As we can see from the episodes themselves, Cam vision is very similar to human vision! One way it differs is that when Cams are physically hit or shocked, their vision greys out with scanlines and other artefacts.
You can see a real-life example of this in this video, in which a skydriver drops a video camera and it keeps recording until it lands (in a pig pen!)
As the camera tumbles, it spins faster than its sensors can keep up with, resulting in the video turning into a series of diagonal lines. Most digital cameras record in a 'sweep', going from one edge or corner to the opposite one. It looks as though this camera records in a diagonal sweep.
This also means that Cameramen possibly don't see things like propellors and hummingbird wings as a blur, as we do - they'd see it as a choppy effect, depending on how the movement syncs with the Cameraman's framerate.
Here's a bird appearing to hover perfectly because its wings happen to match the camera's shutter speed:
As far as Cameramen are concerned, that's probably just how birds are! They don't see bird wings as a flippa-flappa or propellors as a rounda-rounda blur.
Presumably, Cameramen cannot perceive the chimerical colours, which are colours humans can see thanks to how our retinas form afterimages.
Soundkind
I headcanon that Soundkind primarily perceive the world through echolocation. It's not really possible for humans to grasp what this 'looks' like. We've all heard of Thomas Nagel's essay 'What is it like to be a bat?' But at least a bat is a mammal like us, so it'll have mostly the same neural pathways as us. It's not outside the realms of possibility that bats 'see' the world in the same colours as us, but the colours are triggered in the brain's perception by wavelengths of sound rather than light. But Soundkind aren't even animals! Who knows how a tech brain processes this data?
I also headcanon that the Soundkind can see with light to a degree, just not as well as a human, Cameraman or TVman. Their vision is dim, low-res and is heavy on the reds (hence their use of red as an emblematic colour), similar to a human with tritanomaly. They're like cats, in that going blind isn’t really a big deal. If a housecat loses its sight, well, it was mostly using smell and whisker-touch to navigate anyway; its vision was more of a nice-to-have. Speakermen whose visual receptors stop working don’t appear to be particularly impaired - sure, they can’t read printed text any more, but they could read only high-contrast very large print anyway.
Here's Tri-Cam, approximately as a human or Cameraman would perceive them (from this SketchFab model):
Here's how Soundkind might 'see' Tri-Cam with echolocation. (I took the 'matcap + surface' version of the model to get just the contours without textures, then noodled with it a little in GIMP.)
Here's a scene approximately how a human or Cameraman would perceive it (source):
And how it might look to a Soundkind using their visual processing - mainly red (achieved using a tritanopia simulator), and low resolution:
Here's another 3D model, a bit more colourful than Tri-Cam, as a human or Cameraman might perceive it (source):
Putting together the above info, here's how a Soundkind might perceive it using their echolocation overlaid with their optical vision:
TVs
TVs are primarily optic-based, like humans and Cameras. Their vision overlaps less with the human range than the Cameramen's vision does. TVs see further into the infra-red than humans and Cams (unless said Cams have specialised thermal vision), and their vision has much more of a purple cast, hence their use of purple as an emblematic colour.
TVs can see even in complete darkness (by our standards), because everything above absolute zero gives off IR radiation. However, an environment of absolutely uniform temperature would appear blank, because there would be no differences in IR for the TVman to see. In addition, TVs can't see the temperature of stuff behind glass - they would only see the temperature of the glass itself (same as the limitations for real thermal cameras).
My infra-red headcanon is purely based on the fact that TV remotes are usually IR-based (more modern ones use Bluetooth).
…I don't have a speculative image to show, because to my surprise I couldn't find good quality pictures online of the same scene taken with RGB and with thermal (so I can combine them the way I want them). They were all either postage stamp-sized or they didn't align properly. Hopefully you can imagine something!
This video shows something a bit like what I'm thinking of:
YouTube link — How to use the FLIR ONE Thermal Imaging Camera
The video shows that the thermal cam has something called 'MSX', which combines the thermal image with the light image, giving you the thermal imaging info combined with the surface patterns from the light imaging info. However, my headcanon is closer to 'human vision (but a bit more purple) combined with a black-hot greyscale thermal image', as opposed to the purple and yellow mess seen here.
