First try with flat, bias, darks in DSS. M13. Whats happened!

[StargeezerTim]

Hi folks,

This is my DSS output from last night. 22 subs 110 secs, 15 darks, flats and bias. iso 1600. Flats taken this morning in daylight with white t shirt over opening, bias also taken this morning.

The vignetting seems to have changed to a gradient. Any ideas what might have happened?

Also, at bottom right is a kind of trail. it was on all the subs. Any ideas what this is?

Hurrumph!   Tim. 

Here is one of the flats. Does this look OK, it's the first time I have tried this!

I should say, the orientation of the camera may have changed very slightly though the session when I changed battery, but I cant see that would have caused this!

[VikkiFord]

Can't really help you there other than to say that that is waayyy better than my first attempt. I've yet to get out and try again. You inspire me, which isn't that difficult since I have M13 firmly in my sights and I'm gunning for that critter. You do, however, show me what I am aiming for and I thank you 

Vikki.

[D4N]

Moving the camera invalidates the flats, they need to be taken with the camera in the exact same position as the lights or you will end up over correcting some areas and not correcting others.

That said unless you turned it 180 degrees I don't think it would have caused that much gradient to happen.

The pattern of the gradient does seem to match the flat so either the flat is wrong or it has been applied incorrectly.

So are the flats just pointing the scope at the sky or did you use an artificial light source? Was the focuser and camera in the exact same position as when you took the light frames?

Bias in the day is a little risky, you need to be really careful about light leaks.

The little streak on the right looks like an internal reflection, do you always get that on light frames or was it specific to that night?

Sent from my iPad using Tapatalk

[StargeezerTim]

Moving the camera invalidates the flats, they need to be taken with the camera in the exact same position as the lights or you will end up over correcting some areas and not correcting others.

That said unless you turned it 180 degrees I don't think it would have caused that much gradient to happen.

The pattern of the gradient does seem to match the flat so either the flat is wrong or it has been applied incorrectly.

So are the flats just pointing the scope at the sky or did you use an artificial light source? Was the focuser and camera in the exact same position as when you took the light frames?

Bias in the day is a little risky, you need to be really careful about light leaks.

The little streak on the right looks like an internal reflection, do you always get that on light frames or was it specific to that night?

Sent from my iPad using Tapatalk

Thanks,

That streak was just on this target, on all the subs. The other targets showed no streaks. Curious...

You know, anything is possible, it could have turned the camera 180 degrees this morning when I took a cover off the setup. I'll see if I can invert the images and see if that makes a difference. The bias were taken the same time as the flats. 

The camera turned a little when I changed the battery during the flats. The T joint thingy that screws into the camera attachment t thingy (sorry for the technical jargon) is either tight or loose, no in between, like it needs a rubber washer or something.

Thanks again. Tim. 

[VikkiFord]

This was my second try (first was only 5 second exposures). 30 x 10s subs, 10 darks. I would call it a pathetic effort but in the short time I had to grab it and not knowing any better I would say (most definitely) that my exposure was a little (lot) short.

As you can see, it's your image that is giving me the clues as to what going wrong with mine. Exposure 

[StargeezerTim]

This was my second try (first was only 5 second exposures). 30 x 10s subs, 10 darks. I would call it a pathetic effort but in the short time I had to grab it and not knowing any better I would say (most definitely) that my exposure was a little (lot) short.

As you can see, it's your image that is giving me the clues as to what going wrong with mine. Exposure 

I reckon that is good for such short exposures. 10 seconds here would get.... nothing! It's all trial and error. The time I spent on Bis and flats was wasted, as I didn't do it properly... but I'll try not to make the same mistakes again!

Tim. 

[VikkiFord]

Hi, Tim. Thanks.

Did you use a DSLR for your photo? I used my ZWO ASI120MC with a 0.5 focal reducer to bring that lovely little M13 within the frame. Without the FR it kinda sprawls and hangs its arms over the sides of the frame. 8" 'scope. If I remember correctly the FR will increase the light hitting the sensor, maybe that's why I got away with 10 seconds. I'm bursting to try again.

I wish I could help you with yours but I already think what you got is great. Aw heck! You even got that little galaxy in there as well. My respect increases 

[StargeezerTim]

Hi, Tim. Thanks.

Did you use a DSLR for your photo? I used my ZWO ASI120MC with a 0.5 focal reducer to bring that lovely little M13 within the frame. Without the FR it kinda sprawls and hangs its arms over the sides of the frame. 8" 'scope. If I remember correctly the FR will increase the light hitting the sensor, maybe that's why I got away with 10 seconds. I'm bursting to try again.

I wish I could help you with yours but I already think what you got is great. Aw heck! You even got that little galaxy in there as well. My respect increases 

Yes Vikki, my trusty 550D. I've processed it (without the defunct flats and bias frames....over processed probably!) but here is what it looks like now. Tim.

[D4N]

Now that's a nice image, I'd be proud of that

You can clearly see NGC6207 bottom left and IC4617 is also showing between the cluster and NGC6207.

Star colour is a little bit weak though, you might be able to boost the saturation a bit to get some more colour.

I don't see much vignetting there, you might not need flats, or did you used dbe to remove it?

Since the little light streak was only this session it will likely be either a local light that causes a reflection in the OTA in that position or possibly light from Vega causing it.

If your using a newt is it flocked?

Sent from my iPad using Tapatalk

[StargeezerTim]

Now that's a nice image, I'd be proud of that

You can clearly see NGC6207 bottom left and IC4617 is also showing between the cluster and NGC6207.

Star colour is a little bit weak though, you might be able to boost the saturation a bit to get some more colour.

I don't see much vignetting there, you might not need flats, or did you used dbe to remove it?

Since the little light streak was only this session it will likely be either a local light that causes a reflection in the OTA in that position or possibly light from Vega causing it.

If your using a newt is it flocked?

Sent from my iPad using Tapatalk

Hi...I cropped it so the vignetting has been cut!. I wondered about another star... I don't really know how near it is but I took photos of vega as well last night, using the same settings, and it produced spectacular crosses. I attach a sub from that. I assumed that if a star wasn't in the field of the scope, it wouldn't have that effect?

I don't know what flocking is, so I assume it isn't! Lol. 

What is dbe?

As far as colour goes, I don't seem to able to get much from the DSS outputs. If I use a single raw file, for instance, I have got some nice colour in stars. Is there a way to modify the DSS output or any other way of accessing colour? I tried using the colour saturation setting in elements. The only colour that responded to the slider was red, and it looked awful!

Cheers, Tim. 

[D4N]

Looks like a lot of coma at the edge of the field anyway so you will probably want to crop that until you get a coma corrector.

The star itself might be off the sensor but it's light can still be entering the tube and you may be seeing a reflection or part of a diffraction spike. Or it could be a terrestrial light.

Flocking is when you line the tube with a non reflective material to prevent reflections messing up your images.

It is normal for the colours to be faded after stacking in DSS, what are you using to process the image?

DBE is dynamic background extraction, it is great for removing gradients.

Sent from my iPad using Tapatalk

[StuartJPP]

To me DSS always seemed to discard almost all colour but it could have been my settings. I use other stacking software these days because of it.

The flats in your first attempt are definitely not flat. Flats are important and can really help if done properly. I prefer to take flats using an artificial light source (a monitor). If using natural light then I use a bright, evenly lit part of the sky using a white T-shirt stretched evenly...don't know why people use twilight/dawn when the light is a natural gradient, probably what happened to your flats. A nice bright day that is evenly illuminated has worked best for me and I basically pointed straight up, though you need to keep your imaging train constant until then.

That is a nice 2nd image by the way...

[StargeezerTim]

To me DSS always seemed to discard almost all colour but it could have been my settings. I use other stacking software these days because of it.

The flats in your first attempt are definitely not flat. Flats are important and can really help if done properly. I prefer to take flats using an artificial light source (a monitor). If using natural light then I use a bright, evenly lit part of the sky using a white T-shirt stretched evenly...don't know why people use twilight/dawn when the light is a natural gradient, probably what happened to your flats. A nice bright day that is evenly illuminated has worked best for me and I basically pointed straight up, though you need to keep your imaging train constant until then.

That is a nice 2nd image by the way...

Hi Stuart,

Thanks..How do you use a monitor? Do you have a setting to have a flat bright display, then lay it on top of the shirt/opening? Sorry... I cant envisage how to do it!

Tim. 

[StargeezerTim]

Looks like a lot of coma at the edge of the field anyway so you will probably want to crop that until you get a coma corrector.

The star itself might be off the sensor but it's light can still be entering the tube and you may be seeing a reflection or part of a diffraction spike. Or it could be a terrestrial light.

Flocking is when you line the tube with a non reflective material to prevent reflections messing up your images.

It is normal for the colours to be faded after stacking in DSS, what are you using to process the image?

DBE is dynamic background extraction, it is great for removing gradients.

Sent from my iPad using Tapatalk

Thanks. A combination of elements, Gimp, paint.net... freebies mainly!

[VikkiFord]

Thank you for the inspiration. I look forward to producing pictures of such quality. Fantastic.

[StuartJPP]

Hi Stuart,

Thanks..How do you use a monitor? Do you have a setting to have a flat bright display, then lay it on top of the shirt/opening? Sorry... I cant envisage how to do it!

Tim. 

Because my aperture is quite small I can use my netbook's monitor with just a plain white display held up flat against the OTA. You can use Paint or whatever to produce a plain white canvas/display. I change the camera settings to take a longish exposure of about 1 second which I do by adjusting ISO to its lowest (leave aperture alone if imaging with a lens). The idea is to get an evenly illuminated image other than the imperfections in your imaging train, choosing a fast shutter speed will capture the monitor refreshing and will not be evenly illuminated. I will also move/rotate the display around a bit to make sure that any potential dark spots in the backlight of the LCD are averaged out.

A flat should look similar to something like this:

[D4N]

Make sure you perform your levels and curves in a program that can process 16 bits per channel. Paint.net and Gimp 2.8 can't do this (Gimp 2.9 can). Not sure about elements. Stretching the image in 8 bits will cause data loss and nasty edges in dark regions.

I use my iPad and a white T-shirt for flats, works great on a frac as I can just point at the zenith and stick it on top. I leave the ISO the same as the lights though and just reduce exposure to get the right illumination.

Sent from my iPad using Tapatalk

[StargeezerTim]

Make sure you perform your levels and curves in a program that can process 16 bits per channel. Paint.net and Gimp 2.8 can't do this (Gimp 2.9 can). Not sure about elements. Stretching the image in 8 bits will cause data loss and nasty edges in dark regions.

I use my iPad and a white T-shirt for flats, works great on a frac as I can just point at the zenith and stick it on top. I leave the ISO the same as the lights though and just reduce exposure to get the right illumination.

Sent from my iPad using Tapatalk

Is Gimp 2.9 stable and OK to use. I read that it might be unstable and that is why they still download 2.8. What programmes can support stretching an image in 8 bits?

Thanks, Tim.

[D4N]

There is plenty of colour in the image by the way, I've done a fast tweak in photos on my iPad, normally I'd use Gimp but there's not much point getting my laptop out as it is limited in what it can do with a JPEG anyway.

Very crude processing but you can see the colour is in the data even in the JPEG file, you just need to tease it out a bit more

Sent from my iPad using Tapatalk

[D4N]

Gimp 2.9 works fine so long as you aren't using any of the experimental features that aren't in 2.8.

What's the worst that can happen? It crashes and you lose your work since the last save, save often and it's not a problem

Photoshop supports higher bit depths but is costly.

I don't know what OS you use so can't suggest anything else.

Sent from my iPad using Tapatalk

[dph1nm]

using a white T-shirt stretched evenly...don't know why people use twilight/dawn when the light is a natural gradient

It is pretty flat if you point near the zenith. And you are, of course, relying on the reflectivity & transmission of your T-Shirt to be independent of incident angle ... !

NigelM

[Freddie]

To me DSS always seemed to discard almost all colour but it could have been my settings. I use other stacking software these days because of it.

The flats in your first attempt are definitely not flat. Flats are important and can really help if done properly. I prefer to take flats using an artificial light source (a monitor). If using natural light then I use a bright, evenly lit part of the sky using a white T-shirt stretched evenly...don't know why people use twilight/dawn when the light is a natural gradient, probably what happened to your flats. A nice bright day that is evenly illuminated has worked best for me and I basically pointed straight up, though you need to keep your imaging train constant until then.

That is a nice 2nd image by the way...

You wouldn't expect a flat to be flat. That's the point. The flat picks out the imperfections of the system and then applies it to the images to make them flat.

[StuartJPP]

You wouldn't expect a flat to be flat. That's the point. The flat picks out the imperfections of the system and then applies it to the images to make them flat.

Sorry meant "Uniform", totally unlike the OP's original "flat".

It is pretty flat if you point near the zenith. And you are, of course, relying on the reflectivity & transmission of your T-Shirt to be independent of incident angle ... !

NigelM

Can't be flat, that's the point but yes, what I said...point it upwards towards zenith...except that if it isn't yet bright enough the tendency might be to point it towards the brighter part of the sky which if it is late dusk/early dawn means pointing towards a gradient.

[piprees]

That's a nice one Tim and you can see that there's lots of info in your image. Keep up the good work. P.