> Many other cameras, particularly those with aggressive UV-IR cut filters, underespond to H-a, resulting in dim and blueish nebula. Often people rip out those filters (astro-modification), but this usually results in the camera overresponding instead.
Hmm... astrophotographers do not use cameras with UV-IR cut filters at all. For example, I owned a few of these:
They also generally do not use sensors that have Bayer filters. This also screws things up.
Instead they use monochromatic sensors with narrowband filters (either one band or multiple) over them keyed to specific celestial emissions. The reason for this is that it gets rid of light pollution that is extensive and bumps up the signal to noise for the celestial items, especially the small faint details. Stuff like this:
Often these are combined with a true color capture (or individual RGBL narrowband) just to get the stars coloured properly.
Almost everything you see in high end astrophotography is false color because they map these individual narrowband captures on the monochrome sensors to interesting colours and often spending a lot of time manipulating the individual channels.
Recently I've been on a bit of a deep dive regarding human color vision and cameras. This left me with the general impression that RGB bayer filters are vastly over-utilized (mostly due to market share), and are they are usually not great for tasks other than mimicking human vision! For example, if you have a stationary scene, why not put a whole bunch of filters in front of a mono camera and get much more frequency information?
These same things apply to satellite images of the Earth as well. Even when you have optical bands that roughly correspond to human eye sensitivity, they're a quite different response pattern. You're also often not working with those wavelength bands in the visualizations you make.
Scientific sensors want as "square" a spectral response as possible. That's quite different than human eye response. Getting a realistic RGB visualization from a sensor is very much an artform.
It's worth noting that many NASA images use the "HSO" palette which is false color imagery. In particular the sulfur (S) and hydrogen (H) lines are both red to the human eye, so NASA assigns them to different colors (hydrogen->red, sulfur->green, oxygen->blue) for interpretability.
The proper color of an image would be a multispectral radiograph similar to a waterfall plot for each point. Each FFT bin would be 100GHz in size, and the range would be over 1000THz. And in a way, that'd what a color sensor is doing at the CCD level too - collapsing and averaging the radio energy its susceptible to a specific color.
Having dabbled a bit in astrophotography, I would suggest that color is best used to bring out the structure (and beauty) of the object. Trying to faithfully match the human eye would, unfortunately, cause a lot of that data to be harder to see/understand. This is especially true in narrowband.
It is not just in space where nothing is lit by a uniform light source or with a uniform brightness. This is also true for many casual photos you would take on this planet.
Outside of a set of scenarios like “daylight” or “cloudy”, and especially if you shoot with a mix of disparate artificial existing light sources at night, you have a very similar problem. Shooting raw somewhat moves this problem to development stage, but it remains a challenge: balance for one, make the others look weird. Yet (and this is a paradox not present in deep space photography) astoundingly the same scene can look beautiful to the human eye!
In the end, it is always a subjective creative job that concerns your interpretation of light and what you want people to see.
> Because there’s a lot of overlap between the red and green cones, our brain subtracts some green from red, yielding this spectral response:
No, cones do not produce a negative response. The graph shows the intensity of the primaries required to recreate the spectral colour at that wavelength. The negative implies that the primary was added to the spectral colour to match it with itself, instead of adding it with the other primaries.
Yes, they do, after the photoreceptors. Those CIE colorspace curves aren't biology, and shouldn't be interpreted as such.
LMS colorspace is the (currently understood) biological colorspace [1], and contains inhibitions, from the opponent process [2] found in the meatware [3]:
red-green: L - M
blue-yellow: S - (L + M)
This contains a nice introduction to biological colorspace [4].
That's not a inherent property of deep space photos though, just of the sensors that are commonly used. There's no physical reason you couldn't build a telescope with a human response curve. It just maybe doesn't make a lot of sense from a science standpoint.
The question that the general public always wants answered is roughly, "If I was floating in a safe glass bubble in outer space looking at this object with my eyes, what would I see?"
Does anyone know the answer to this? Would it just be black? Or just a bright white star?
Another thing not mentioned is all the stuff further away gets redshifted down into IR thanks to the universe expanding. So while the original output might have been in the visible part of the spectrum its wavelengths have now been stretched out to something your eye can no longer see.
If you want to see all the cool shit 4 billion light years away, you are gonna have to get those retinal IR implants you keep asking for each Christmas installed.
depends on how close you are to say, the sun or a supernova.
also, without an atmosphere or magnetic field or day/night your eyes might not have as much longevity due to no shielding from harmful radiation/uv/etc.
23 comments
[ 3.3 ms ] story [ 40.3 ms ] threadHmm... astrophotographers do not use cameras with UV-IR cut filters at all. For example, I owned a few of these:
https://www.zwoastro.com/product-category/cameras/dso_cooled...
They also generally do not use sensors that have Bayer filters. This also screws things up.
Instead they use monochromatic sensors with narrowband filters (either one band or multiple) over them keyed to specific celestial emissions. The reason for this is that it gets rid of light pollution that is extensive and bumps up the signal to noise for the celestial items, especially the small faint details. Stuff like this:
https://telescopescanada.ca/products/zwo-4-piece-31mm-ha-sii...
https://telescopescanada.ca/products/zwo-duo-band-filter
Often these are combined with a true color capture (or individual RGBL narrowband) just to get the stars coloured properly.
Almost everything you see in high end astrophotography is false color because they map these individual narrowband captures on the monochrome sensors to interesting colours and often spending a lot of time manipulating the individual channels.
This is done at the medium to high end using the PixInsight software - including by NASA for the recent James Webb images: https://www.pbs.org/video/new-eye-on-the-universe-zvzqn1/
The James Web telescope has a set of 29 narrowband filters for its main sensor: https://jwst-docs.stsci.edu/jwst-near-infrared-camera/nircam...
Hubble pictures were famously coloured in a particular way that it has a formal name:
https://www.astronomymark.com/hubble_palette.htm
(My shots: https://app.astrobin.com/u/bhouston#gallery)
Canon has made a few astrophotography cameras:
https://en.wikipedia.org/wiki/Canon_EOS_R#Variants
There are also modified cameras available with the filters removed:
https://www.lifepixel.com/
Scientific sensors want as "square" a spectral response as possible. That's quite different than human eye response. Getting a realistic RGB visualization from a sensor is very much an artform.
Outside of a set of scenarios like “daylight” or “cloudy”, and especially if you shoot with a mix of disparate artificial existing light sources at night, you have a very similar problem. Shooting raw somewhat moves this problem to development stage, but it remains a challenge: balance for one, make the others look weird. Yet (and this is a paradox not present in deep space photography) astoundingly the same scene can look beautiful to the human eye!
In the end, it is always a subjective creative job that concerns your interpretation of light and what you want people to see.
No, cones do not produce a negative response. The graph shows the intensity of the primaries required to recreate the spectral colour at that wavelength. The negative implies that the primary was added to the spectral colour to match it with itself, instead of adding it with the other primaries.
https://en.wikipedia.org/wiki/CIE_1931_color_space#Color_mat...
Yes, they do, after the photoreceptors. Those CIE colorspace curves aren't biology, and shouldn't be interpreted as such.
LMS colorspace is the (currently understood) biological colorspace [1], and contains inhibitions, from the opponent process [2] found in the meatware [3]:
red-green: L - M
blue-yellow: S - (L + M)
This contains a nice introduction to biological colorspace [4].
[1] https://en.wikipedia.org/wiki/LMS_color_space
[2] https://en.wikipedia.org/wiki/Opponent_process
[3] https://en.wikipedia.org/wiki/Lateral_geniculate_nucleus#Col...
[4] https://color2.psych.upenn.edu/brainard/papers/Stockman_Brai...
Does anyone know the answer to this? Would it just be black? Or just a bright white star?
If you want to see all the cool shit 4 billion light years away, you are gonna have to get those retinal IR implants you keep asking for each Christmas installed.
also, without an atmosphere or magnetic field or day/night your eyes might not have as much longevity due to no shielding from harmful radiation/uv/etc.