Of Camera Sensors and Speed Boosters
In my last blog entry, I wrote about exposure and dynamic range and how to manipulate both in order to obtain visually compelling images. However, while both concepts seem to be rather theoretical at a first glance, their successful application in imagery is still very much based on the technical understanding of components, optics, etc.
Without a proper and thorough understanding of these elements, it is not only difficult to consistently produce good images that are capable of telling a story, it is also incredibly difficult to grow as a filmmaker and especially as a cinematographer.
Since one of my goals for this module is conducting camera tests in order to improve and deepen my understanding of the technical, physical, and optical aspects of cinematography and improving my craft, I decided to have a deeper look into the heart of a camera, which is the camera sensor. It is here, where the image ultimately gets produced and the sensor’s individual capabilities as well as limitations will define what the final outcome of an image will be like depending on various conditions.
Now this is going to be a big chapter. As you will see, there is a lot of (very technical) history without which you cannot truly understand a sensor’s importance. Thus, let me dive straight into it and guide you through my learning process.
What do Camera Sensors Do?
Camera sensors are the brain and the heart of any camera. Traditionally, shooting with film would require you to use physical film stock in order to capture the incoming light with. In contrast to that, digital cameras use a sensor to capture light, turning it into an electric signal. As such camera sensors are not only the brain and the heart of any camera, they are also the most important primal factor in defining exposure, which has to do with their physical size, resolution and technical build.
Sensor Formats
Camera sensors come in various sizes. However, the most common ones are currently (not an exhaustive list):
16mm,
Micro 4/3 (MFT),
APS-C,
Super 35mm, and
Full Frame,
the last of which equals the size of the frame of a 35mm film strip (measuring 36mm x 24mm), which is the oldest format and now considered the standard. For the Super 35mm cinema format, however, there are different measurements available, as delineated in this video below:
As a matter of practice, all other sensor sizes are usually referred to the full frame sensor in terms of size relation, which is commonly expressed by a crop factor. Crop factor describes how much the image of a certain sensor will be cropped in relation to a full frame sensor. As such, a MFT-sensor, for example – which is roughly 70% the size of a full frame sensor – has a crop factor of x2. In contrast to that, the APS-C sensor that the manufacturer Canon uses has a crop factor of x1.6 whereas the APS-C sensor of the manufacturer Nikon has a crop factor of x1.5.
This is important as the crop factor of any given sensor does have an impact on the effective focal length you are receiving. This can however be mitigated by the choice of focal length on the lens. Thus, a MFT sensor at 12mm focal length will have the same field of view than a full frame sensor at 24mm focal length, although it will affect other aspects, such as e.g. depth of field and bokeh.
Furthermore, certain sensor sizes come with a defined aspect ratio. As such, a small MFT sensor does have a 4/3 aspect ratio, whereas APS-C and full frame sensors tend to have 3/2 aspect ratio.
And if that was not confusing enough for starters, here’s an overview of a couple more sensor sizes of some popular cameras:
As you might glean from that chart, the exact size of the sensor size (expressed in mm) usually comes down to the individual manufacturer and may thus vary in exact size. The same goes for the crop factor, which I’ve mentioned a bit further up.
Does Size Matter?
In general, it has been argued that a physically bigger sensor outperforms a smaller one in low-light conditions. While sensor size itself does not make that much of a difference in good lighting conditions, the big differences become visible in low-light. As cameras with bigger sensors capture more information (meaning: light) than smaller sensors (assuming the same resolution), they are better suited in low-light conditions, offering more of a dynamic range while also needing less signal amplification, resulting in less visual noise as well.
Furthermore, bigger sensors also tend to have a bigger angle of view compared to cameras with smaller sensors (that is, if both cameras are placed at the same distance and have the same lens applied). That, in turn, means that different sensors will require different lenses in order to capture the same field of view.
Another interesting aspect of bigger sensors is the fact that a camera with a bigger sensor usually also requires lenses which glasses are bigger in diameter, which in turn has an impact on how the image appears itself in terms of depth of field. Many people tend to find images shot with bigger sensors to be better in regard to their depth of field performance and the quality of the bokeh, like Gary Sims points out:
However, there are also downsides to big-sensor-cameras: Cameras with bigger sensors are also more expensive than smaller ones, especially when they come with tailored (equally expensive) equipment. Which needs to be noted in terms of budget. Also, their weight increases, as bigger sensors require bigger camera bodies, lenses, and supporting equipment.
Native ISO
A quick look back to the image of the exposure triangle above reveals that ISO is one of the three important aspects of the exposure triad, having an equal influence on exposure next to shutter speed and aperture.
ISO, as a bastardised abbreviation, stands for International Organization of Standardization, formerly referred to as ASA (American Standards Association), and was brought over from film. As such, it is usually referred to as “film speed”, describing the light sensitivity of a film emulsion.
However, while it is true that it stands as a factor of the sensitivity to light in the context of film stock and chemically based emulsions, ISO means something slightly different in terms of the digital camera sensor. While the inherent principle between both would not be too different from each other, with digital sensors a more accurate definition of the ISO would be the degree of amplification of the signal (light) entering the sensor.
Since this value is selected dependent on the actual light conditions, most digital cameras will cover a range of ISOs that can be manually selected. A low ISO number, such as e.g. 100 or 200 – means that the signal is not amplified much – and would thus be used for normal to moderate lighting conditions. The higher the number, the more amplification of the sensor ensues, making these settings useful for low-light conditions.
Or so goes the theory.
Practically, because ISO in digital terms does not really equal sensitivity to light, dialling up the ISO – or gain – on a camera sensor only tells the software/camera to amplify the incoming electrical signal. Because of this, smaller electrical currents that are usually not picked up and thus not replicated in the later image, are now also amplified, resulting in ugly digital noise. The higher the number of the ISO, the stronger the amplification, the more issues, as you can see exemplified in this tutorial here:
Looking now at the concept of native ISOs or even dual native ISOs (cameras that have two optimal ISO settings), these turn out to be different for every camera and/or manufacturer.
The Canon EOS 600D has a native ISO of 100, the Blackmagic Ursa Mini 4.6k and Blackmagic Cinema Camera come with a native ISO of 800, whereas the Sony FS7 has a native ISO of whooping 2000. The term native here, however, essentially tells the filmmaker that the best result (with the smallest amount of noise) is to be achieved when filming with 800.
Technical Build – The History of Camera Sensors
Film
During the history of filmmaking, and for the longest time span within, movies were filmed on physical film stock, consisting of a sheet of cellulose and an emulsion based on crystals of silver halide. As such, and due to a lot of experimentation of film stock by companies such as Kodak, Agfa and Fuji, the recipe of the chemical emulsion – and the physical carrier material of that emulsion itself – changed quite a bit, becoming more and more refined as time passed.
As a result of that process, film stock not only has its own individual properties – both in procedure as well as visually – its 35mm version has also been considered the standard due to grain, resolution, colour rendition and also – here it comes – in dynamic range.
For the longest time, and until digital film cameras evolved drastically, the quality of film stock in regard to dynamic range and resolution was unmatched. Due to the physical-chemical size of the silver halide used within the chemical emulsion of film stock, 35mm film had a resolution just short of what 4k is today, meaning that it was not only able to portray more colours, but also in a higher dynamic range. Yes, there was such thing as film grain, but whilst it advanced to become a stylistic element of the medium of film – and is thus often romanticised as an aesthetic to the point that artificial film grain effects are sometimes added during post in digital production – the film stock material itself advanced as well.
Video
Now, with the advancement of digital film, film stock was replaced by a digital solution, which is nowadays our camera sensor. In the early days, this was a difficult endeavour as to the manufacturing of them. Digital camera sensors initially had to be built with photo diodes – essentially receptors – that catch the incoming light, turning it into electrical signals that can then be read out (or ‘converted’) into a digital image signal.
In the beginning, these photo diodes – much like film – were still based on chemicals such as silicon, brome and phosphor, using their atomic propensities to generate and control the flow of electrons – which is essentially a current – as a so-called P-N-Junction diode. Whilst I won’t be able to go into massive detail about the chemical and physical aspects, more detailed information on this process is provided in this video by Filmmaker IQ down below:
Types of camera sensors
Having delineated the basic principle of how a photo diode works, the only question that remains is how to get the ‘image’ off the diode and into something that remotely resembles an image?
While wiring together diodes seems like the quickest solution to this problem, it turns out that this is actually not an option as the wiring would create an electric current, which, in turn, would be captured on the image as well, creating visual noise. Thus, the transfer from the diode needs to be achieved in a different way.
Throughout the years, the technology of these sensors changed drastically. Aspects such as light sensitivity, noise reduction, and image processing itself changed, leading to the two main types of camera sensors, which I will quickly introduce here for matters of completion – the CCD and the CMOS sensor.
Charged Coupled Device (CCD)
These CCD sensors use no wires, but instead consist of a single piece of silicon, forming one massive diode. With this sensor, electrons are shifted over a register line by line, which then reads the energy levels of each line, translating it into the digital image information. This created less ‘spill’ of current than would have been with a wired solution, allowing for a relatively noise-free reading.
While the first CCD-sensors seemed like a stroke of genius, reading out the signal was not completely fault-free and took comparatively more time than filming with film stock, since every single line of every single image had to be read serially. Also, shifting and ‘reading’ the lines would cost a lot of charge; especially with higher resolutions as they require more lines (and thus pixels) to be read and translated.
Furthermore, due to the above-mentioned readout process, some signal loss in between the lines was still given; meaning that digital noise was recorded within the information of the image as well.
Complementary Metal Oxide Semiconductor (CMOS)
This sensor has been a major step-up from CCD sensors. Although they are built similar to them, CMOS sensors now include integrated circuits, such as SRAM, DRAM, and microprocessors. Now instead of physically shifting electrons to a conductor, each pixel on the sensor has its own conductor. This not only costs less power, but also allows for simultaneous reading of all image pixels. However, it also means higher noise levels in the visuals due the grid and transportation method.
Due to new processes and sensor types, the readings became progressively quicker, less fault-prone, and the sensors themselves – along with their cameras – became cheaper, allowing for a democratisation of the filmmaking industry as consumers and prosumers were now increasingly able to produce films on a smaller budget.
A Bonus Round: Speed Boosters
While our university stores do not yet own a speed booster and I do neither currently have the budget to either rent one from a third party provider, nor to buy one myself, I still thought I might add this piece of technology to my research of sensors, dynamic range and exposure.
So what is a speed booster?
According to DL Cade, a speed booster has essentially three effects:
It makes lenses faster
It makes lenses wider
It makes lenses sharper
Sounds like a bold claim, right?
Speed boosters essentially work like inverted teleconverters. Teleconverters consist of a combination of negative lenses (biconcave lenses) that are attached between the camera body and the lens, thus working like magnifying glasses. Because of that, they increase the apparent focal length while decreasing the lens speed.
This essentially creates the same effect as a telephoto lens, minimising the angle of view. Thus, a teleconverter with x2 magnification will double the effective focal length, but will also do so at the cost of two stops of light. Because of this, a 300mm lens with 2.8f will effectively behave like a 600mm lens with 5.6f. This however, also comes at a loss of overall sharpness.
Speed boosters now do act the opposite: Although they are usually attached in-between the camera mount and the lens, they consist of a combination of positive lenses (biconvex glasses). Due to the shape of its internal lens, a speed booster concentrates the light rays into one spot, increasing the amount of light falling onto the sensor, thus adding a stop of exposure to the dynamic range of the camera while simultaneously changing the focal length.
As such, a speed booster with the magnification of x0.71 will turn a 50mm lens with 1.4f to a 35mm lens with 1.0f. This also affects the angle of view in the camera, allowing the attachment a full frame lens onto an APS-C or MFT-sensor; gaining a nearly full-frame coverage without having to pay the price of a full frame camera.
References:
Benjamin Jaworskyj (2018) Does CAMERA SENSOR SIZE matter? 2018 [online] Available at: https://www.youtube.com/watch?v=UnLyg7TZEuY&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=38&t=0s [Accessed on 13 October 2019]
Cade, D. (2013) Metabones Speed Booster Adapter Makes Your Lenses Faster, Wider and Sharper [online] Image taken from: https://petapixel.com/2013/01/14/metabones-announces-revolutionary-adaptor-makes-ff-lenses-faster-and-wider/ [Accessed on 13 October 2019]
Channel 8 (2017) Camera Sensor Size Explained [online] Available at: https://www.youtube.com/watch?v=cBV8wpNmnpY&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=42&t=0s [Accessed on 13 October 2019]
Channel 8 (2017) What is the Metabones Speed Booster? [online] Available at: https://www.youtube.com/watch?v=louG2350wmo&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=27&t=0s [Accessed on 13 October 2019]
D3Sshooter (2017) How to measure the dynamic range of your DSLR [online] Available at: https://www.youtube.com/watch?v=n8ggmnLxlbk&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=76&t=0s [Accessed on 13 October 2019]
Diallo, A. and Butler, R. (2013) First Impressions: Metabones Speed Booster [online] Available at: https://www.dpreview.com/articles/2667195592/first-impressions-metabones-speed-booster [Accessed on 13 October 2019]
Filmmaker IQ (2015) The Science of Camera Sensors [online] Available at: https://www.youtube.com/watch?v=MytCfECfqWc&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=40&t=0s [Accessed on 13 October 2019]
Fstoppers (2018) No, Larger Sensors Do Not Produce Shallower Depth of Field [online] Available at: https://www.youtube.com/watch?v=ZUbU6exONdU&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=36&t=0s [Accessed on 13 October 2019]
Gary Explains (2018) Camera Sensor Sizes: Micro Four Thirds vs APS-C vs Full Frame [online] Available at: https://www.youtube.com/watch?v=N3-MtoKBZfo&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=35&t=0s [Accessed on 13 October 2019]
KINETEK (2016) Tutorial on Cinematography - Are Full Frame Cameras Better Than MFT [online] Available at: https://www.youtube.com/watch?v=kh3SOpziEDQ&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=77&t=22s [Accessed on 13 October 2019]
Malhotra, K. (n.d.) Full Frame VS Crop Sensor VS Micro Four Thirds: Camera Sensors Explained [online] Available at: https://digital-photography-school.com/camera-sensors-explained/ [Accessed on 13 October 2019]
Mikulec, T. (2019) Pushing and Pulling Film [online] Available at: https://thedarkroom.com/pushing-and-pulling-film/ [Accessed on 20 October 2019]
LensProToGo (2019) BLACKMAGIC POCKET 6k - LOW LIGHT and EXPOSURE TESTS [online] Available at: https://www.youtube.com/watch?v=mZUCPIuD4rA [Accessed on 25 October 2019]
Newman, B. (2015) Speed booster: Advance technology guide [online] Available at: https://www.whatdigitalcamera.com/technology_guides/speed-boosters-advance-technology-guide-60994 [Accessed on 13 October 2019]
Park Cameras (2016) Explained: Camera Sensor Sizes [online] Available at: https://www.youtube.com/watch?v=3t1IsLuHuhY&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=37&t=0s [Accessed on 13 October 2019]
Stronz Vanderbloeg (2016) How does a Speed Booster work? [online] Available at: https://www.youtube.com/watch?v=B3ywLouK-6g&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=26&t=0s [Accessed on 13 October 2019]
Sydney Portraits (2017) What is 'correct exposure'? Must-know photo video tutorial [online] Available at: https://www.youtube.com/watch?v=HTre17Q0pp0&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=3&t=185s [Accessed on 13 October 2019]
The Slanted Lens (2018) How to Push and Pull Film [online] Available at: https://www.youtube.com/watch?v=90bo5bb77_g [Accessed on 13 October 2019]
VMI (n.d.) A Filmmaker's Guide to Sensor Sizes and Lens Formats [online] Image taken from: https://vmi.tv/training/useful-stuff/Guide-to-Sensor-Sizes-and-Lens-Formats [Accessed on 13 October 2019]
Wolfcrow (2016) How does Camera Sensor Size Impact Cinematography and Photography [online] Available at: https://www.youtube.com/watch?v=O313wHbxYNA&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=41&t=0s [Accessed on 13 October 2019]
ZY Productions (2015) Full Frame vs crop - What's the difference? [online] Available at: https://www.youtube.com/watch?v=WEGA6yrHcM0&list=PLRG4t0YYtkIzejIqbZ-qKOLjQchEri9ev&index=39&t=0s [Accessed on 13 October 2019]