OSC - What a unexpected disappointment

[swag72]

I recently purchased a OSC in order to compliment my mono. The same chip size as my Atik 314L, the Atik 16HR OSC seemed like a good addition to my imaging setup. After the first night out - How wrong I am.

My thoughts so far are as follows.

1) Bayer matrix - What the hell is all that about? On Artemis capture, the OSC looks like I am looking through a cotton handkerchief. What a mess. Mono is far cleaner to look at in general.

2) 600s sub with an IDAS filter - I had collected some Ha data on IC1396 with my mono camera and so wanted to collect some RGB. I decided that I would try the OSC. With a 600s sub in Maxim, the screen was blank (apart from loads of stars!) - I could see NOTHING, but I knew I was in the right place. I returned to Artemis and after some serious screen stretching I finally saw something very faint indeed. My expectation after using a modded DSLR was that this would be more sensitive than the DSLR - I was very disappointed.

3) So I just wanted to do a quick stack this morning to see whether I had any data at all. Hardly had anything, except a lovely green screen. Thanks to help on here I may have sorted out the bayer pattern prior to stacking, but what the hell's that about? What a complete palaver. I could have stacked 20 mono images in the time it took me to work that out!

4) And to top it all, this was supposed to be a colour CCD right? Well, the pictures suggest that in 21 x 10 min subs, that's over 2 hours worth, there's not much colour at all.

5) In Maxim, it would appear that my amximum ADU is about 55,500. Virtually none of the stars were saturated, yet they are TOTALLY white, no colour in them at all.

So I have rambled on and probably not in the right order. Will I be keeping this OSC? I'm not sure. If anyone ever asks which is better, in my experience now, a total no brainer, mono all the way. This OSC colour lark is utter pants.

So, here's the rather fetching colour of the first stack and after DBE I managed to get it to a more natural colour!!

Then tried to see whether there was any colour there at all. Mmm, a little, but not where I expected it.

So, are my expectations just too high? Or am I being a complete numpty with this OSC? Would welcome your thoughts before this never sees the light of day again.

[DMouse]

Thanks for the heads up! I have been toying with the idea of getting a CCD and now I think OSC is out for sure. I know this means filter wheels and filters but if the final image is better then all is good.

Thanks

[Stephen]

Out of curiosity, did you first nights with a mono camera and processing go so much better?

Just trying to suggest that one nights imaging and one quick stacking session might not be the best criteria for making decisions?

[jgs001]

I'm not really up on OSC CCD capture and processing (only dSLR) and DSS handles all the hard work for me on that... so I can't really help on Maxim... Are you debayering the images ? If I've understood how this all works, then the sensor is mono, and the Bayer matrix that is overlaid is used to determine the RGB channels for the image... (dSLR I think is RGGB per pixel). To get a colour result from the Bayered raw data, you have to debayer the image, which removes the cloth mesh look.

[ollypenrice]

I agree that you see less on the capure screen. I said this in my shoot out for Astronomy Now, but you do get used to it.

There is not much colour in the ET region, the nebulosity being mostly Ha. (HaRGB is probably as good or better than HaLRGB.)

You may, I don't know, be damaging the colour information by debayering before stacking? It's important not to do this. The debayer action must come after the stacking or you do get very thin colour. So it's calibrate-stack-debayer.

Send me a single FITS sub by all means and I'll have a look at it. I have some ET colour data for comparison.

As I also implied in my AN article, finding out what to do with OSC data is, at first, a bit daunting. Once you get the right routine it's a lot easier. The single flat and the lack of aligning are the perks. But is OSC easier? Nope! Not once you're inside the house.

Olly

Edit, I went back to look at my own OSC data on this one. This had 6.6 hours of OSC in the TEC140 at F7. You see, there's not a ton of colour to be had. The Ha makes this talk.

PS, Lucky escape there. I just spotted a typo on FITS. I'd begun with a T by mistake!!

[r3i]

You should've got some colour with the quantity of data you captured. Below is my effort from last year using a QHY8 OSC. If I recall the clouds came over early and I only got 11*5 minutes exposures so the result is rather rough but the colour is there. I used Nebulosity for the capture, stacking and debayering, followed by some fumbling about in CS3. I've found Nebulosity to be nice and easy to use for the initial processing phase.

[JamesF]

Class rant, Sara

James

[Mark72]

I'd say you first calibrate each, then debayer each and then stack them all.

OSC is really a lot less sensitive than mono but adding LPS filter makes it even less so.

This object would really respond very well with mod. DSLR.

Mark

[r3i]

You may, I don't know, be damaging the colour information by debayering before stacking? It's important not to do this. The debayer action must come after the stacking or you do get very thin colour. So it's calibrate-stack-debayer.

The prescribed workflow in the documentation for both Nebulosity and MaximDL is Calibrate - Debayer - Stack.

Never tried it the other way round.

[ianaiken]

Sara - great rant, that's how I felt pretty much when moving from dSLR to OSC

Maybe you could share your data somehow and we could have a play with the processing?

[ollypenrice]

The prescribed workflow in the documentation for both Nebulosity and MaximDL is Calibrate - Debayer - Stack.

Never tried it the other way round.

OK, maybe I'm wrong. (Sorry!) I do the last two in one AA5 pass and so I don't actually know which order it does them in. However, if I ask AA5 to do the entire process in one go the colour is not as strong.

It isn't universally true that OSC is a lot less sensitive. It varies from chip to chip. Mine are not very different from each other. When I did the Astronomy Now shootout I did two 5.5 hour images of M42 and they were remarkably hard to distinguish. The mono is faster but not by a lot. It also depends on the object, in practice. Dietmar Haager's website has some good research on this. I've heard that LP filters hurt OSC but have no experience of them.

Modded DSLRs can give a very strong signal on emission nebulae but you have little control over the Ha by the looks of it. When I imaged this target I wanted, at all costs, to avoid the featurelss sea of pink that you often see. With Ha-OSC I was able to preserve some of the look and colour of the natural RGB version but intensify it. Maybe the chip makers use red filters for the Bayer which exclude a bit more Ha

than we'd like. However, I prefer to have full contol of how the Ha is processed and applied.

So where does this leave Sara? Maybe there's a workflow or software issue. Maybe it's the LP filter?

Olly

[Mark72]

Olly,

You are right. The sea of red with mod. camera. It's after reading your post that I realize it's not the best way.

Sara,

Your subs are very long. You probably do not have a lot of them. May be you can zip them all, provide us with a link and we'll sort the rest out.

Mark

[RobH]

I thought it was calibrate, debayer and then stack....I must say, I've always had to work hard to get colour good and strong.

Must try it your way Olly.

Sara....I had both an atik 16HR, and 16HRC, and used one for luminance and one for colour.

Yes, the OSC is a lot less sensitive than the mono, but I get round this by shooting lots of OSC.

This of course negates, to a certain extent, the fact that you're grabbing colour data quicker withe the OSC than with the mono.

I got rid of the OSC in the end, not because it wasn't sensitive, but because I thought it lacked resolution, although I got hold of an M25C last year, and am experimenting with blending OSC and mono again....it seems to work very well for Deitmar Hagar, so it's darned well going to work for me

[swag72]

Thanks all - Just come back from my Spanish class so have had a read and a smile!

Perhaps I am doing it the wrong way round - I aligned in PI, then debayered then stacked.

Is it my IDAS filter? I hope not, it cost me a bomb!!! I figured that I'd need one as I did with the DSLR and it worked rather well. I am surprised at the lack of colour in the stars as this was something I wanted to try to capture. Maybe there's a point of capturing a bit of short RGB and then adding it in, purely for star colour.

I will gladly put my data into a dropbox folder and you can have a play. I did no processing on it to make it pretty or acceptable, I just wanted to see what amount of data I'd captured.

Is it fair to slate a OSC after one night? Probably not in all honesty and I certainly hope no one will take anything I say with anything more than a pinch of salt. But as I am happy to work with a mono CCD I figured that the move to a OSC would be fairly simple and intuitive, I am using the same capture software afterall, not like moving from the DSLR where that was all alien to me.

Off to load up to drop box and I'll put the link in later!

[Mark72]

Sara,

Are you sure you picked up a correct debayer pattern in PI? It's very tricky there and if it's wrong the color is gone. Can you also add one calibrated but non-debayered sub?

Mark

[swag72]

I'm just uploading all 21 subs as they are out of the camera, there's no calibration data to go with it as I just wanted to see what I had.

[Earl]

T

Perhaps I am doing it the wrong way round - I aligned in PI, then debayered then stacked.

I used this work flow with OSC Calibrate > Debayer > Align > Stack

If you align OSC before Debayering you will mess the colour up.

[swag72]

I think this is the link for the dropbox folder that has all 21 subs in it.

https://www.dropbox.com/sh/e7hnohcmdjp2ysi/cw0pmlxhp9/IC1396%20OSC

[ollypenrice]

I thought it was calibrate, debayer and then stack....I must say, I've always had to work hard to get colour good and strong.

Must try it your way Olly.

No, don't Rob, I think I got it round my neck there!! My workflow is semi automated so I suspect I've mis-remembered what I set up. Earl has it; if you align before debayering you muck up the colour, that's the point.

I use various techniques for intensifying colour without boosting noise. For some reason I seem to be in need of them these days but I really don't know why... I used to drop L onto RGB and that was that.

Olly

[swag72]

Looks like I'll have to retry my process then as I aligned before debayering.

Can't decide now whether to give it another try tonight or put the mono back on!! Decisions!!

[Earl]

Its not to bad, just needs lots more data and calibration

[harry page]

Hi

Yes calibration , debayer and the alighn and stack

If you use the batch processing in Pixinsight it will do this automaticaly

Also pic your targets with a OSC galaxys do not respond as well ( small ones anyway) , but a good nebula works well.

Of course you can use you OSC colour and mix some lum into it

Harry

[ianaiken]

The problem was without calibration data batch processing wouldn't run, so Sara had a go manually. I think I may have said align then debayer in the other thread, which is wrong if I did and my fault Sara!

Edit: No, I said debayer then align :-)

[Lars]

I'm have used an IDAS LPS II filter and a CLS CCD clip-in filter with my Canon 450. The IDAS seems to cut some of the red so I tend to use the CLS filter for red nebulas(the IDAS seems to be a little better for blue objects though).

For the Elephant trunk, the IDAS may not the best choice. If you are not in a very light polluted area with difficult gradients, I think you'd be better without it.

/Lars

[swag72]

Just removed the IDAS, lets see whether that makes much difference!