DSLR Horizontal Banding...?

[AndyUK]

Hiya - I've noticed some horizontal banding on my subs, which even when stacked with Bias frames is still evident in the completed stack...

Here's the result of a DSS stack of 22x600s subs, 40 flats (obviously they're not very good either as there's quite a bad gradient on here - please ignore that for now!) and 20 bias frames. I was expecting that I might be able to get a much better final result out of >3.5hrs data, but as soon as I started trying to stretch the image via curves, these bands start to raise their ugly head. I've tried Noel's actions on them (which temporarily pushes them back down again after 3 iterations) but the noise is still there and I'm always struggling to keep them away...

I've given this DSS output a hefty tweak here in CS5 to show the banding... Has anyone any idea what this might be (and more importantly, how I get rid of it )

[3ggman]

I have been getting a similar issue with the images I have been taking over the last week. (the banding is going vertical in mine, canon 1000D). I am not sure what is causing it but it is causing me great heartache trying to process the images.

The banding is not present in either the flats or darks frames that I have taken over the week so I don't think that its a camera issue.

In PixInsight, when I seperate the RGB channels the banding is only present in the blue channel and both red and green are perfectly clear and very noise free

While conditions have been quite clear there has still been a lot of moisture in the upper atmosphere which I think may be causing it.

I am hoping that if I try to get my total exposure up to and over 5hrs I may be able to process them and not have too much of this showing up.

It really is a pain and I am on the verge of deleting everything I took this week

I know thats not really much help but you are not alone.

[GrantEb]

Are you subtracting darks?

Also, remember, it's getting warmer now, so noise will be a lot more evident.

HTH

[Deneb]

Send the stack over , I would like see if i can replicate that amount of banding... Have you just tried stacking just the subs to see if you get the same result.. & what ISO was you using ?

Nadeem.

[AndyUK]

Cheers Johnny - It's sort of comforting to know I'm not alone, but obviously it doesn't get either of us to the bottom of it... :rolleyes:However, I have to admit that I was very suspicious about the conditions last night although I'm not sure why it would cause this banding(?). I was of the same thought as you though - Capture double the subs and see if that salvages anything, or bin them .

Grant - Admittedly this stack didn't include darks (just bias and poor flats) but I did try a subsequent stack with some darks and also some older flats from another night (same optical light path) and the result was slightly better... but the banding was still there which prevented me from stretching the image as much as I would have liked / I would have thought should have been possible with 3.5hrs of subs - I didn't notice these bands a month ago .

Nadeem - Sure - I'll whack them onto dropbox - I'll let you know when they're there (it'll take a while!). I haven't tried stacking just the subs, but will give it a whirl - I was using ISO 800 (which has pretty much been my norm to date)

[AndyUK]

Here's a stack of just 14 of the best subs... (again highly tweaked to show bands). It could be me, but are they not QUITE as bad(?).

Perhaps I'm also getting confused about subtracting bias signal - As I understand it, you EITHER subtract bias frames OR flat darks, but not both(?). As my flats have such a short exposure (and therefore surely noise would be severely limited?), I usually subtract only bias...

(Nadeem - Still uploading - Will let you know )

[GrantEb]

Apparently, you don't need to take bias frames with a DSLR? don't take my word for it though.

[GrantEb]

Do you know what the temp was when you were shooting? You should be able to extract it from the RAW file, I'm not sure how though.

[AndyUK]

Hi Grant - I can do that (KUSO EXIF Viewer) - Hmmm - At the start of the run the in-camera temp was 8C, but by the end it was 18C... (Wow - I had no idea the chip got that hot - The battery was external and the liveview was switched off ).

Obviously that wouldn't have helped, but would this create horizontal band artifacts?

NOTE: It would also explain why the 600s stock darks I took in my fridge came out at 20C... and goes to show the value of taking darks before and after imaging...

[GrantEb]

I think it might.

Open a single 600s dark in photoshop, and stretch it, see what you get?

[AndyUK]

Hi Grant I think you may well have something here...

I opened up 2 frames - One at the start of the run (6C) and one from the end (18C). For starters, the noise difference across the frames was very evident, and I wasn't able to see the banding in the first frame (maybe my eyes?) but in the last, I think I could faintly see it.

I've also just checked the temp fluctuation from start/finish of a few recent 600s sessions from the last few weeks (but none of these displayed this problem?):

2-3

-2 to +4

4-4

4-7

3-12

3-8

2-8

7-12

10-16

12-16

Maybe the temps involved in this session crossed some kind of threshold? I had a particular close look at the 3-12 session to compare and also the 2 latter ones, but still no signs of the banding I got here...

[GrantEb]

I think maybe the banding is coming out of either your dark, flats or bias frames.

There doesn't seem to be much banding in the light frames

[GrantEb]

Sorry, I am rambling on here.

Try this:

Take flat darks, ie dark frames taken at same exposure and iso as your falt frames, temp shouldn't matter.

Don't use bias frames

Stack all your lights, flats and dark flats in DSS.

HTH

[focaldepth]

Just a thought.

Are you running on batteries, are they healthy and fully charged?

Are you running on a mains adaptor, is it decent quality?

[AndyUK]

Hi Focaldepth - I use one of these- I don't think this is power related as again, as I wouldn have thought I'd seen this effect in ALL my images...(?)

Cheers Grant - Please, ramble away...! I did have a go at processing these again with darks (admittedly stock frames at 20C), flats and flat darks and it did come out a bit better... but the bands were still there.

I know the general recommended practise is to take darks before/after the session, and it's recommended to take some frames to begin with to get the camera at "operating" temp which would, I assume, be dependent upon ambient temp... I'll have to investigate from the subs I have how long that takes and THEN take the "before" set (and then allow the 90mins+ after for the "afters"...)

(Nadeem - Still not completed the uploads... will let you know )

[focaldepth]

I am a new-commer to astro photography so I will let the experts advise you on stacking.

Electronics however is a different matter.

If the stackers can't help then consider the following:

Noise on the camera power supply.

Noise coupled from the mount motors.

Noise on the light source used for the light frames.

It is just that it reminds me of a TV, computer monitor or CCTV camera just before the power supply fails.

[Starflyer]

I used to be plagued by horizontal banding with my 450D. I didn't spend time trying to to figure out where it comes from. I cool my camera now and haven't had a single banding problem since I started doing this.

[AndyUK]

Hi Focalpath - Yes - I see what you mean... it does look like "interference" doesn't it...?

However, I've now processed this stack of subs with 2 different sets of flats (one taken at the time via white t-shirt, the other "stock" via laptop screen) and also with and without darks (admittedly stock from the fridge, but although fridge temp is about 3C (ish), the exif on the files notes them as being 20C due to camera chip heating up). I've also tried it with 2 sets of bias (one set taken a few months ago, the other set taken yesterday) and 2 sets of flat darks (both sets taken at time of flats).

Maybe I need to set a longer duration between each image to allow the chip time to cool down - At the moment I've been setting 60s on the timer, but maybe I should make this 120s?

Starflyer - Oh yes... Peltier cooling... how I would LOVE to do that! Unfortunately I can't spil DYI... I looked that up from another thread and also noted something called a peltier "finger" whereby (I think), a small sheet of metal is basically "poked" into the camera providing cooling almost directly to the back of the chip. However, as my 40D is my only DSLR, I couldn't justify that particular mod (even if I could find someone to do it for me), but casing it in a pelter-cooled box would be a vast improvement. Again though, sadly I'd need someone who could make it for me (maybe Andy Ellis @ astronomiser knows of someone... I'll drop him a line... It MUST be cheaper than a CCD...!

[3ggman]

Hey Andy,

I was having very similar problems with my camera and to be honest I think it is unavoidable when doing long exposures unless you can cool the camera (i am looking into building a peltier cooling box now). Last night I did just what you suggested and gave the camera a much longer cooling time between shots (4mins) while it didnt kill the problem it certianly did reduce a lot more, even longer would probably reduce it even more.

I think the cooling box is the only way to really get rid of it altogether.

[dmahon]

I only have 15s between my subs, using APT on a Canon 450D with an internal battery. I have used 600s ISO400, 300s ISO400, 30s ISO400, 300s ISO800, and 30s ISO800. None of them have shown the banding that you have.

Are you using the frames direct from the cameras SSD card or are you downloading them to your computer over the USB cable?

[AndyUK]

Are you using the frames direct from the cameras SSD card or are you downloading them to your computer over the USB cable?

I'm capturing on the card (Compact Flash) and then at the end of the session, I remove it from the camera and put it into a USB card reader... (Would that make a difference?)

I haven't tried using APT on an imaging night yet, but it is something I intend trying out one day - I've been resisting "laptop-attached" imaging to date (but I guess it's only a matter of time...)

I think the cooling box is the only way to really get rid of it altogether

I have a feeling that you may well be right - It was okay in cooler temps but when the weather gets warmer, if this is the norm for 600s exposures (plus then having to wait 4-5mins between frames) this will seriously cut down the number of subs I was hoping to be able to get... As I'm cr*p at DIY (and wouldn't have a hope of putting a peltier box together), I guess I'll just have to resign myself to all objects being 3 or 4 night "projects" or spending the summer using the mak 180 on the Moon and Saturn (and maybe Mars later). Ho hum...

[MikeWilson]

Hi Andy,

I've had banding before but nothing like this.

I only use 10 seconds inbetween frames, although given your experience I may double it to 20 seconds for the moment.

Has anyone developed a cooling mod that doesn't require a cool box or is a full-on-peltier or long delays inbetween subs the only way forward?

Best,

Mike

[Earl]

I have had a quick play, i hope you dont mind

[AndyUK]

Hi Mike - Yes, I'm quite depressed about this... . It's rather going to stop me in my tracks... :). (Have you seen the peltier finger mod - Hyper-Modification of Canon 450D (XSi) Cameras and here: Canon XSi 450D Peltier Cooler Modification | Rankinstudio - I thought filter removal / modification was scary, but that looks REALLY SCARY to me .

Hi Earl - I have to say, I'm REALLY impressed that you produced that out of a raw stack that I'd WELL over tweaked! Here's the image I did finally manage to get out of it without the banding coming through (much)

[Earl]

the photoshop tools i bought do a good job on banding