Narrowband imaging

[mik]

Hi over recent weeks there have been lots of images posted on the forum described as Narrowband and as a non imager i am wondering if someone can give a breif explanation as to what this is what equipment is used ie cameras filters and what the benefits of this method are thanks.

Mick.

[Keagle]

Narrowband is taking pictures in very specific wavelengths of light.

This is all done using filters. You can get various types of narrowband filters, such as Hydrogen Alpha (Ha) which is probably the most common, Hydrogen Beta (Hb), Sulpher III (SIII), or Oxygen III (OIII).

These filters simply let in a very narrow band of light (hence the name). They are named after gases because that is the type of gas that produces the wavelength of light that these filters let in.

For example, a Hyrodgen Atom is excited and as the electron jumps up to the next level and back down, the energy given off as it jumps back down a level is emitted as light. This light is a specific wavelength, of which a Hydrogen Alpha filter will let through while blocking out any other light not at this wavelenth.

Imaging wise, many use Hydrogen Alpha filters to create a luminence layer. Then standard RGB exposures are used to colour it, or imagers will do normal LRGB and then add Ha to add detail.

Now, the filters let in light at a specific wavelength pass.

Ha is near the red end of the light spectrum, as is SIII, so these are normally used as red in true-colour images.

Hb is near the blue end of the spectrum, so in true-colour narrowband images these are used for blue.

OIII is in the green spectrum, and in true-colour images used as green.

False-colour, like Hubble, is where they've got exposures in say Ha, SII and OII.

Hubble uses:

OII = Blue

SIII = Red

Ha = Green

This creates a false-colour image, usually used to emphasise a specific image or simply for asthetic purposes.

Camera Wise, most narrowband imagers use mono cameras as these are much more sensitive to light so putting a filter over it utilises the whole chip. It can be done with a colour camera but it is not as effective, as a bayer matrix is there too so, for example, if you have a Ha filter on, in a bayer normally only 1/4 of the bayer is red and as such you are only getting 1/4 of the total Ha light coming in (as Ha is near the red spectrum).

The advantages of this method are you can do it even during a full moon. As it is "narrowband" the wavelengths of light coming in are very narrow, it does not let in moonlight as the light coming from the moon is not the right wavelength to pass through the filter. You can play around with the data, false-colour, true-colour etc.

The disadvantages are it's a lot more technical, utilising sets of filters so unless you have a filter wheel you need to refocus every filter change, requiring more post-processing time, and longer exposures as you are only getting very specific wavelengths of light, cutting out a lot of the others, so you have less light than you normally would hitting the sensor (only the light coming through the filter, as opposed to having no filter at all).

That's how I understand it, I've tried to make it as clear as possible but I've probably rambled it. I'm probably wrong, too - but I hope not!

Ah, teatime now!

Kurt

[Kaptain Klevtsov]

Hi Mik, here goes then.

Nebulae give off light due to stuff happening at an atomic scale, in a similar way to a neon lamp being red, and only ever red. If you use a filter than only lets that particular sort of red light through, you could image a cityscape at night and only see the neon lights in the image. You wouldnt see streetlights, car headlights or buildings with the lights on. In the same way, using a filter tuned to nebula wavelengths you can cut out all the other light which gets in the way of a nice image. This includes street lighting, the moon etc.

To get a colour image, you image through several different filters and get several different black and white images. Using something like Photoshop, you then asign a colour to each B&W image and put them on top of each other to give the final coloured result.

Kaptain Klevtsov

[narrowbandpaul]

Thats pretty much it kurt...

though strictly the reason that moonlight is cut out is not because the does not emit ha (I have an image of the moon in ha posted on this forum), but because the moon is a broadband source (light over all wavelengths). So whilst the total amount over all wavelengths is large (the moon is bright), the total amount passing through the filter is low, since the filters only pass a little bit of light around the central wavelength.

The filters have a property called Full Width half Max...which tells us how 'narrowband' the filter is. To be classed as narrowband, the fwhm has to be around 13nm or less. For example take Ha at 656nm, and a FWHM of 10nm and a peak transmission of 90%. Transmission is just the ratio of light that gets through the filter, as gets rejected.

So the max value is 90%, so the half max is 45%. The filter passes light with 45% transmission at 656 +- 5nm, with 90% at 656nm.

The narrower the filter the more moonlight gets blocked. i use a 6nm filter, and can image in midsummer with full moon for 20mins, and still achieve a black background.

There is something subtle here...the ha line is not alone. it is flanked on either side by N[iI] (say Nitrogen 2) lines, only around 1nm apart. So with a typical filter, the ha is actually a combo of Ha and N[iI], which are farly bright.

There is nothing wrong with this..it looks as if it were pure Ha, but from a scientific viewpoint, the N[iI] is actually a noise source, since we dont want it, we just want Ha. To remove this line, we need a very narrowband filter, a 3nm filter will work, but even better would be a 1.5nm filter from the likes of Andover Corp, but this could be pricey.

So we make do with our N[iI] ande Ha combo, and are completely happy

Another point, which is not in any way important for imaging, but its nice to know, is what the square brackets round the emission lines mean

For example...

S[iI]

O[iII]

He[iI]

He

firstly the roman numeral refers to the ionisation state of the element...I means unionised (all electrons accounted for), II means singly ionised (one elctron removed) and so on

The square brackets mean forbidden line...meaning that we do not see this line in a laboratory or anywhere on earth. It is only produced in the cold, low density conditions of space, where collisions are very infrequent.

Why use these lines in particular, ie Ha, S[iI] and O[iII]...well they are uncorrelated...the intensity of line one does not affect the intensity of another...

What do these lines show us...heres my theory

I view ha as a kind of background stuff thats always there, on top of which we place the O[iII] and S[iI] data.

I read a paper that claimed that S[iI] was boosted in regions of compression. I agree with this. Time after time one sees a good S[iI] signal coming from the edges of nebula...like the wall in cygnus

O[iII], is a doubly ionised state, meaning it is found in high temperature regions...therefore, the presence of O[iII] indictates high temps and high energy events.

When this data is combined with the background ha, we see that the centre of nebula are often hot, and are surrounded by a wall of compression. this to me creates a very 3d view...more so than conventional RGB imaging

It is a more scientific approach, that yield more scientific looking images (one can instantly see high energy regions), but images that look really cool. and thats why i like it.

To see the O[iII] and S[iI] generally you need longer exposures...if you use one hour of ha (say 6x10mins) subs, then i would recommend 2 hours each of O[iII] and S[iI], and good darks, to remove noise, and flats

but the effect really are cool.

Thats a wee bit more science behind the theory of narrowband imaging

Hope that helps...no doubt you have seen the cool images taken on this forum.

For more great pics, check out narrowbandimaging.com, from richard crisp

this guy invented ( no joke) the subject

Paul

[mik]

So all in all Kurt knows stuff, KK knows stuff and makes me laugh & Paul knows lots of stuff and made my head hurt :?

Thank you very much for the expanations i will read through a few more times but i do get the idea.

Cheers Mick.

[narrowbandpaul]

what didnt you understand fully...its easier to answer specific questions

[mik]

This bit :?

The filters have a property called Full Width half Max...which tells us how 'narrowband' the filter is. To be classed as narrowband, the fwhm has to be around 13nm or less. For example take Ha at 656nm, and a FWHM of 10nm and a peak transmission of 90%. Transmission is just the ratio of light that gets through the filter, as gets rejected.

What does nm stand for ?

Mick.

[Eddie]

nm = nanometre

Eddie

[mik]

The reason i ask the question is i am thinking of taking my first steps into dso imaging with a dedicated ccd camera and wasn't sure what narrowband was but what i do know is that is for the more seasoned imager and not for the beginner, yes i can go down the dslr route but photography is my other hobby and a dedicated ccd will be a nice learning curve for me but i don't want to spend a fortune, i dont have auto guiding so will be limited to short exposures of 30-40 seconds and my targets of interest are star clusters so my thoughts are to get the Atik 16ic Mono camera to get me going because i think if i am right this will be more sensative than the colour cam and if i want to get into LRGB at a later date i just have to get filters & wheel i will be useing my zs66 Petzval scope for this but please bare in mind this is just a see if i like it project.

Mick.

[narrowbandpaul]

star clusters dont do well with narrowband filters, you are just blocking out the light

[RichieJarvis]

How does Narrowband do with Galaxies as well?

Thanks,

Richie

[Whippy]

AFAIK Richie, galaxies with star forming reigons in them have areas of Ha which can be imaged. Yfront's recent M33 is a good example: http://stargazerslounge.com/index.php/topic,33569.0.html .

Mik, If you get the Atik camera, you could use that in your newt and use your webcam in the 66 for guiding. All you'd have to do is piggyback the 66 or get a side by side bar and long exposures are yours .

Tony..

[mik]

Thanks Tony but i cant guide on my CG-4 though and that would require a new mount like HEQ5 and i cant afford that also my Newtonian OTA is out on loan

Cheers Mick.

[Whippy]

Of course Mik, I didn't spot your mount in your sig. D'oh!

Tony..