Cameras With 4:3 Anamorphic Mode?

[Tito Ferradans]

I wanted to understand why is everyone so obsessed with 4:3.

Is it because there's more height pixels, or because of the aspect ratio itself?

[webrunner5]

An anamorphic lens is a wide-angle lens that optically distorts the 16:9 image to fit into a 4:3 frame before sending it to your camcorder. So yeah you take advantage of all the pixels on 4:3.

[elgabogomez]

If you're going for 2.40:1 you are going to discard a lot of the recorded image in a 16:9 sensor read, if you are doing raw it is a smaller file size if the camera has a 4:3 crop, it's also a potentially minimized rolling shutter. If you're going as wide as your ana adapter allows, the sides of the image are dark anyhow. And finally, yes, it takes away a lot of the guessing in post de-squeeze cause there are presets for anamorphic in adobe ae and pp as well as in resolve. In resolve i do a new project and set the resolution to say 1920 by 800 and the method to see images to fill or stretch (I don't recall which) and I still have to select the files and horizontally stretch them manually a bit more till it looks normal as I'm dealing with 3.56:1 footage.

In the Alexa cameras that do anamorphic mode there is actually more resolution and wider field of view (18mm vertically). That's why I am closely waiting for the gh5 to come out and see the 4:3 Ana crop factor,  not really for the resolution but for the chance to see the 13mm vertical size of the sensor (if true there is no crop).

[Tito Ferradans]

Ok, I get it that you take advantage of all the pixels in 4:3, but if you have no EXTRA pixels is there really an advantage? You get bigger files, but, are they big enough to impact your workflow? I don't mean that for high-end cameras, like RED or Alexa, because these will indeed clog your edit, but for DSLRs and mirrorless cameras the difference is pretty small.

It's just like filming 4k for outputing 1080p. You get a lot of room to play with reframing. So, if anything, shooting 16:9 would give you an edge because you can reframe your shots in post.

Also, cropping the sides is not really hard in any editing software, and framing on set isn't hard either since most cameras have guide frames and you can enable a 4:3 guide in order to see what's "in your shot". For monitoring, it takes 5 minutes to make an accurate cutout of a 2.4:1 frame (I've done it many times for directors who couldn't understand a 3.56:1 frame.

So, if we're not really getting any EXTRA data, just "making less leftovers", is it really a problem?

With a 16:9 readout, you can use any anamorphic. With a 4:3, you can only use 1.78x compression or stronger.

[elgabogomez]

It's not a "big" problem you are right, and the reframing across the 16:9 to 3.56:1 is helpful indeed. But take my personal camera the a6300 as an example for clear benefits of a 4:3 read: it's the worst rolling shutter of any camera out there, if sony made a paid app to do 4:3 video and still keep everything else the same (codec, bitrate, picture profiles, log, zebras, etc,.) I would be the first to buy it, if you have ever played with the clear zoom of a recent sony and compare it with the regular recording, the most obvious change is less rolling shutter, a bit more noise and an almost invisible less resolution, so imagine the 4:3 crop again, being a crop of the 16:9 frame but with faster readout in the same 100mbps and you have the advantages of the clear image zoom without the disadvantages. Even possibly less chance of banding if the bitrate is properly used. 

Now take the ursa minis or the kinefinity terras as examples, it is a lot less data (either raw or prores) in the anamorphic mode. Enough to impact your workflow because you can have 10 minutes more recording on the same 128gb card and trust me, 10 minutes per card is a big deal in a shooting day for a short or long fiction film.

[webrunner5]

Yeah pissing away data on some goofy crop is hard to justify unless you are making a unlimited budget film.

[BrooklynDan]

I think that it's important to look less at the specific aspect ratio of the sensor than at the height of the sensor, especially when it comes to using professional anamorphic lenses. When shooting for a 2.40 aspect ratio, having a taller sensor means using more of the field of view of the lens. And this comes into play when using older lenses that may not have the wide angle options, or have wide angle lenses that are unusually large or oddly shaped. At my job, I have compared Lomo anamorphics on the Alexa (18x24mm sensor) and the Red Dragon (15x30mm sensor). On the Alexa, you can more or less get way with using the 50mm as your widest lens. On the Red, the 50mm is more like a 60mm, and you're gonna be relying more on the 35mm Lomo, which is a big heavy lens with a giant front diameter, and a lot more barrel distortion and softer edges. It's just more practical to have a sensor that matches the full height of the 35mm academy format, regardless of whether it is 4:3 or not. And with how high resolution have gotten lately, there's no need to fear cropping. No use crying over spilled pixels.

[Tito Ferradans]

7 hours ago, BrooklynDan said:

I think that it's important to look less at the specific aspect ratio of the sensor than at the height of the sensor, especially when it comes to using professional anamorphic lenses. When shooting for a 2.40 aspect ratio, having a taller sensor means using more of the field of view of the lens. And this comes into play when using older lenses that may not have the wide angle options, or have wide angle lenses that are unusually large or oddly shaped. At my job, I have compared Lomo anamorphics on the Alexa (18x24mm sensor) and the Red Dragon (15x30mm sensor). On the Alexa, you can more or less get way with using the 50mm as your widest lens. On the Red, the 50mm is more like a 60mm, and you're gonna be relying more on the 35mm Lomo, which is a big heavy lens with a giant front diameter, and a lot more barrel distortion and softer edges. It's just more practical to have a sensor that matches the full height of the 35mm academy format, regardless of whether it is 4:3 or not. And with how high resolution have gotten lately, there's no need to fear cropping. No use crying over spilled pixels.

now THIS is more like the arguments I was looking for. When it comes to the GH4 (and 5 - as well as any Canon shooting ML RAW) I know they have increased sensor height when shooting 4:3, but most people requesting the feature just seem to neglect that.

I didn't mention before, but this thread is a case study for an upcoming episode. I want to make people realize that "just" a 4:3 crop is no big deal. :P

[elgabogomez]

Maybe on a 4:3 crop you feel less guilty of the crop, when reviewing your footage the 3.56:1 composition sometimes looks so great that you don't want to adhere to the cinema "standard", if the camera is doing it for you is as we say in México "eyes that don't see, heart that doesn't feel " (ojos que no ven, corazón que no siente) :D

[Justin Bacle]

On 2/14/2017 at 4:23 PM, Tito Ferradans said:

I wanted to understand why is everyone so obsessed with 4:3.

Is it because there's more height pixels, or because of the aspect ratio itself?

I'd say both, it's great to have more vertical resolution, but having to crop less is more "storage efficient" to me. (with magic lantern being limited by the card write speeds, it's easier to achieve more vertical resolution in 3:2 or 4:3 than in 16:9  )

I actually shoot more in 3:2 with magic lantern as it allows slight movement in the shot in post if wanted, but 16:9 is just too much cropping to me.

[Tito Ferradans]

10 hours ago, elgabogomez said:

is as we say in México "eyes that don't see, heart that doesn't feel " (ojos que no ven, corazón que no siente) :D

We have the same saying in Brazil! :D

8 hours ago, Justin Bacle said:

I'd say both, it's great to have more vertical resolution, but having to crop less is more "storage efficient" to me. (with magic lantern being limited by the card write speeds, it's easier to achieve more vertical resolution in 3:2 or 4:3 than in 16:9  )

I actually shoot more in 3:2 with magic lantern as it allows slight movement in the shot in post if wanted, but 16:9 is just too much cropping to me.

Yeah, in the case of ML, cropping goes smoother on the card.

I was thinking and maybe I didn't express myself quite right at the start. I started this rant because of the amount of messages I get from people saying "My camera doesn't have a 4:3 mode, so I'm not able to use 2x stretch lenses" and stuff like that. I don't know what leads them to think that.

[Justin Bacle]

6 hours ago, Tito Ferradans said:

Yeah, in the case of ML, cropping goes smoother on the card.

I was thinking and maybe I didn't express myself quite right at the start. I started this rant because of the amount of messages I get from people saying "My camera doesn't have a 4:3 mode, so I'm not able to use 2x stretch lenses" and stuff like that. I don't know what leads them to think that.

There is also the bitrate problem, let say you record in FullHD, AVCHD (24 Mbps).
Using a 2x strech scope, you would end up cropping the sides to get a 2.4 aspect ratio which implies cropping 1/3 (approximately) of the image you get. Decreasing the overall bitrate in use. You then end up with a 16 Mbps 2.35:1 image to work in post, which doesn't really allow a proper grade.
That is the reason why I don't use the anamorphic on my main camera (AF100).

I may very well be wrong, but I'd like to be proven otherwise. Any thoughts on that ?

[christrad]

On 14/02/2017 at 4:23 PM, Tito Ferradans said:

I wanted to understand why is everyone so obsessed with 4:3.

Is it because there's more height pixels, or because of the aspect ratio itself?

One answer : 4:3 don't crop horizontal flares!

I quit GH4 for the colors, now i'm happy with my NX1 except the 16:9 crop my flares :(.