RED future...

[Radoslav Karapetkov]

Export AVIs from RedCine.

 

[HD Hildebrand]

I second Axel's request. I want to do stop-motion using a custom push to make switch, and also want to get it on a local mo-control rig ASAP I take delivery to dial it in for future jobs.

Is there a way to "external" tigger the Red yet for single frame?

 

[David Mullen ASC]

Oh sorry David, I apologise, I missed your last statement! Lost what I wrote and then had to write it again. But I think what I'm saying makes even more sense, because to do a shot with equivalent FOV to a 35mm you would need something like a 21mm. And then you DEFINATELY are going to get distortion on a MCU, and you will also need to use Close-Focus lenses. What I have been saying is not about the DOF as much as FOV. And the issue about resolution I discussed and the requirement to process the sequences at 4K is even more necessary.

I think you're missing the point -- because the RED sensor is slightly smaller than 4-perf 35mm anamorphic, you are cropping the sides of the lens image, so going to a shorter focal length to compensate doesn't force you to move the camera closer, you've already matched the longer lens just by switching to the shorter focal length and then cropping it to the same view, plus by using less of the sides of the lens, the extra barrel distortion of the shorter lens will be minimized. So there would be no extra distortion in the MCU since you didn't move the camera closer (plus how often do you shoot close-ups on a 35mm anamorphic lens???

That's like shooting your close-ups on an 18mm lens in Super-35. In fact, the problem with close-focusing anamorphic lenses for wide-angle close-ups is the reason most directors prefer Super-35 for that style of shooting.)

I still think you are overthinking this problem. Most anamorphic photography is in the 40mm to 100mm range, so to match that, you'd have to work in the 28mm to 75mm range. Most anamorphic close-ups are shot in the 75mm to 100mm range, so to match that, you'd be using 50mm to 75mm range. And there are close-focus 50mm anamorphic lenses.

Practically speaking, you could compensate for the size of the RED sensor 95% of the time, so this is not a tragic situation. You are under some illusion that it really makes a huge visible difference to use a 35mm instead of a 50mm when they are both giving you the same field of view. The difference will be in depth of field (about 1.5 stops more depth of field, which can be compensated for usually) and how much edge distortion there is (which is compensated somewhat by the fact that you are using less of the lens coverage).

Otherwise, in terms of perspective, "feel", you've made the 35mm anamorphic look like the 50mm anamorphic by putting it onto the RED camera -- just like a 25mm looks like a 50mm when put on a Super-16 camera, just that it has the depth of field of a 25mm.

I could probably shoot a test with some anamorphic lenses on a RED camera and shoot wide shots, medium, close-ups, wide-angle close-ups, etc. just like any other anamorphic feature and you'd look at the footage and know it was an anamorphic-shot movie. Which is all that ultimately matters. Yes, there may be a rare occasion where you can't get as super wide-angle as you'd like, or the super-wide angles are too distorted, but that's about it. Truth is that many people avoid the 35mm anamorphic for this reason, it has a lot of barrel distortion -- in fact, it's not until you get to 50mm and above that it starts to go away.

On the other hand, people use anamorphic zooms to get around this barrel distortion problem and then the footage looks less like it was shot in anamorphic. So for all you know, you may prefer the look of the barrel-distorted anamorphic primes over the optically correct image in the anamorphic zoom.

And yes, a 4x3 sensor that is 18.59mm tall would be one solution, the other would be to build a few more wide-angle anamorphic lenses optimized for the shorter sensor.

I've exhausted this topic mathematically-speaking so there is nothing left to say other than to shoot your own tests and decide if it is unworkable or not.

 

[Brent J. Craig]

...Thankfully that is not happening where I am.

No? Is there a Panavision or Clairmont shop where you are?

 

[thornben]

No? Is there a Panavision or Clairmont shop where you are?

Not Clairmont, but yeah.


I hope that it works out.

 

[Jay A. Kelley]

Export AVIs from RedCine.

I 2nd this.

Jay

 

[thornben]

And yes, a 4x3 sensor that is 18.59mm tall would be one solution, the other would be to build a few more wide-angle anamorphic lenses optimized for the shorter sensor.

I've exhausted this topic mathematically-speaking so there is nothing left to say other than to shoot your own tests and decide if it is unworkable or not.

Sorry David,

I was never asking you if it was unworkable or not, but thank you for explaining it. I said it was crazy, and I am right, but not for the reasons we went down. As far as the camera is concerned, there is no doubt you know more than me. I do get it now though, I was incorrect. I see what you mean now. Sorry for being boring. Yes you are right there is no extra distortion, and yes you are right you gain a stop or more, which is actually a good side benefit. That was pretty stupid and do I apologize for wasting so much time. On the plus side, we all got some interesting information out of it. However, fundamentally I feel you will lose the largness of the way the lens sees, but practically speaking you are absolutely right, it's not really a problem per se.

Yeah, I do overthink stuff. It is an "issue" I am aware of! Haha! I'm sorry you got drawn into that spiral, I'm sure you have better things to do.
David, thanks very much for a stimulating and informative dialogue. Thank you for correcting me, too.

But on the plus side, thinking it through did lead me to the real problem.

Namely that if one wants to shoot anamorphic on the RED, as I am sure many directors like myself do want to, and it is a reason that RED should be concerned with as it is a reason people want to choose the camera for a job.
(And that point of view wasn't being expressed in these forums so that's why I posted my comments in the first place.)

Well, then:

Building the larger sensor is actually the only good solution once you take into account the back-end issues I was describing in my previous post.

To better explain:

You shoot 2X Anamorphics inside the Super-35 area, right? I was wrong, it will work 95% of the time without any noticeable difference at the camera stage.

BUT:

Every shot will probably need to be processed as a 4K image sequence. Why?

In unsqueezing the image, and cropping it, you lose a lot of information. If you do 2K image sequences, the resultant output will still be 2K after the processing, but it will be noticeably softer than desired.

So you have to do everything at 4K, right, and then probably downscale it to 2K wtih sharpening to get the crispness. This isn't a problem for the DP now, it is a time and budgeting problem, which I do know about.

4K image sequences take more than twice as long to process, which means it costs more than twice as much (either because the post house adds more machines to process it and they charge you for it, or they take longer and they charge you for it).

Even if you do all this in-house, you still have to charge the client more as it takes longer, or absorb the extra costs – and anyway they are also waiting longer for the deliverables which isn't what they like and may be a reason you don't use Anamorphic. Which is a good reason from a production perspective, but a bad reason from a director's perspective.

And then on top of this if you want to crop the image area to reduce bandwidth of the Quicktimes you will be working with, then that takes longer as well. I personally wouldn't because you might as well use the extra area for reframing considerations, but if you don't crop it there then you have to render every single shot in your timeline even when just doing an edit - and then you have more time and money again.

Which brings you back to the usual problem for Anamorphic - it costs more than shooting Super-35.

But, if you have an Academy-height sensor, this problem doesn't exist. You can do 2K's, and you can even use the Proxy Files straight into Final Cut. No extra processing than needed, no rendering of the timeline. Money. That is always the most important deciding factor in any project. Once the problem enters the post side we are talking mega bucks. And that is a deciding factor. Also, without forgetting, the image quality is going to improve as well as you are using the lens as it was designed and getting a cleaner image, and a larger image.

Also, they should have an option to window that sensor to 1.78:1/3-perf size, so people don't have any unecessary data to handle.

Building wider lenses that aren't as good quality is not the answer. I'm sorry, I'm not trying to be rude, please don't take it that way.

Anyway, at some point Panavision or somebody else is going to either have to build 1.33X Anamorphic lenses or, much more intelligently: bring out a camera that uses an Academy-sized sensor. I mean if that isn't what they are doing for the Genesis v2.0 then they are definately crazy - Anamorphic is what they do best.

I think these are all very good suggestions and reasons for Jannard and the RED TEAM to consider, and hopefully do.

Thanks again!

-Ben

irate:

 

[thornben]

...and before anyone says "Anamorphic lenses are generally softer than Spherical Lenses anyway" you will also admit that the reason an anamorphic picture seems sharper is because it has more resolution in the first place. If you are taking this element away I think there is a problem.

 

[thornben]

Actually you wouldn't be outputting 2K sequences in the usual size, you would be outputting sequences with 2K width in unsqueezed form, or possibly a little less as Super-35 width is greater than Academy/Standard 35 (which is what the lenses are designed for) width. But the height would be much greater than standard 2K and the unsqueezed width would also be much greater than 2K. That would also mean for an HD picture you would get a much cleaner and sharper image, too.

You would also have the standard DOF/vignetting/lens barreling of anamorphic, which some people might not like, but the industry is used to those issues. And personally, yes, I like those things. They look interesting and can be exploited for a look.

 

[David Mullen ASC]

I'd point out that the Phantom HD and the Arri D20 both have square-ish sensors.

Yes, in an ideal world, all of these new digital cameras would all have 4K / 2K 4x3 sensors, the height of 4-perf 35mm, to accomodate the few people shooting in anamorphic (and the fewer who want a 4x3 image). They always have the option of only using a 16x9 portion of the sensor most of the time (like when the Arri D20 is used in HD mode instead of Data mode.)

Trouble is that the industry as a whole is moving towards 16x9 negative formats (3-perf 35mm, Super-16) and digital sensors. That's just the trend because most people want 16x9 images for HDTV broadcast, and you can crop for 1.85 (barely) and 2.40 (more.)

So I simply don't see much impetus for camera makers to address the smaller crowd that needs 4x3 sensors.

This is why I think someone should be addressing the issue of making 1.34X anamorphic lenses.

Truth is that most people are going to use spherical lenses and crop to get 2.40, just as they do in 35mm -- Super-35 is a lot more popular than anamorphic. But there is a certain hardcore crowd of anamorphic lovers that aren't going away.

In fact, today I saw "Sweeney Todd" (Super 1.85) and the first three trailers were all movies shot in anamorphic lenses -- two romantic comedies and one horror film. So there is a die-hard contingent of anamorphic shooters that I think these camera and lens makers need to address.

---

"2K" scans of 35mm anamorphic are 1828 x 1556; "4K" scans are 3656 x 3112. Of course, they may actually scan the soundtrack area too (so 2048 across for 2K or 4096 across for 4K) but they won't use that area in the final film recording.

 

[vinodkumar]

Heard about scarlet i know red will make best of the best but i have only one request jimm plz dont make it too small which shakes in hand...

once again thanks for making a camera like redone i would like to buy 2 next year .....

vinodkumar@gmail.com
india

 

[Häakon]

Yes, in an ideal world, all of these new digital cameras would all have 4K / 2K 4x3 sensors, the height of 4-perf 35mm, to accomodate the few people shooting in anamorphic (and the fewer who want a 4x3 image).

Which is bigger, a square or a rectangle? This depends of course on the dimensions of each since it is possible to fit a square inside a rectangle or vice-versa if the proportions are correct.

The bottom line is that one is not "better" than the other, it just depends on your ratio of choice and maximizing resolution from your gear. A 4x3 sensor is a complete waste of resolution if you are shooting non-anamorphic widescreen movies. Since my suspicion is that this is the way the large majority of RED shooters will be using the camera, I think it would be a very poor choice to incorporate a 4:3 sensor in a future release. Absolutely it would be beneficial for anamorphic or 4:3 shooting, but I see those applications being a small slice of the overall RED pie. Just my two cents.

 

[Casey Green]

Which is bigger, a square or a rectangle? This depends of course on the dimensions of each since it is possible to fit a square inside a rectangle or vice-versa if the proportions are correct.

The bottom line is that one is not "better" than the other, it just depends on your ratio of choice and maximizing resolution from your gear. A 4x3 sensor is a complete waste of resolution if you are shooting non-anamorphic widescreen movies. Since my suspicion is that this is the way the large majority of RED shooters will be using the camera, I think it would be a very poor choice to incorporate a 4:3 sensor in a future release. Absolutely it would be beneficial for anamorphic or 4:3 shooting, but I see those applications being a small slice of the overall RED pie. Just my two cents.

What about the look-around area. By having the 4:3 sensor, don't we also get additional benefits for the Camera operator by having the "surround view" area around the frame? And since RED only uses an Electronic View Finder, this may be important, no? I suppose an extra large 16:9 sensor could be made in a similar way.

 

[Axel Mertes]

I'd point out that the Phantom HD and the Arri D20 both have square-ish sensors.

[..]

David,

the Phantom has a 4096x2440y pixels (~1.6:1), the ARRI is 3072x2048y (so 3:2). Not exactly sqarish...

There are cameras with squarish sensors, but thast typically industrial stuff.

We just shot finished a shooting in Berlin in some close bunkers, all is aimed for 2.4:1 super cinemascope results (we just took the full image for later pan&scan option at full 4K).

See some footage posted here:

http://www.reduser.net/forum/showthread.php?p=139223&posted=1#post139223

We already printed on film, currently waiting to see the results returned from the lab.

I agree that having anamorphic lense would make sense here to gain the last little piece of image quality. On the other hand you'll have hard times to get the full 4K to the release prints anyway, so have the 4Kx2Ky being converted to an anamorphing 2Kx1.7Ky (roughly ) will still be a very valid workflow for many productions.

Regards,
Axel

 

[conrad gaunt]

... I need the RAW port. I honestly need it for what I want to do... ... The RED RAID is also in the wings for use with the RAW port so I have to factor that in as well.[ ]

I`m not sure you do need the RAW port?, the Red Raid won`t be able to capture data from the RAW port, only RedCode RAW (compressed) data. The RAW port requires fridge sized capture gear, somebody correct me if I`m way out and they come in microwave oven sizes etc, and will probably cost more than the rest of the camera and accessories. The good news, Redcode RAW (the compressed stuff), should suffice for most high end SFX etc, so probably yourself too.

, Edit: RAW capture would be nice though, but cost of raw port will get you a nice lens, probably money better spent.

 

[Häakon]

What about the look-around area. By having the 4:3 sensor, don't we also get additional benefits for the Camera operator by having the "surround view" area around the frame? And since RED only uses an Electronic View Finder, this may be important, no? I suppose an extra large 16:9 sensor could be made in a similar way.

If you put a 16:9 recording area inside a 4:3 sensor, you'd have look "around" on the top and bottom, but then one might argue that you should be able to record that entire area if your application is 4:3 or anamorphic and you want the best res possible. Of course, if you make it 4:3, you might just as well make it 16:9 so you have "look around" on the sides... and then, well, you get the picture. It's just best to make the sensor in the shape and size of the format that you'll be doing the most recording in, which for this crowd, I believe is a widescreen ratio. Hopefully that makes sense...

 

[Casey Green]

Makes sense, Häakon.

 

[Jarred Land]

you nailed it Haakon.

 

[Zach Hilton]

I`m not sure you do need the RAW port?, the Red Raid won`t be able to capture data from the RAW port, only RedCode RAW (compressed) data. The RAW port requires fridge sized capture gear, somebody correct me if I`m way out and they come in microwave oven sizes etc, and will probably cost more than the rest of the camera and accessories. The good news, Redcode RAW (the compressed stuff), should suffice for most high end SFX etc, so probably yourself too.

, Edit: RAW capture would be nice though, but cost of raw port will get you a nice lens, probably money better spent.

Wasn't the Raw port canceled for lack of interest? Or something better which has yet been announced?

 

[Zack Birlew]

Yup, sure was and I'm waiting to see what RED will do instead of the RAW Port, especially since it was stated that the alternative will be MUCH cheaper. Also, back when I had originally posted about needing the RAW Port, the RED RAID was, as far as I understood it, going to be a seperate product from the RED Drive. When people started calling the RED Drive the RED RAID, I have no idea, but I always distinguished them as two seperate products.

The need for high speed frame rates is still there. 60fps at 4K is important but probably not as much as 120fps at 2K was going to be and 4.5K was a biggie too. I didn't care about the $6,500 price added on for the RAW Port, even though that meant giving up a lense or two and several accessories. With the RED, the best image quality is what I'm after and that includes both resolution and frame rate speed. The last thing I want is to be limited in what I can do. Now, I'm not saying I need a 1000fps Phantom camera or anything but having the ability to do really slow motion stuff at 2K+ resolutions is very important in several things I'm planning on doing with my RED.

Don't read this all the wrong way, I support the RED team heavily and I'm very grateful for what they've done with the making of this camera. I'm simply stating my needs and what I want out of the camera, which RED is still providing as I've stated with a simpler and cheaper solution, whatever that may be. In any case, I'm sure NAB will be very exciting this year! :bleh: