RSS
 

How does anamorphic photography work?

25 Aug

Last time we posted about an anamorphic lens there were lots of questions about what anamorphic photography is and how it works, so we thought we’d offer a simple guide to the basics to help everyone understand what it is.

The word ‘anamorphic’ comes from the Greek words ‘ana’ and ‘morph’ which together suggest something that alters its shape but then changes back to normal again. In photography, it relates to a situation in which an image is distorted as it is taken, but is then undistorted to a normal shape when it is projected or displayed. A common example of anamorphic imagery can be seen in the cinema, when a movie is shown in that long letterbox format that stretches across the screen. This characterizes what most of us think of as anamorphic.

A 2.35:1 format image shot using the Sirui 35mm F2.8 1.33x anamorphic lens. The image was recorded in 4K video, so it started life in the 16×9 format.

The optical anamorphic process was invented during World War I to help observers in tanks get a wider view of the battleground without having to make the observation hole any larger, and the system was used on-and-off in cinema once the war was over. Anamorphic films became more popular in the early 1950s with the process re-energized to provide an exciting alternative to the almost-square format of television.

With the growing number of TVs in homes, Hollywood wanted to ensure cinema retained some unique qualities that would still make people leave their living room and part with their money to watch a movie. The long letter-box format is now synonymous with epic cinema all over the world and is a subconscious indicator for the audience that the movie they are watching has high production values.

That link with the atmosphere of ‘serious’ filmmaking is why anamorphic photography is so attractive to amateur and professional filmmaker alike, as it can lift production value in the eyes of the audience (or client), and elevate the filmmaker from the throng of video-makers shooting 16:9 or 4:3 on ‘lower-end’ equipment. Warranted or not, many people see using the anamorphic process as a leg-up on the way to artistic greatness.

The [anamorphic] letter-box format is now synonymous with epic cinema all over the world

The image formats associated with anamorphic cinema are also pretty interesting in their own right and engage some different compositional properties that are genuinely useful and unique compared to those that apply to a typical still image – in the same way that panoramic formats work in still photography.

Anamorphic aspect ratios

While in stills photography we tend to use whole numbers when discussing the aspect ratio of any given format, such as 3×2, 4×3, 5×4, 10×8, in anamorphic cinema these things are measured using 1 as the height of the frame. So, popular aspect ratios these days include 2.35:1, 2.39:1 and 2.40:1, though the official standard according to the Society of Motion Picture and Television Engineers (SMPTE) specifies 2.39:1 for widescreen projection.

CinemaScope is 2.66:1 and belongs to 20th Century Fox, but there were a whole load of other formats devised by other studios that didn’t fancy paying Fox for the license to use 2.66:1. Hasselblad fans will recognize the CinemaScope proportions as they are approximately the same as the XPan format that lives on via the 65:24 ratio in the X1D II camera.

This is a 2.66:1 CinemaScope format image, created by using a 2x anamorphic lens while recording 4:3 full-sensor video in a Micro Four Thirds camera

In this digital age, filmmakers can use whatever format suits them, though there is some value in sticking to an established ratio just for familiarity and what it might mean to the audience. The movie La La Land, for example, is shown in CinemaScope to help invoke a sense of the age it portrays – audiences, not just of a certain age, pick up on these things subconsciously and it adds something to the picture.

This diagram shows how different popular projection formats compare. The 4:3 aspect ratio was popular in film and still is in digital sensor formats, while 3:2 is what you get when you shoot full frame and with APS-C/Super 35 sensors, while 16:9 is the standard for most digital cameras in video mode and what we see most in popular video. Widescreen really starts at 2.35:1 and 2.39:1 with moderate anamorphic lenses, and 2.66:1 provides a really long and thin widescreen format.

In film-based cinema, the ends of a wide format might be cropped from the picture to meet the 2.39:1 requirement, especially when a 2x anamorphic lens is in use, but in digital video, a timeline of any proportions can be created to show a finished product in 3.5:1 if desired.

How the format is made

Normal, spherical, lenses look all around themselves in equal measure – viewing at the same angle left/right as they do up/down. Anamorphic lenses capture an elongated horizontal field of view. To achieve this, the lens squeezes the image horizontally to fit within the constraints of the sensor’s dimensions.

This anamorphic image was captured using an anamorphic lens on an iPhone 11 Pro. Use the slider to compare the desqueezed image (L) with the squeezed image (R).
Photo by Dale Baskin

This effect can be seen in the anamorphic video clips below.

That squeezed view has to fit on to a relatively square sensor, such as a 4:3 Micro Four Thirds chip, so the anamorphic element group in the lens squeezes/distorts the horizontal view so that it will fit into the available sensor space. To do this a cylindrical element is used that has the shape of a section cut from a tube – it is bent in only one plane rather than being convex all round as a normal lens would be.

This is clearly a piece of paper and not a glass lens element, but it gives you an idea of the shape of the anamorphic cylinder element that creates the wider horizontal field of view without changing the vertical field of view.

That cylindrical lens is the shape produced when you bend a sheet of paper – bowed in the horizontal aspect but still flat in the vertical aspect – allowing it to capture a wider field of view horizontally than it does vertically. Like in a Hall of Mirrors, this distorted surface creates a distorted image on the sensor or film. When projected to show the audience, that distorted image is passed through another anamorphic lens to distort the view once again, but this time in reverse – un-distorting it so that it looks normal. Historically, in anamorphic cinema, both the camera and projector are fitted with anamorphic lenses.

In the digital world an anamorphic lens is needed only to record the image, as software can be used to stretch the recorded image and make the subjects look geometrically correct again.

This picture was taken with a 1.33x anamorphic lens in stills mode on the Lumix GH5. The recorded image measured 5184 pixels wide and 3456 high, as shown in the Image Size window of Photoshop
To find the length that the image needs to be for the subject to look normal you multiple the recorded length by the anamorphic factor – in this case 1.33x
With the width and height dimensions unlinked you just enter the new width dimensions. In this case 5184 x 1.33 = 6895 pixels. Hit ‘OK’ and the image stretches to the right anamorphic format

In still photography de-squeezing a picture is pretty straight forward. You simply multiply the horizontal pixel count by the squeeze factor of the lens. So, if your original image measures 4000×3000 pixels, for example, you multiply 4000 pixels by the squeeze factor to get the width the final image should be. If the lens had a 1.33x factor we multiply 4000 x 1.33 to get 5320 pixels. In the Resize dialogue of your editing software, unlink the horizontal and vertical resolution figures so the aspect ratio can change, and then replace the 4000 with 5320 for the horizontal dimension, keeping the 3000 pixel (vertical) dimension unchanged.

Why not just crop?

You would think it would be easier just to crop a normal picture to make a letterbox format than going to all the bother of getting special lenses – and you’d be right. The issue though is that when you crop you create a lower resolution image – whether on film or on a digital sensor – and either waste film or pixels in doing so. Anamorphic lenses create an image that fills the film frame/sensor area so all those pixels you paid for are used.

This is a frame from a 4K video recorded with the Sirui 35mm 1.33x anamorphic lens. It uses all eight million of the sensor’s pixels. In contrast, cropping a 16:9 video frame to this 2.35:1 format would give us an image with roughly 6MP of data.

Shooting video using a 4K camera produces frames that are each about 8MP. Once you crop that 4K image to an anamorphic format, such as 2.39:1 for example, you end up with footage containing far fewer pixels. 4K frames shot in 16:9 (1.78:1) are 3840 x 2160 pixels, but when that frame is cropped to 2.39:1 it becomes 3840 x 1606, which is only 6.2MP. Using an anamorphic lens allows you to record using the full 4K area of the sensor, thus retaining all those pixels so the resulting 2.39:1 footage retains 8 million captured pixels instead of just 6.2MP.

This is a still image recorded on the GH5 through the Sirui 35mm 1.33x anamorphic lens. The top image represents what the view looked like, and the second image is how the image looks once the lens has squeezed the wide aspect onto the 4:3 sensor. In software, I de-squeezed the 4:3 captured frame to 16:9 so that the subject would look normal.

Some cameras, like the Panasonic Lumix GH5 and GH5s offer a specific Anamorphic mode that allows the whole 4:3 sensor area to be used to record the footage. In this mode, the GH5 can create 6K footage in which each frame contains the full 20MP resolution of the sensor. When that image is de-squeezed to produce the anamorphic final result those 18 million pixels will still be present.

If you were to use a 1.33x anamorphic lens like the Sirui 35mm F1.8 the footage de-squeezes to a 16:9 format, but one that contains 18MP instead of the 14MP you’d get by simply cropping the full frame to 16:9. Even then, cropping this 16:9 image to 2.35:1 will deliver a higher resolution frame than shooting with a 16:9 area of the sensor in 4K – 10MP instead of 8MP.

Here’s the Cooke 32mm T2.3 Anamorphic/i 2x lens on the Lumix GH5 – well, it’s more like the camera is on the lens rather than the other way round. With the camera in its Anamorphic Mode the 4:3 sensor-captured image de-squeezes to make a 2.66:1 CinemaScope format picture.

The difference is more pronounced when using lenses with a greater than 1.33x anamorphic squeeze factor. A 2x lens, such as the Cooke 32mm Anamorphic/i would create a 2.66:1 output from the full area of a 4:3 sensor or a 3.5:1 final result from a 16:9 area. So, if you were cropping 4K footage to match those aspect ratios you’d end up with 3840 x 1444 pixels (5.5MP) for a 2.6:1 format or 3840 x 1098 pixels (4.2MP) for 3.5:1 format. Both of those represent a significant drop in resolution from the original 8MP of 4K footage – which is all preserved when using an anamorphic lens.

Resolution isn’t the only benefit

Retaining decent resolution isn’t the only reason to shoot with an anamorphic lens: these lenses have specific characteristics many people find attractive.

The almost trademark blue streak extending across the frame of an anamorphic picture comes from point light sources reflecting in the surface of the anamorphic cylinder and spreading out across the scene.

The most commonly recognized characteristic is a blue streak that shoots across the frame when a point light source is aimed at the camera – a car headlight for example. This is caused by direct light reflecting off the anamorphic cylinder and then spreading out left and right across the frame.

Obviously, these blue lines are more prevalent in lenses that have the anamorphic cylinder at the front of the construction, and much less obvious in those that place the cylinder at the rear. The new Arri/Zeiss anamorphics spread the cylinder effect throughout the lens construction, rather than having a specific group of elements to do the job, which allows a degree of control over how dominant the blue streaks will be. In more regular anamorphics the blue streak effect can be played up with reflective coatings inside the forward elements to enhance the color of the streak and how easily it can be ‘activated’.

The bulbous anamorphic cylinder can be a magnet for light and can reduce contrast when even off-center lights are pointed towards the lens.

As well as this specific type of flare, light falling on the front element will create an overall flare that in turn can give anamorphic footage a low contrast atmosphere even when contrast is quite high. This again depends on the design of the lens. Older lenses tend to flare more easily while newer designs aim for more contrast and allow filters to be used when lower contrast is desirable.

The oval shape of out-of-focus highlights is usually demonstrated in night scenes with distant car lights, but this characteristic is also visible during daylight hours. Here you can see the light between the trees – which would usually appear round – takes on an upright oval shape.

The other immediately recognizable characteristic of anamorphic lenses is the elongated shape of out-of-focus highlights. These highlights – a street light in the distance for example – would reproduce bright discs in pictures taken with a normal spherical lens, but when shot using an anamorphic lens they appear as ovals. In fact, all out-of-focus details are reproduced with an elongated shape that exaggerates the degree to which things are out of focus. This in turn only makes the focused subject stand out more.

The appearance of an extra-shallow depth-of-field is further enhanced by the complications of the altered angles of view we get with an anamorphic lens. A lens with a 1.33x anamorphic effect will have its marked focal length widened by the anamorphic factor – so a 100mm 1.33x lens would deliver the angle of view of a 75mm (100 divided by 1.33 = 75). With more dramatic anamorphic lenses the effect is more pronounced too, so a 1.8x which would give that 100mm the view of a 56mm. The final look is of a 56mm lens that exhibits depth-of-field characteristics similar to those we would expect from a 100mm lens.

Above you can see how the same scene is reproduced differently by a normal spherical lens and an anamorphic lens of the same focal length. I used the Lumix X-Vario 12-35mm set to 35mm to compare with the new Sirui 35mm 1.33x anamorphic lens.

The camera-to-subject distance remained the same, as did the F2.8 aperture, but there is a slight difference in the degree to which the background appears out-of-focus. As you can see, the subject appears much smaller in the anamorphic images due to the extra width of the view, so naturally, a photographer would normally get closer to make the subject fill the frame, and thus increase the shallow depth-of-field effect simply by using a closer focus distance to achieve the same subject magnification.

What is also clear from these images is that the anamorphic lens delivers a considerably wider view for the same marked focal length. This comparison also shows the shape of out-of-focus highlights from the same scene rendered quite differently.

Is it worth the effort?

That’s a matter of opinion of course, but those wanting to make the most of all the tools available to influence the audience will say ‘yes’. The look is special and it can add something very substantial to the atmosphere of a film. As mentioned earlier though, an anamorphic lens can’t make a poor film into a good film, compensate for bad lighting, primitive camera work or wooden acting – it is only a part of the many elements that can make a movie an award winner or a rotten tomato.

Street lights just out of the frame (and a high ISO setting) contribute to a nice soft contrast in this scene, even though the actual scene was filled with deep shadows. The look and feel of the shot are different enough to that which we would expect from a regular spherical lens that we can tell there is a certain something else about it. The highlights and background details look a bit different and there is a wide feel but without the usual distortion of a close perspective.

Anamorphic photography also isn’t suited to all subject types, and while not a fast rule it tends to work best with drama rather than documentary. The widescreen says ‘now I’m going to tell you a story’ and can prepare the audience for all the exaggeration that makes a story moving, dramatic and emotional, while more regular formats might be better for presenting strictly factual information.

There are in-between cross-over areas though that still work well, such as those old wildlife films that present factual information with a deep Hollywood voice-over and in which all the lions in the family have a name and roam the grasslands to the sound of a full studio orchestra.

There’s also a sense of cinema about a still shown in anamorphic format

In stills photography, what an anamorphic lens will give us is something a bit different. ‘Different’ is something I value, though obviously ‘good different’ rather than the other. ‘Different’ makes our work stand out from the rest, and as there aren’t many stills photographers using anamorphic lenses ‘different’ is what you will get.

There’s also a sense of cinema about a still shown in anamorphic format, and with the built-in characteristics of an anamorphic lens that inherent atmosphere will feel stronger, making it possible to present movie-stills filled with an implied storyline – without actually having to go to the bother of shooting the movie.

Articles: Digital Photography Review (dpreview.com)

 
Comments Off on How does anamorphic photography work?

Posted in Uncategorized

 

Tags: , ,

Comments are closed.