Hubble Pallete.... Possible with DSLR imaging?

[StargeezerTim]

Hi all,

I love the colours and effect of the Hubble Pallete images I have seen. Is it possible to get this effect/colours with DSLR imaging or do you need a mono ccd and filters and other expensive stuff.

I anticipate this being a short thread! Ha ha...  Tim. 

[kalasinman]

No, yes

[glowingturnip]

like he said !

Afraid not old bean - Hubble palette is a way of falsely representing specific narrowband wavelengths from Hydrogen Ha, Sulphur SII and Oxygen OIII.  Mapping is I believe Ha=Green, OIII=Blue, SII=Red, but to the native eye, both the Ha and SII will look red/infra-red and are pretty close to each other.  OIII is a teal colour.

So your red channel on the DSLR contains both Ha and SII, and the blue+green together contain OIII (all along with a whole bunch of other non-narrowband stuff).

You could map your colo9ur channels to different colours if you wanted, but that would just be 'art', not 'science'

[pipnina]

I've been wondering about something similar myself...

Since the orange 580nm skyglow still shows up strongly in my blue and green channels, doesn't that mean narrowbands would still show up slightly in all my bayer'd DSLR pixels, by using the filters and changing each image to contain no saturation, I could make my own narrowband images...

Not sure if it would quite work like that... but I'm interested to try...

[Pompey Monkey]

I think you can. In theory. Here is a video that explains it: https://youtu.be/E842Rs-iB2I

However, it is just not an efficient way of doing so.

There is nothing to stop you using Ha, OIII and SII filters in the imaging train, processing them as monochrome and combining using the Hubble false colour pallette.

The issue are manifold: The IR filter in the camera reduces the Ha transmission by a lot (and from what the turnip said it will also reduce the SII a lot as well); the bayer matrix also limits the sensors' sensitivity to individual wavelengths depending on the colour and also effectively reduces the resolution of the sensor.

Of course the filters can be removed and the sensor de-bayered, but these are not jobs for the feint-hearted. There are threads on the subjects here on SGL: the de-bayering one is over a hundred pages....

[pipnina]

I think you can. In theory. Here is a video that explains it: https://youtu.be/E842Rs-iB2I

However, it is just not an efficient way of doing so.

There is nothing to stop you using Ha, OIII and SII filters in the imaging train, processing them as monochrome and combining using the Hubble false colour pallette.

The issue are manifold: The IR filter in the camera reduces the Ha transmission by a lot (and from what the turnip said it will also reduce the SII a lot as well); the bayer matrix also limits the sensors' sensitivity to individual wavelengths depending on the colour and also effectively reduces the resolution of the sensor.

Of course the filters can be removed and the sensor de-bayered, but these are not jobs for the feint-hearted. There are threads on the subjects here on SGL: the de-bayering one is over a hundred pages....

You can simulate bayer removal in DSS, stack your bayered Ha images with it and turn the bayer interpolation off (for debayered DSLRs) then reduce the saturation to 0

Or at least... That should work.

[Leveye]

With editing you can change the colors to anything you wish. I wouldn't do it but many do. It's your work you decide.

[Pompey Monkey]

You can simulate bayer removal in DSS, stack your bayered Ha images with it and turn the bayer interpolation off (for debayered DSLRs) then reduce the saturation to 0

Or at least... That should work.

But you'll still only have had one in four pixels illuminated by Ha or SII, reducing the effective resolution even if the resulting image has the same number of pixels as the whole sensor?

[pipnina]

But you'll still only have had one in four pixels illuminated by Ha or SII, reducing the effective resolution even if the resulting image has the same number of pixels as the whole sensor?

I get the feeling the bayer filter (on mine at least) isn't especially strong. Some data would still get through and I imagine if dithering is used (as it should be with a DSLR!) you shouldn't see any artifacts.

I will admit, it definately will NOT be as effective as a modded camera or a mono CCD. But if on a budget and a little wary it should provide some good results if you're willing to work for it!

Also, it seems like Psychobilly is having success with his: http://stargazerslounge.com/topic/68412-ha-filter-on-dslr/although I', unsure if he's modded or not.

[kalasinman]

Sorry, but with all due respect, a bad idea.

Filters can only remove data.

As was pointed out above the Bayer filter mechanically cuts the signal for red to 1/4, blue to 1/4 and green, which I believe contains none of the NB channels by 1/2.

What I've heard Oiii and Sii are quite weak, and even with a full mono/filter system may require hours of integration time. One not particularly deep image could take over a week of integration. Then there's tracking. Can your rig pull off clean subs of an hour or more each?

Many of the filters aren't parfocal and focus changes with time. Got auto focus?

My approach with my admittedly low budget rig is to make the best images I can by picking a task that fits the capabilities of my gear.

[StargeezerTim]

I think you can. In theory. Here is a video that explains it: https://youtu.be/E842Rs-iB2I

Lovely video, doesn't sound straightforward though from the posts on here. Think I will get on and start to try and get the best from my new modded 100d before committing to any more expense. Having said that, I really do need a coma corrector, especially with these larger nebs. Can't get away with cropping forever!

[StuartJPP]

Would you want to do it? Probably not, you would spend an entire year's worth of UK imaging time (and then some) to get pretty poor results on a single target, but of course it is possible.

You would need the 3 said filters to do it, and yes you will lose a lot of data but it is still possible...there are still at least 1 in 4 pixels that will be (in)sensitive to the wavelengths. Far from ideal, but not "impossible".

Have a look at Scott Rosen's website "Astronomers Do It In The Dark"...not Hubble Palette (SII missing) but shows what could be possible, but just look at his exposure times...

http://www.astronomersdoitinthedark.com/index.php?c=159&p=526