8 ms·

Anyone know how much of a difference are due to post processing of images when comparing webb and hubble? For example in the image of NGC7496 galaxy, Webb has m

Anyone know how much of a difference are due to post processing of images when comparing webb and hubble? For example in the image of NGC7496 galaxy, Webb has more resolution right, but did they make the colours more intense/shifted, or are these pictures a sort of representative of what we could see with our eyes https://d2r55xnwy6nx47.cloudfront.net/uploads/2022/07/NGC7496_Comparison.jpg https://d2r55xnwy6nx47.cloudfront.net/uploads/2022/07/NGC749...

jakear 4y ago
All the colors are false, it’s imaging inferred.
- z9znz 4y ago
  Politely suggesting that false has some negative connotations, perhaps "artificial" would be a better word... or even illustrative or representational.
  - jefftk 4y ago
    "False" is the traditional word here: https://en.wikipedia.org/wiki/False_color https://en.wikipedia.org/wiki/False_color
    - z9znz 4y ago
      Fair enough. Then it's really a matter of intended audience. It is a correct technical term for the color-domain educated audience. Could be that I'm one of the few outside that audience, but probably not :). The wikipedia article does note that pseudo-color is an alternative, and that carries less generally negative connotation. This is all nitpicking anyway, but thanks for the bit of education! [added - although now I can pick at the definition of "true color", since it is based on the idea of how a color would appaer to a human viewer. We know pretty well that there's quite a variance of perception of color amongst humans, so there arguably is no "true color". There's just X% of people see this color. :) ]
    - lapetitejort 4y ago
      For a similar term that can confuse outside audiences, see: https://en.wikipedia.org/wiki/Fictitious_force https://en.wikipedia.org/wiki/Fictitious_force
      - z9znz 4y ago
        Interesting. Perhaps it's like most other things in life. In hindsight, we might have done it differently. In this case, there are probably better, more descriptive (and less confusing) adjectives than "fictitious" which would succinctly describe a non-existent but perceived (or imagined) force. Even so, I never visualized something pushing me back into the seat. I always visualized the seat pressing against me. After all, I didn't feel sensations on the front of my body when accelerating. The sensations are the same as if someone has come from behind and starting pushing you.
        jakear 4y ago
        Removing yourself from the equation makes it easier to see: imagine a ball on a string in a car. Something happens and the ball moves forward relative to you. It’s not hard to see how you might assume a force pushed the ball forward.
- TremendousJudge 4y ago
  But aren't the percieved colors redshifted? Aren't these "false color" images just blueshifting the colors back to a visible light image, which is how they would look if they were moving at the same relative speed as us?
  - jakear 4y ago
    Roughly speaking sure, but the color mapping function they used just paints the longest wavelength the telescope can see “red” and the shortest “violet”. No attempt is made to properly “undo” redshifting. I’ve been unable to find any technical reports on the technique, but the gist of it is here: https://www.axios.com/2022/07/17/james-webb-space-telescope-nasa-photos-processing https://www.axios.com/2022/07/17/james-webb-space-telescope-...
    - heavenlyblue 4y ago
      It does seem like the redshift would need to be fixed on a star-by-star basis depending on how far away is the star.
      - falseprofit 4y ago
        I think you mean galaxy-by-galaxy.
  - adgjlsfhk1 4y ago
    depends on the observation. Anything in our galaxy or nearby galaxies is actual infra-red light at the source.
    - giantrobot 4y ago
      That is not true. The light reaching us from most objects in our and nearby galaxies is pretty close to the original wavelength emitted. It takes extremely relative velocity to red shift for visible wavelengths to shift to infrared. Most stellar objects emit a whole bunch of wavelengths of light, some emit more IR than visible. Interstellar dust doesn't red shift light due to velocity but because the molecules absorb higher energy photons and re-emit lower energy photons.
- ordu 4y ago
  This phrase implies that there are some "true colors", but are they? People see colors in weird ways, and it is a bold move to declare them to be the true ways. https://www.science.org/content/blog-post/chemistry-colors-images-and-reality https://www.science.org/content/blog-post/chemistry-colors-i...
  - jakear 4y ago
    Monitors output color via modulating three color channels. The modulation can either vaguely correspond to how the original wavelength would activate the three cone varietals in most humans, or can bear absolutely no relation to that. Equating the two is possible, but lazy.
    - ordu 4y ago
      I do not believe that human perception has anything to do with "true" colors. Such a complex system with so many quirks cannot be called the true way to deal with wavelengths. And yes it starts with cones in a retina, which wipe out most of information about a spectrum of a light. Some information got lost completely, some is mixed up, to a point where it is just vaguely represents the source signal. But if it seems to you too vague, then I can come to a point from a different angle: did you hear about tetrachromacy? "A study from 2010 actually suggests that a rare condition known as tetrachromacy exists in up to 12% of women in the world."[1] Don't you think that these women see colors in a "more true" way than other people? I personally think they are. It means, at least, that a more common way of a trichromacy is not the true way to see colors. It gives us not true colors, but some approximation which is very far from reality. You mentioned "most humans" as a reference point, so probably you believe that the truth can be established by a voting. I do not. I know that the objective truth is an unattainable ideal, but I do not believe that the majority of votes would be enough to establish an acceptable approximation to the ideal. It needs something more than that. [1] https://azretina.sites.arizona.edu/index.php/node/852 https://azretina.sites.arizona.edu/index.php/node/852
peanutz454 4y ago
The pictures are rarely what you will see with the human eyes (not as much colour).
krisoft 4y ago
> are these pictures a sort of representative of what we could see with our eyes The JWST sees in infrared. Our eyes don’t. This makes it not representative of what we would see with our eyes.
- Rastonbury 4y ago
  So when comparing Webb and Hubble images, we cannot rule out that the colours have been enhanced in post processing and comparisons taken with a grain of salt?
  - BurningFrog 4y ago
    The Webb pics would be invisible to humans without enhancing the colors. Though most things probably look pretty similar in visible light.
    - simonh 4y ago
      The main problem with visible light is it's absorbed and scattered by dust. It's Webb's ability to see through all that dust in the IR spectrum that reveals a huge amount of information and images Hubble and our eyes would never be able to see.
  - biorach 4y ago
    Hubble images are also frequently "enhanced". Not all the wavelengths these telescopes capture can be seen by the human eye so some post processing is necessary for the "publicity shots" A lot of the real science is not done visually but by analysing the data
  - jasonwatkinspdx 4y ago
    False color images are common in astronomy as they're working with spectrum ranges wider than our eyes. There's nothing about this that requires a grain of salt style skepticism.
  - whatshisface 4y ago
    There is no naked eye picture to compare it with, so you can't say enhanced. To answer your question, though, the JWST pictures and the Hubble pictures were turned into RGB images in the same way.
visarga 4y ago
> or are these pictures a sort of representative of what we could see with our eyes Webb can do "imaging spectroscopy", where it can take an image, but it will take a spectrum and every pixel of the image as well. In imaging spectroscopy, there is information on the spectrum of wavelengths present in each tiny piece of the image. This can help clue scientists in as to what elements or chemicals might have created that spectrum. https://www.space.com/james-webb-space-telescope-hubble-images-comparison https://www.space.com/james-webb-space-telescope-hubble-imag...
giantrobot 4y ago
First, all astronomy PR images are edited to look good to normal people. Sharpness is enhanced to bring out fine details and wavelengths are mapped to attractive colors to make an attractive composite. If you are a scientist working with data from Hubble or Webb you're working on the raw data and ignore the PR images entirely. The PR images are derived from the scientific data but they are not themselves scientific. Both Hubble and Webb also take observations with fairly narrow filters. In a PR image these essentially grayscale images are mapped to a color channel, red, green, or blue. The wavelength mapped to the "red" channel might be a mix of 650nm and 600nm filters. This blend will not look like what your eye would see but makes for a pretty desktop wallpaper. Webb in particular only sees in the IR bands which your eye can't see. Making a PR images from Webb images arbitrarily maps IR wavelengths to RGB channels. That's not to say a researcher won't filter and process the raw data. But their goal is to extract useful scientific information and not just produce an attractive image. They'll spend more time looking at graphs from the raw data than the actual 2D array of pixels. For instance, if you know the composition for an object you know what it's spectra should look like. If you compare that spectra to what's received you can calculate the red shift of every pixel. Doing this for the whole image can then let you tell the red shift of an entire object giving an idea of its proper motion relative to us. For say a nebula this can show how it's moving and give an understanding of its 3D structure even though the images are all 2D. I don't mean to disparage the PR images. They can often be used to explain phenomena to non-experts (not that I'm an expert). But they're never really going to show what the human eye might see, most stellar objects would look like mostly diffuse white blobs even "up close". Most things emit a number of wavelengths which just mix and look white to us. It's only with narrow band filters that telescopes can actually pick out fine details of most objects.
infogulch 4y ago
To anyone wondering how the sausage is made I recommend this video [1] by Nebula Photos who reprocessed the same raw data that NASA used to create the released nebula picture. At around 19m he talks about how he arranges the images in layers based on the wavelength of the filter that was used during its capture. About false color... Real things in the universe have real color -- they emit light in a variety of wavelengths -- even when that color is outside of the perceptual range of your eyes. The colorized photos like the ones released by NASA or Nebula Photos just expose the real relative color differences present in the original data, just shifted into a range that is you are able to perceive. "false color" images are definitely not someone taking a colored pencil to a black and white photo of a nebula and "coloring it in" to look nice, the colors are actually meaningful and give you more information about everything that the telescope actually detected than what a black and white photo could show alone. [1]: https://www.youtube.com/watch?v=DVuonz26P0w https://www.youtube.com/watch?v=DVuonz26P0w
- zoomablemind 4y ago
  > ...The colorized photos like the ones released by NASA or Nebula Photos just expose the real relative color differences present in the original data, just shifted into a range that is you are able to perceive. It's amazing to see the structure in such colorized images or even the picture of our Milky Way. I wonder, without the colorization, say, to a human eye, would such images appear mostly black (black-and-white)?
  - colechristensen 4y ago
    You just would miss a lot of detail with human limitations.
lattalayta 4y ago
Would love for someone to correct me if I'm wrong - but isn't this similar to how we look at nightvision or other heat seeking camera images? We falsely color them so that the temperature ranges are visible in the image, but it's not something that our human eyes perceive.

jakear 4y ago

All the colors are false, it’s imaging inferred.

z9znz 4y ago
Politely suggesting that false has some negative connotations, perhaps "artificial" would be a better word... or even illustrative or representational.
- jefftk 4y ago
  "False" is the traditional word here: https://en.wikipedia.org/wiki/False_color https://en.wikipedia.org/wiki/False_color
  - z9znz 4y ago
    Fair enough. Then it's really a matter of intended audience. It is a correct technical term for the color-domain educated audience. Could be that I'm one of the few outside that audience, but probably not :). The wikipedia article does note that pseudo-color is an alternative, and that carries less generally negative connotation. This is all nitpicking anyway, but thanks for the bit of education! [added - although now I can pick at the definition of "true color", since it is based on the idea of how a color would appaer to a human viewer. We know pretty well that there's quite a variance of perception of color amongst humans, so there arguably is no "true color". There's just X% of people see this color. :) ]
  - lapetitejort 4y ago
    For a similar term that can confuse outside audiences, see: https://en.wikipedia.org/wiki/Fictitious_force https://en.wikipedia.org/wiki/Fictitious_force
    - z9znz 4y ago
      Interesting. Perhaps it's like most other things in life. In hindsight, we might have done it differently. In this case, there are probably better, more descriptive (and less confusing) adjectives than "fictitious" which would succinctly describe a non-existent but perceived (or imagined) force. Even so, I never visualized something pushing me back into the seat. I always visualized the seat pressing against me. After all, I didn't feel sensations on the front of my body when accelerating. The sensations are the same as if someone has come from behind and starting pushing you.
      - jakear 4y ago
        Removing yourself from the equation makes it easier to see: imagine a ball on a string in a car. Something happens and the ball moves forward relative to you. It’s not hard to see how you might assume a force pushed the ball forward.
TremendousJudge 4y ago
But aren't the percieved colors redshifted? Aren't these "false color" images just blueshifting the colors back to a visible light image, which is how they would look if they were moving at the same relative speed as us?
- jakear 4y ago
  Roughly speaking sure, but the color mapping function they used just paints the longest wavelength the telescope can see “red” and the shortest “violet”. No attempt is made to properly “undo” redshifting. I’ve been unable to find any technical reports on the technique, but the gist of it is here: https://www.axios.com/2022/07/17/james-webb-space-telescope-nasa-photos-processing https://www.axios.com/2022/07/17/james-webb-space-telescope-...
  - heavenlyblue 4y ago
    It does seem like the redshift would need to be fixed on a star-by-star basis depending on how far away is the star.
    - falseprofit 4y ago
      I think you mean galaxy-by-galaxy.
- adgjlsfhk1 4y ago
  depends on the observation. Anything in our galaxy or nearby galaxies is actual infra-red light at the source.
  - giantrobot 4y ago
    That is not true. The light reaching us from most objects in our and nearby galaxies is pretty close to the original wavelength emitted. It takes extremely relative velocity to red shift for visible wavelengths to shift to infrared. Most stellar objects emit a whole bunch of wavelengths of light, some emit more IR than visible. Interstellar dust doesn't red shift light due to velocity but because the molecules absorb higher energy photons and re-emit lower energy photons.
ordu 4y ago
This phrase implies that there are some "true colors", but are they? People see colors in weird ways, and it is a bold move to declare them to be the true ways. https://www.science.org/content/blog-post/chemistry-colors-images-and-reality https://www.science.org/content/blog-post/chemistry-colors-i...
- jakear 4y ago
  Monitors output color via modulating three color channels. The modulation can either vaguely correspond to how the original wavelength would activate the three cone varietals in most humans, or can bear absolutely no relation to that. Equating the two is possible, but lazy.
  - ordu 4y ago
    I do not believe that human perception has anything to do with "true" colors. Such a complex system with so many quirks cannot be called the true way to deal with wavelengths. And yes it starts with cones in a retina, which wipe out most of information about a spectrum of a light. Some information got lost completely, some is mixed up, to a point where it is just vaguely represents the source signal. But if it seems to you too vague, then I can come to a point from a different angle: did you hear about tetrachromacy? "A study from 2010 actually suggests that a rare condition known as tetrachromacy exists in up to 12% of women in the world."[1] Don't you think that these women see colors in a "more true" way than other people? I personally think they are. It means, at least, that a more common way of a trichromacy is not the true way to see colors. It gives us not true colors, but some approximation which is very far from reality. You mentioned "most humans" as a reference point, so probably you believe that the truth can be established by a voting. I do not. I know that the objective truth is an unattainable ideal, but I do not believe that the majority of votes would be enough to establish an acceptable approximation to the ideal. It needs something more than that. [1] https://azretina.sites.arizona.edu/index.php/node/852 https://azretina.sites.arizona.edu/index.php/node/852

peanutz454 4y ago

The pictures are rarely what you will see with the human eyes (not as much colour).

krisoft 4y ago

> are these pictures a sort of representative of what we could see with our eyes The JWST sees in infrared. Our eyes don’t. This makes it not representative of what we would see with our eyes.

Rastonbury 4y ago
So when comparing Webb and Hubble images, we cannot rule out that the colours have been enhanced in post processing and comparisons taken with a grain of salt?
- BurningFrog 4y ago
  The Webb pics would be invisible to humans without enhancing the colors. Though most things probably look pretty similar in visible light.
  - simonh 4y ago
    The main problem with visible light is it's absorbed and scattered by dust. It's Webb's ability to see through all that dust in the IR spectrum that reveals a huge amount of information and images Hubble and our eyes would never be able to see.
- biorach 4y ago
  Hubble images are also frequently "enhanced". Not all the wavelengths these telescopes capture can be seen by the human eye so some post processing is necessary for the "publicity shots" A lot of the real science is not done visually but by analysing the data
- jasonwatkinspdx 4y ago
  False color images are common in astronomy as they're working with spectrum ranges wider than our eyes. There's nothing about this that requires a grain of salt style skepticism.
- whatshisface 4y ago
  There is no naked eye picture to compare it with, so you can't say enhanced. To answer your question, though, the JWST pictures and the Hubble pictures were turned into RGB images in the same way.

visarga 4y ago

> or are these pictures a sort of representative of what we could see with our eyes Webb can do "imaging spectroscopy", where it can take an image, but it will take a spectrum and every pixel of the image as well. In imaging spectroscopy, there is information on the spectrum of wavelengths present in each tiny piece of the image. This can help clue scientists in as to what elements or chemicals might have created that spectrum. https://www.space.com/james-webb-space-telescope-hubble-images-comparison https://www.space.com/james-webb-space-telescope-hubble-imag...

giantrobot 4y ago

First, all astronomy PR images are edited to look good to normal people. Sharpness is enhanced to bring out fine details and wavelengths are mapped to attractive colors to make an attractive composite. If you are a scientist working with data from Hubble or Webb you're working on the raw data and ignore the PR images entirely. The PR images are derived from the scientific data but they are not themselves scientific. Both Hubble and Webb also take observations with fairly narrow filters. In a PR image these essentially grayscale images are mapped to a color channel, red, green, or blue. The wavelength mapped to the "red" channel might be a mix of 650nm and 600nm filters. This blend will not look like what your eye would see but makes for a pretty desktop wallpaper. Webb in particular only sees in the IR bands which your eye can't see. Making a PR images from Webb images arbitrarily maps IR wavelengths to RGB channels. That's not to say a researcher won't filter and process the raw data. But their goal is to extract useful scientific information and not just produce an attractive image. They'll spend more time looking at graphs from the raw data than the actual 2D array of pixels. For instance, if you know the composition for an object you know what it's spectra should look like. If you compare that spectra to what's received you can calculate the red shift of every pixel. Doing this for the whole image can then let you tell the red shift of an entire object giving an idea of its proper motion relative to us. For say a nebula this can show how it's moving and give an understanding of its 3D structure even though the images are all 2D. I don't mean to disparage the PR images. They can often be used to explain phenomena to non-experts (not that I'm an expert). But they're never really going to show what the human eye might see, most stellar objects would look like mostly diffuse white blobs even "up close". Most things emit a number of wavelengths which just mix and look white to us. It's only with narrow band filters that telescopes can actually pick out fine details of most objects.

infogulch 4y ago

To anyone wondering how the sausage is made I recommend this video [1] by Nebula Photos who reprocessed the same raw data that NASA used to create the released nebula picture. At around 19m he talks about how he arranges the images in layers based on the wavelength of the filter that was used during its capture. About false color... Real things in the universe have real color -- they emit light in a variety of wavelengths -- even when that color is outside of the perceptual range of your eyes. The colorized photos like the ones released by NASA or Nebula Photos just expose the real relative color differences present in the original data, just shifted into a range that is you are able to perceive. "false color" images are definitely not someone taking a colored pencil to a black and white photo of a nebula and "coloring it in" to look nice, the colors are actually meaningful and give you more information about everything that the telescope actually detected than what a black and white photo could show alone. [1]: https://www.youtube.com/watch?v=DVuonz26P0w https://www.youtube.com/watch?v=DVuonz26P0w

zoomablemind 4y ago
> ...The colorized photos like the ones released by NASA or Nebula Photos just expose the real relative color differences present in the original data, just shifted into a range that is you are able to perceive. It's amazing to see the structure in such colorized images or even the picture of our Milky Way. I wonder, without the colorization, say, to a human eye, would such images appear mostly black (black-and-white)?
- colechristensen 4y ago
  You just would miss a lot of detail with human limitations.

lattalayta 4y ago

Would love for someone to correct me if I'm wrong - but isn't this similar to how we look at nightvision or other heat seeking camera images? We falsely color them so that the temperature ranges are visible in the image, but it's not something that our human eyes perceive.