stops, something Ansel Adams discovered and used to his great advantage.
 
Kodak already disclaimer the graph so to further remove a portion is very strange. I think you misreprent the zone system though. The printing process cannot increase the DR over that which was recorded. You can print it so as to help the viewer see the recorded DR at it's best but not actually improve it. On the chart, although graph extends far to the left, the actual point of no more shadow detail is where it goes flat, not beyond.
 
linear vs. gamma

linear vs. gamma

To say: "RED currently has no way to skew the gamma curve to have more lattitude in the highlights than in the shadows" is not just false and misleading, it shows that you really don't understand dynamic range, image sensors or how the RED is working. And of course, because the sensor is linear, unlike with film, under or over exposure doesn't effect contrast as you're not putting different parts of the tonal range on a curve.
Graeme[/QUOTE]
I understand exactly that the sensor is linear and in fact does not have the gamma curve configuration of the f900 or varicam. What I'm suggesting is that this is perhaps a flaw since the camera is definitely showing it's midpoint to be closer to the highlight end of the scale. That means the median 0 point (properly exposed skin tones) falls closer to the highlights which means there are more stops latitude in the shadows than in the highlights. Of course you can just change the exposure and effectively shift the entire linear scale to a better zone where highlights will be best recorded but then skin tones will be darker. I've been through this extensively with the Varicam, which basically when you switch the Panasonic monitor to "Film Gamma" it essentially corrects not what goes down to tape but just what appears on the monitor so that the director/prod. can feel safe. Of course I know this is in fact the best way to work the camera- have you ever seen the RAW files out of the Panavision Genesis? Flat and with skin tones 2/3 to a stop down. That's exactly where you want to be with a RAW type configuration.
Problem is that on set, directors and producers get nervous. A friend of mine who is shooting ABC's "Brothers and Sisters" originally was going to use the Genesis, but because the ideal raw files (in fact you can't mess with the genesis either) looked the way they did in uncorrected post, and because the producers just didn't understand what they were looking at, and because they couldn't slow production down to a pace that would allow an on-set DIT to tweek every set up, they canned the Genesis and went for 35mm. I guess I got used to the versitality of the Varicam and being able to call where that 0 mid point was going to fall. I could give it 6 stops in the highlights and 4 1/2 to the shadows for day exterior and then for night ext. split the scale exactly opposite. Having presets that can load in a camera quickly is usually a lot faster than skewing the directors/clients/editors monitor between locations. But hey, maybe my 20 years of experience, the half dozen HD features I've shot and my degree in cinematography from Brooks Institute really doesn't put me on the map with you guys as someone who may have a thing or two to say of interest. Everyone seems so involved in defending the system- STOP - It's not necessary, I'm already sold and know that the camera is going to change the whole face of film making. If you guys read the title of this post it is; "R&D Priorities" and that's because I'd like to see the camera shine on set and sure - take over the world- as soon as possible. But there are some serious shortcomings and I don't see a problem with discussing what needs to be fixed first, and what second and so forth for the system to gain the reputation that it will deserve.
 
Kodak graph

Kodak graph

Graeme Nattress said:
I think you misreprent the zone system though. The printing process cannot increase the DR over that which was recorded. You can print it so as to help the viewer see the recorded DR at it's best but not actually improve it. On the chart, although graph extends far to the left, the actual point of no more shadow detail is where it goes flat, not beyond.
Click to expand...
Exposing the negative and processing it has nothing to do with printing. Of course you cannot improve the DR in printing, but you certainly can improve the DR by overexposing the negative and then underdevelopinng it in a cooler temp chemistry and for a longer time. You are essentially lowering the ASA rating of the film but taking advantage of the higher DR built into the emulsion. The thing is Kodak of course pushes the ASA rating to the limit usually of what the film can possibly do while retaining acceptable DR range. This is less important in a low speed stock. And the graph you looked at yes- on the left side it is the shadows and it goes flat, but I was commenting on the right side where you see the graphed response simply ends. There is more there in the highlights, a lot more in fact.
 
shot 8 toyota spots w my camera at brooks three weeks ago and everyone loved it and praised it....

the senior gaffer; an older asian gentlemen, was very impressed w the range, look, workflow et. al,

dont forget sound and story...picture is not everything in story telling.

onset at brooks...with gaffer in view...
 
Technically the Panavision Genesis does not shoot & record "RAW" files (RAW referring to a Bayer-filtered image that has not gone color-processing to become RGB), it converts the signal from its RGB striped sensor into RGB and records it as 10-bit 4:4:4 1080P HD video, with a LOG gamma curve (Panalog).

And Panavision has a LUT box so that the LOG image is gamma corrected for viewing on a monitor in REC 709 color space.
 
i love it when you talk dirty like that...
 
Hi there,

Constructive criticism is cool but why do you have to be so angry and pissed about it?

"This needs to be solved immediately"
The camera is Goddamn loud.
FORGET ABOUT DEVELOPING A SET OF LENSES FOR NOW.
You cannot use this camera to do stunts, fights, car action, pyro or anything else that requires slo mo. Did I say really disappointing?
11.3 stops latitude. Again a real disappointment

For someone who is so unhappy about this camera, I would recommend that you simply use another camera. I recall all of the skepticism surrounding the camera before it came out. It didn't surprise me..however, I never expected such resentment towards a 17,500 4k camera. What other 4k digital cameras are there? and how much do they cost? No one is forcing you to use it. I'm sorry but your rants complete ignore the fact that there is really NO OTHER CAMERA comparable in price/features/performance. You might as well complain that RED should have made an 8k camera. The work is not done yet..and RED will continually improve and enable more and more features. Everyone at RED is working around the clock. Demanding everything to be enabled right now is just unrealistic.
 
I agree with everybody that Morpheus tone makes it hard for us to exactly understand his position but the point might just be that this could happen to ANYBODY who picks up the camera without going through these boards everyday and without direct access to technical support from someone from RED. How brilliantly the camera was designed and how great value for money it is, at the moment it still a technical challenge to work with and we have to give Morpheus some credit for pointing that out. I am trying to imagine myself on a set with fifty people all waiting for me to load different firmware that might or might not solve a problem :waaa:

Also the fact that not all cameras seem to respond to new firmware in the same way worries me a bit. How are new camera owners supposed to find out whether there problem is unique?

I hate the waiting game but it might be a good thing that RED is not shipping new cameras until certain things are worked out.

Graeme: I always understood that you could upload custom curves for display to the camera? Wouldn't that enable you to push the underexposed skin tones up so they look good again? Am I totally missing the point...?
 
You must log in or register to reply here.
Back
Top