RPAs...

[brian hanson]

wow, red just doesn't sleep!

 

[Jeff Kilgroe]

You wouldn't need the full FF35 coverage, even for the height of the frame. At first it was difficult to say if these should be 2X or 1.5X anamorphics. The more I think about it, the more I'm inclined to agree with David's post above about 1.5X being a good compromise. This makes the most sense for serving the majority of RED's customers. With Epic-X/S35 being a 2:1 sensor and it looks like the maximum recording area will be about 27.65mm x 13.83mm. A 1.5X anamorphic used to shoot 2.40:1 would yield a native pixel resolution of 4096x2560. So anamorphically, that is still higher resolution than native 4K 16:9 on the RED One. Using a 2X anamorphic lens would give us 3072x2560 natively, so a loss of 1K in horizontal resolution, with probably the only benefit being slightly "enhanced" optical "features", maybe but that is also dependent on the optical design.

If you were to shoot on an FF35 sensor, it becomes a question of lens coverage. To pull off the 2.40:1 example above on Epic-X, the lens would need at least a 26.1mm image circle. But what if the RPA's were to have a 30mm image circle? Then you could get away with a recording area of 25.44mm x 15.9mm. Making the ASSumption that Monstro FF35 will still have the slightly larger photosites (6um), that gives us 4240 x 2650 native pixels.

So almost identical, really, even with the larger image circle. It will be easier to maintain a larger image circle with less anamorphic compression (typically). And 2X anamorphics will provide for less usable resolution on Mysterium and Mysterium-X sensors, while only providing a resolution benefit if it can cover the full height, or very close, of the FF35 sensor.

Anyway, after that little thought exercise, I think a 1.5X may just be the ticket here and keep the coverage at a 30mm+ circle like the RPPs.

If RED does go for 2X anamorphics and they do cover the full height of the FF35 image plane (24mm), that would yield an image size of 4800x4000 native pixels, using the same 6um ASSumption above. That would require a 37.5mm image circle, possible I guess. But the 2X doesn't make a whole lot of sense for resolution on the Mysterium-X sensor.

Edit > I pulled out a couple numbers for un-squeezed image sizes as it was confusing and gave the wrong impression about resolution. Thank you, Mr. Mullen.

 

[David Mullen ASC]

With Epic-X/S35 being a 2:1 sensor and it looks like the maximum recording area will be about 27.65mm x 13.83mm. A 1.5X anamorphic used to shoot 2.40:1 would yield a native pixel resolution of 4096x2560 and an un-squeezed resolution of 6144x2560. So anamorphically, that is still higher resolution than native 4K 16:9 on the RED One.

I'm never comfortable with calculations that increase the pixel size of the frame by unsqueezing an anamorphic image because that gives the false impression of increased resolution. The area used on a sensor when using a 2X anamorphic lens will be a 1.20 : 1 shape for a final 2.40 image. There is no increase in true resolution if you decide to double/uprez the horizontal size of the file by unsqueezing it, so I'm not sure of the point of saying "unsqueezed resolution of 6144x2560". That's simply a larger size file you can upscale the horizontal dimensions to, not an increase in resolution.

Besides, for a film-out, you'd leave the file 1.20 : 1 with a 2X optical squeeze, record that to 35mm. And for DCI projection, you'd have to unsqueeze it and convert it to one of the standard DCI file sizes, 4K or 2K.

 

[Stephen Williams]

As a student in boston, I really appreciate this -- there simply arent any affordable anamorphic solutions up here!
arty: ANAMORPHIC PARTY! BYOG (BRING YOUR OWN GLASS)

I have seen LOMO's go on Ebay for $1000 per lens with PL mount, I suspect Red lenses will be more.
I am pleased RED is doing this, for 2 years I have been told cropping is good enough.

 

[Ryan E. Walters]

Simply amazing- I swear nobody sleeps at Red ...

I'm looking forward to hearing more about these lenses- what the cover, the squeeze factor, and what kind of flares & bokeh they produce. Thanks for continues to provide option Red!

 

[Ryan S]

Now I'm looking forward to the first stylized, backlit renders from Jarred!

 

[Jeff Kilgroe]

I'm never comfortable with calculations that increase the pixel size of the frame by unsqueezing an anamorphic image because that gives the false impression of increased resolution. The area used on a sensor when using a 2X anamorphic lens will be a 1.20 : 1 shape for a final 2.40 image. There is no increase in true resolution if you decide to double/uprez the horizontal size of the file by unsqueezing it, so I'm not sure of the point of saying "unsqueezed resolution of 6144x2560". That's simply a larger size file you can upscale the horizontal dimensions to, not an increase in resolution.

That's correct, I wasn't trying to imply higher resolution. Although I see how it can be interpreted that way. And in most cases for un-squeezing the image, you wouldn't stretch the image horizontally, but rather compress it vertically to take advantage of some oversampling. For example, the 4096x2560 from a 1.5X anamorphic on 5K-MX, would be best re-sampled to 4K 2.40:1 for DCI or if you needed a 4K 2.40:1 for DCI or a DI for FX work.

 

[Miguel "Macgregor" De Olaso]

I know a lot of people believe this, but if you look at old 5-perf 65mm Ultra Panavision movies shot with 1.25X anamorphic lenses, you can see squeezed backgrounds and horizontal flares. Same goes for old 8-perf 35mm Technirama movies shot with 1.5X anamorphic lenses. I've posted frames from "The Battle of the Bulge" and "El Cid" that show this.

It's just that the distortion is not quite as extreme as with 2X anamorphics. On the other hand, the distortions increase as you open up the lens aperture, so I wouldn't be surprised if a 1.33X anamorphic lens shot at T/2.0, for example, had similar distortions to a 2X anamorphic lens shot at T/2.8. Though the shape of the out-of-focus ovals may never get as skinny on the 1.3X lens.

Anyway, it would be easy enough to test a 1.33X Hawk lens against a 2X Hawk lens to compare the level of anamorphic distortions.

Yes, ask Hawk/vantage for their demo tests comparing the 1.3x and the 2x. They shot the same shots with their different front attachments so you can see the bokeh difference (in 35mm, in d21 and in Red).

Not that the x1.3 are bad at all, but they almost look spherical.

Edit: i tried attaching some screenshots, i hope it works.

 

[Miguel "Macgregor" De Olaso]

testng

 

[joshua csehak]

I'm never comfortable with calculations that increase the pixel size of the frame by unsqueezing an anamorphic image because that gives the false impression of increased resolution. The area used on a sensor when using a 2X anamorphic lens will be a 1.20 : 1 shape for a final 2.40 image. There is no increase in true resolution if you decide to double/uprez the horizontal size of the file by unsqueezing it, so I'm not sure of the point of saying "unsqueezed resolution of 6144x2560". That's simply a larger size file you can upscale the horizontal dimensions to, not an increase in resolution.

But wouldn't perceived resolution increase if the sensor resolution was higher than that of the lens optics?

I don't know the technical details of the Red sensor and lenses, but if I take a shot in 4k and scale it down to 2k and then rescale it to 4k again and compare it to the original 4k image, the original is a little sharper but not twice as sharp. If that's because of the lens, and the anamorphic lens is capable of squeezing more information onto the sensor, then you would get more horizontal resolution, right? But if it was because of the sensor, then it wouldn't make any difference.

But don't anamorphics work the opposite way? The horizontal part is the same, but the vertical part is stretched to get more vertical resolution, right? So wouldn't these only be useful on the FF35 sensor? The Mysterium is at (or close to) the final delivery aspect ratio.

 

[James T Mather]

Not that the x1.3 are bad at all, but they almost look spherical.

I agree - going 1.3 is or even 1.5 seems IMVHO pointless optically.

For the pixel junkies the chip resolution is a small reason to shoot anamorphic as the lenses (owing to bending light paths etc) have visibly lesser actual res than spherical anyway - you hit an optical wall. Maximum res pixelwise is not the sole purpose of the endeavor - the reasons for shooting anamorphic has more to do with "look" and these are mostly by-products of the optical squeeze/desqueeze process.

This might not bother the pixelpushers (who seem to measure quality solely in terms of array dimensions and neg size) but there are other aspects to the process. I fail to understand why using the "whole" neg is so vitally important unless it is some rainman like inability to grasp that by cropping off a bit of horizontal info you end up with a nice pic - who cares so long as the images look good and hold up and at 5k they will. It seems a kind of Aspergers mechano lust.

Also a mite contradictory to say give us anamorphic (a format developed to combat a square neg) and then argue why doesn't it work on our widescreen camera.

Also with all the demand for FF anamorphic I wonder how many have used anamorphic - The DOF is crazy low - you need to work at T4 to get any kind of chance to get the lens sharp to begin with and then it is still thin - at FF you will have sharp eyes/soft nose - yes some very pretty tilt/shift looking garage band pop videos will come of it but for budget scheduled narrative it seems bonkers.

 

[David Mullen ASC]

The other argument for sticking to 2X anamorphics is simply that you may need to augment your Red anamorphic lens package with some other lenses (Hawk, Technovision, JDC, Lomo, etc.).

Even on the 5K 30x15mm sensor, that area used for 2.40 with a 2X anamorphic lens will be 18x15mm, which is smaller than the 22x18.5mm area of 4-perf 35mm anamorphic, so there shouldn't be any less depth of field.

Yes, on the Monstro sensor, it would be 29x24mm, which is larger, but only by 1.3X compared to 22x18.5mm.

Of course, if the reason people like anamorphic are the artifacts and the super shallow focus at wide apertures, then I guess they'll be even happier on the Monstro, though the 1st AC might not be so happy...

 

[nick allsop]

Im hoping for a look similar to the JDC or Lomo's but with the build quality of the RPP. The look of the Hawks is too perfect, they don't flare enough.

 

[Harry Clark]

Wow. This is exciting. Count me in for a set.
Cheers,
Harry

 

[Stephen Williams]

The look of the Hawks is too perfect, they don't flare enough.

I understand there are 2 versions, 1 set that flars & one that don't.

 

[Michael Lofquist]

Brilliant - another Red master piece. Love the way you guys disrupt the establishment

 

[PaulClements]

Good news, but I doubt they'll be 2x

Good news, but I doubt they'll be 2x

Interesting lenses. Coulda sworn Jim said they weren't going to be making any, glad he changed his mind. I'm kinda hoping the sensor sizes of the Epics are changed to accomodate.

My guess is they will be 1.5x or 1.66x at most... I say this because a 35mm would have a similar fov to a spherical 23mm/21mm and the "possible" 24mm would have a similar fov to a 16/14.5mm spherical lens. If the lenses were 2x that would mean the "possible" 24mm would essentially be a 12mm spherical lens and since Red have made no announcement for such a wide standard prime I'd be suprised if they went to such efforts for the anamorphic crowd.

Paul

 

[Shawn Nelson]

So I guess the million dollar question is what the squeeze factor will be. 1.3x, 1.5x, 1.6x 2x? Interesting that Jim didnt say, could that question still be on the table?

 

[Todd M.]

Tonight we are announcing the development of the RPAs (RED Pro Anamorphics). They will match color, optical quality and mechanical performance characteristics with their little brothers... the RPPs. The weight of the RPAs is expected to be within 1/2 pound of the RPPs.

Initial set will consist of:

35mm T2.4
50mm T2.4
85mm T2.4
100mm T2.4

A 25mm T2.4 is also being studied.

Delivery of the 4 lens set is expected by year's end... although you know the schedule drill.

Prices will be slightly higher than RPPs.

Jim

Forgive the elementary question but would a DP on here care to give a little background on this focal length selection? I'm not a DP and have not worked with anamorphic footage before in the post process. Specificallly, how does a 35mm anamorphic lens differ from a 35mm prime? I understand that the image is squeezed using an additional piece of glass in the lens so it fits the imaging plane and uses more available resolution, but once "un-squeezed" doesn't the image still retain the characteristics of a similar focal length prime?

Thanks for this.

Best,

T

 

[Charles Angus]

I just hope for nice flares. There seems to be an idea these days amongst shooters and lens manufacturers that flares are a bad thing. In my opinion, flares lend a lot to the "character" of a lens, and something like the Canon 4.5x11 Cine that doesn't flare at all just isn't that interesting to me.

You can always flag off light sources, but you can't make a flare-less lens flare.

But since beautiful flares is one of the reasons to choose anamorphic shooting, I'm sure these will flare easily and beautifully.