Narrowband question

[Daz1969]

Hi all, Can i use a one shot colour cam (DSI Mki) for narrowband imaging? Probably a dumb question but ive asked it anyway:)

Daz.

[steppenwolf]

Yes you can but it won't work as well as a mono camera because of the Bayer Matrix filters. This is an example of an Ha filter in use on a colour CCD:-

[Glider]

As Steppenwolf says, basically I've only done some reading on this myself, but as far as I understand it, yes you can.

I think you can set Envisage to give you 'mono' (see comments below), un dark subtracted, FITs, save all RAW images.

This gives you grey scale FITs files that are not deBayered.

Effectively the raw data from the camera is normally sent to the PC, in software Envisage would deBayer to create your normal Red, Green and Blue files, and then combine them to give you your fourth Light file.

Capturing and saving data as describied above gets round the deBayer and results in 1 fits file per exposure.

Have a look at the DSI Yahoo Groups, there are 2 or 3, this subject is a can of worms, but it is well covered by people that know more than me!!

Escpecially, I got really confused over the mono tick box in Envisage and what it does.

Enjoy.

[Daz1969]

Excellent!! Thanks guys.

Fab pic Steve.

Gilder: Shall have a look at the DSI group(never knew there was one)

Thanks again.

Daz

[SteveL]

I think you can set Envisage to give you 'mono' (see comments below), un dark subtracted, FITs, save all RAW images.

This gives you grey scale FITs files that are not deBayered.

...but that method also gives you three out of the four pixels that contain noise (at least for Hydrogen alpha). Ha is red, so in a OSC CCD, only the red pixels collect any real usable data, the other three (2xgreen, 1xblue) is pretty much empty or just noisy.

Two ways to do this. Stack and process and usual then strip out only the red, or strip out the red channel from your lights/flats/darks/bias, and stack only them. The second option does gives a much sharper image, but as its only one pixel from the RGGB 2x2 matrix, the image sizes are only half as big.

It is possible to get really deep narrowband images with a one shot colour, but you do need to pre-process the files to get the best data. The following were taken with a Starlight Xpress M25C OSC CCD, red channel stripped out to greyscale TIFFs, stacked and processed as a grey image.

[steppenwolf]

That Witch's Broom has come out particularly well, Steve.

[Daz1969]

Stunning ,simply stunning Steve!

[Glider]

SteveL, yes, I'm still getting my own head around what you do with the raw data files when you have them, as I say, it's un deBayered data that isn't even really 'mono' as might be implied by Envisage and its tick boxes. Confusing?

I see your point about the channels of noise, if I debayer the single raw FITs data files and split them into LRGB files using the Meade IP program then the green and blue are indeed pretty much nothing but noise. The 'image' is in the red.

From what you say is it the case that the Red on its own therefore has a better SNR than the Light, since the Light is a software generated composite of the red plus the 2 lots of noise from the green and blue?

Does this follow through if I then shoot with an OIII and then an SII filter.

Use just the Green for one and the Blue for the other?

The reason I got into all of this was just trying to do 'simple' darks and flats with old data sets. Envisage has a tick box to automatically collect darks and when capturing lights to automatically subtract them before saving the data. The darks that Envisage collects, saves and uses are raw. The light files that I've saved are LRGB. If I didn't do the dark removal when the data was collected the raw darks don't match the deBayered lights. Funnies appear in the images.

The way I see it I should have captured raw lights to match the darks if I want to post process, or capture debayered darks just like I captured the lights.

However, the Yahoo forum posts seem to imply the latter still introduces funnies because of the deBayer, and that Envisage still tries to do dark subtraction on raw capture files unless the mono box is ticked.

The same goes for flats and bias.

I'm starting to rant!!!!!!

Is this a function of the DSI and it's software, do I need another camera?!

[SteveL]

Does this follow through if I then shoot with an OIII and then an SII filter.

Use just the Green for one and the Blue for the other?

SII is also red... just another shade of red OIII is a greeny/blue, and can fall between the blue and green filtered area, and sometimes you get it on both green and blue! (depending on the color dye and the falloff as you move away from the central wavelength of the dyes)

IMHO (and your mileage may vary) Envisage is a great starting point, but if you are ranting about it, then its time to move onto a better image capture program that allows you to do want *YOU* want to do, and not let Envisage try to do what it wants to do. It tries to be clever, but it fails when you get to a certain advance stage. You want to take darks when you want, as many as you want, and not let it apply darks before it save your data to the disk. you want seperate files that YOU can calibrate when you want to, how you want to, with whatever program you want to. I lasted about a week on envisage (old version) before I moved on, it was restrictive and limiting, and stopped me doing things how I wanted to do it. The application ay well have changed since then, but it does sound like its still trying to do stuff for you, when in fact you should be doing things yourself.

[Euan]

Ha isn't too bad on a OSC CCD as Steve has shown, here is what I've managed to get with a QHY8. OIII and SII are probably pushing it too much though, I've tried and failed. I think the Bayer Matrix blocks too much of the signal