Flats and correcting significant Blobs! (Astroart?)

[Notty]

Hello again, so having got guiding working I took 4x5 minute subs of the Pinwheel Galaxy, along with darks, bias and for the first time flats.  Flats are going to be my next nemesis I think.  I have got some nasty looking blobs on my imaging train somewhere!  I know you aren't supposed to but beforehand yesterday got the baader fluid and microfibre cloth and cleaned the objective of the ED-80 which was where I assumed it was.  Imaging camera is EOS1100D, recently modded and used for the first time this week for AP. Below is my stacked image, but before flats applied, showing my nasty blobs particularly one right in the middle ruining the galaxy:

Tried a few methods of flats, this was 25 shots camera set to AV pointed right against a white laptop screen, clearly showing the culprits:

This is the image once flats applied.  I used Astroart, which is new to me, but there aren't too many ways to misinterpret the settings so I'm confident it's done the right thing.  As an experiment I did process the lot in DSS as well, but the final image was a vertical slither of the total so I'm not sure what's going on there!

Not sure if it's an Astroart thing but i had to apply the flat manually once it had created the master flat, even though I thrrew all the files at it againin preprocessing.  Should a decent set of flats be able to sort them out do you think?  Also now I've established they aren't on the scope objective, would I be right they must be on the sensor of the camera?  Like I say it's only just been modded (professionally) so I would have assumed the sensor should be clean?  Is there anywhere else it might be possible to get contamination like this (eg inside the objective?)

I know I've got 2 potential issues here, the flats themselves and where the bunnies live, but anyone got any ideas?

Thanks!

[Davey-T]

Spots like that are usually close to the sensor, dust on the objective won't show up in an image, don't like the look of the very black one, image is very stretched which magnifies the dust bunnies.

Dave

[cathouse42]

Canon sensor cleaning

http://cpn.canon-europe.com/content/education/infobank/capturing_the_image/sensor_cleaning.do  (3 pages).

[Toxophilus]

I had something similar on my Atik314. It was recommended to send it in for a clean and service. Before sending it away I got a magnifying glass and had a look at the sensor and sure enough you could see a speck of dust on the sensor. When the camera came back to me the dust bunny had gone and all was well

[cathouse42]

I had something similar on my Atik314. It was recommended to send it in for a clean and service. Before sending it away I got a magnifying glass and had a look at the sensor and sure enough you could see a speck of dust on the sensor. When the camera came back to me the dust bunny had gone and all was well

How much did they charge to use a rocket blower?

http://www.amazon.co.uk/s/ref=nb_sb_noss_2?url=search-alias%3Daps&field-keywords=rocket+blower

[Toxophilus]

How much did they charge to use a rocket blower?

http://www.amazon.co.uk/s/ref=nb_sb_noss_2?url=search-alias%3Daps&field-keywords=rocket+blower

That was the killer..

I was quoted €30 + VAT for a clean and service. Not bad I thought, and then they use about the most expensive courier they could so the whole thing cost €72 in the end and took 3 weeks! I was not a happy bunny, but fortunately during that time we had a lot of bad weather so I didn't need the camera. Still it was better than doing it myself and risking damaging a £900+ camera.

[cathouse42]

That was the killer..

I was quoted €30 + VAT for a clean and service. Not bad I thought, and then they use about the most expensive courier they could so the whole thing cost €72 in the end and took 3 weeks! I was not a happy bunny, but fortunately during that time we had a lot of bad weather so I didn't need the camera. Still it was better than doing it myself and risking damaging a £900+ camera.

Well, unless you are using the Rocket Blower as a Javelin, there is low risk.

Hold the camera upside down and give it a puff 

[Thalestris24]

Hi

You should be taking flats through your imaging train and without changing the focus you used during imaging, or changing anything else really! (Apart from exposure). Anything you see in terms of blobs etc that are in focus, will be at or near the sensor. Anything out of focus will be further away but anything on the objective will be so far away from focus that it won't be visible.

Hth

Louise

[Notty]

Thank you all, yes I took all the flats right after the lights during last nights session.  I've blown my sensor with my airbrush, but also had a look down my scope and saw some significant looking debris on the field flattener at the other end, so I am hoping I have done away with the majority of the crud.  If I get a chance to capture anything tonight and it's still there then I guess the camera is definitely the culprit and needs more attention.

Assuming I did my flats correctly though, would it be reasonable for them to clear massive blobs like that anyway or is it asking a bit much?  For example one is clearly obliterating one of the tips of the galaxy arm.

[Thalestris24]

Thank you all, yes I took all the flats right after the lights during last nights session.  I've blown my sensor with my airbrush, but also had a look down my scope and saw some significant looking debris on the field flattener at the other end, so I am hoping I have done away with the majority of the crud.  If I get a chance to capture anything tonight and it's still there then I guess the camera is definitely the culprit and needs more attention.

Assuming I did my flats correctly though, would it be reasonable for them to clear massive blobs like that anyway or is it asking a bit much?  For example one is clearly obliterating one of the tips of the galaxy arm.

Hi Notty

Um, the flat you posted looks much too bright to me. The flats' histogram peak should be left of centre. You should be able to see the vignetting gradient on the flat. But the vignetting is still apparent after you've applied flats to your subs so something not right... If using DSS, make sure you have the latest version - 3.3.4

Louise

[Notty]

Thanks Louise I agree with everything you say, but I just don't know what I've done wrong! I was told however that you don't need to worry about the histogram with a DSLR as setting it to AV gives the correct exposure. Is this not correct? Or are we saying I need to do some manipulating to the flat before applying it?

[Thalestris24]

Hiya

No, you don't want to manipulate the flat. I was thinking, though, that a jpeg of a flat will appear a lot brighter than the raw version. Nevertheless, stacking the raw flats with the lights in dss should remove blemishes and gradients including vignetting. Using AV should be ok unless you have a modded camera. I do mine via a plan in APT. You could check the exif data to see what your flats' exposure was. From memory, I think my flats taken at F5 are about 1/40 sec but obviously will vary according to you laptop screen brightness and focal ratio.

Louise

[Notty]

Ah, that's it then my DSLR is modded, so I need to have another go then. Histogram peak left of centre? Hopefully I can work out how to do that.

It's tedious waiting for the sun to set this time of year. 3 nights in a row of clear skies; I'm absolutely shattered but don't know when my next opportunity will be. Have fun everyone!

[Thalestris24]

Ah, that's it then my DSLR is modded, so I need to have another go then. Histogram peak left of centre? Hopefully I can work out how to do that.

It's tedious waiting for the sun to set this time of year. 3 nights in a row of clear skies; I'm absolutely shattered but don't know when my next opportunity will be. Have fun everyone!

Yeah, my modded 1100d overexposes if I try and use AV. So have to set up a manual exposure flat plan. You probably need to do some trial and error to get the exposure right but once you've ascertained a good exposure then you can just use the same plan Yep, histogram peak about 40% from left, I think.

Louise

[Toxophilus]

Well, unless you are using the Rocket Blower as a Javelin, there is low risk.

Hold the camera upside down and give it a puff 

Unfortunately it is not easy as that as the sensor is behind glass on the Atik 314L+ and so you have to take the camera apart, something that I'm not too keen on doing.

[Davey-T]

Hi Notty, should have stuck to Solar you can take flats in the daytime to see what effect your cleaning efforts have had.

There's an equation for working out the location of dust bunnies in the imaging train, you input your parameters and it works out their distance from the sensor depending on how many pixels they cover.

I'll dig it out if you're into maths, involves a few log' calc's

Dave

[Notty]

Cheers Dave, but I think I blasted 'em out with my airbrush yestarday!  Images much cleaner 

[Notty]

Louise, can I ask with regard to setting the exposure on my modded 1100d, my histogram is basically a vertical column about 6-7% of the width of the graph.  I can move it left to right fine with the shutter speed, but the it always runs all the way to the top of the graph.  I don't seem to be able to reduce the peak any.  I'm guessing this means I'm overexposing it?  I note that if I turn down the lappy screen brightness the flickering is pretty awful and I'm guessing would corrupt the flats, or would it average out over enough?  Thanks!

[Notty]

Forgot to fix this pic of the graph... 

This isn't taken when I actually took the flats, so I didn't have it set to bulb, it's just to illustrate what the histogram looked like

[Thalestris24]

Forgot to fix this pic of the graph... 

This isn't taken when I actually took the flats, so I didn't have it set to bulb, it's just to illustrate what the histogram looked like

IMG_0552.JPG

Hiya

You can probably ignore the peak height - it just needs to be left of centre. Strictly speaking, as I understand it, the flat pixels should measure 25000-30000 ADU (analogue to digital units) but I don't know how to directly measure that value with a dslr image. For a 16 bit image, the max value is 65535 (= saturation) and 25000/65535 = .38. So, 0.38 - 0.40 of the saturation point. I figure you can just convert that to a percentage and apply it to any histogram where the leftmost point is 0 and the rightmost is 100%. So just left of 50% of the way along. With APT, I usually go by the grey 'luminance' histogram peak (on the bottom of the histogram display). I suppose the same applies to the above on-camera histogram which I presume represents luminance.

If anyone knows different, please advise!

Louise

[cathouse42]

Unfortunately it is not easy as that as the sensor is behind glass on the Atik 314L+ and so you have to take the camera apart, something that I'm not too keen on doing.

So, if the sensor is behind glass, how did the dust get in there?

Perhaps it should be sealed?