Horizontal Banding - Color Horizon - Still a problem!

[bwa]

Version 4.2.2.1 of the core software has partially resolved my horizontal banding problem, i.e.: the banding is less severe, but it is still present when one does a strong image stretch!

The attached image is of the Leo Trio of galaxies, M66+: 21x60 sec., Med Gain setting, 4x4 binning (Normal), aligned and stacked from captured subs.  The banding gets considerably worse as the exposure time goes down and is present in all bin modes and all gain settings.  And it doesn't make any difference if an image is live aligned & stacked or processed from subs.

bwa

[Edwardnut]

I want to know Can I fix this? Or wait for it to return to normal.

[bwa]

Spent another night imaging with exactly the same setup as previously; no banding and no idea why!?  It is rather frustrating not knowing how everything is going to work from one night to the next.

[bwa]

I think I've found the problem with horizontal banding, at least for my color Horizon; others might be different.

The banding seems to tie to the Offset used.  The fault Gain/Offset values, per ATIK support are:
...Low: Gain 1, Offset 17
...Med: Gain 6, Offset 29
...High: Gain 30, Offset 139

When I enter the above values in the Custom selection I get distinctly different results than when I select Low, Med or High.  It almost appears the Offset being applied by the Low, Med and High is not being used or is lower than stated; thus, giving rise to horizontal banding of various degrees. 

This is an instance where light pollution is actually a benefit by setting an artificial Offset; thus, imagers with lots of light pollution will probably have no problem with horizontal banding.  However, imagers in a dark site may see banding!

Until this problem is resolved by ATIK Support I would suggest using the Custom selection and setting your own Gain and Offset values.  You can use the ones above or select them from the graph (PDF file) I've attached.

Enjoy!

bwa

[dpaul]

Hi BWA,

I believe banding is quite common with various types of ''cooled'' CMOS cameras (not just Atik). I'm no expert but on some nights I get it and others less so (I'm mainly using the Horizon mono). Maybe its a certain combination of cooling level, gain, etc. I wanted to share details of an amazing tool in Pixinsight - I do all my processing with this software.  The method is called ''canon banding removal'' and can be found as one of the ''script tools''.


Attached is your original frame that you posted in April and in only a few seconds this tool removed all evidence.
I use is much of the time so no issues at all with banding. There are many other reasons for using Pixinsight

Hope this helps.

Dave

[Astrogate]

I guess this is okay if you are using the Horizon for imaging along with post processing but why I held off from buying one of these cameras is it EAA usage. This camera what developed also to be use with the Infinity software for EAA or near real-time observing and if the banding is this bad then I do not see this as a successful EAA device unfortunately. Both my ZWO ASI 183 mono and ASI 1600 color has no banding at all!!

Chris

[bwa]

I guess this is okay if you are using the Horizon for imaging along with post processing but why I held off from buying one of these cameras is it EAA usage. This camera what developed also to be use with the Infinity software for EAA or near real-time observing and if the banding is this bad then I do not see this as a successful EAA device unfortunately. Both my ZWO ASI 183 mono and ASI 1600 color has no banding at all!!

Chris

If you saw my previous post, there is no banding if the Gain/Offset are set properly.  As I said, set the Gain/Offset in Custom.  Using the Low, Med or High defaults can result in banding.  ATIK Support has been notified of the problem BUT has not responded!!

And yes, PixInsight does has a script to reduce/eliminate the banding BUT I would prefer not to have the banding in the first place...

bwa

[Astrogate]

Sorry Brian but I am talking about EAA and not imaging. I am sure Atik will resolve the gain/offset issues so one does not have to manually select the proper matching.

Chris

[CraigG]

bwa-

Any new updates on the banding issue from either your experience or from Atik? I'm also noticing horizontal banding in my images from a Horizon OSC. I've attached a way overexposed image to highlight what I'm seeing with regular stretches in most images with light or nebulous areas. As I am trying to stay as close to simple EAA as possible, I'm not interested in using post processing tools to remove them! This image is 2x2 binned with the stock Medium gain setting (not set in Custom yet as you suggested).

Thanks...
Craig

[dpaul]

Hi Craig,

With the Horizon mono camera how many frames are you taking for each Red, Green, Blue and Luminance filters?
Also wondering if you are taking Bias and Dark frames to calibrate each of the light frames?
Usually I take a minumum of 12 frames per filter and ideally 20, this makes a big difference to the signal/noise ratio of the final integrated image. This won't necessarily get rid of all the banding but proably reduces it.

As per my previous note, horizontally banding seems to be a common feature of CMOS cameras from various manuafucturers.
I haven't played around with the gain setting that BWA suggests (which may indeed be useful) but because they can easily be removed in post processing, its a non-issue for me.

Over the last 18 months I've changed from hating any type of imaging (visual only was my interest). Then I got into semi-real time video and soon realised there was a lot more data in the image if it was post processed. A year ago I then bought an Atik Horizon mono camera and started playing around with post processing. Even simple processing with Pixinsight gives amazing results. I now get equal satisfaction out of the processing as I do the capturing - never thought that would happen.


Just wanted to share my experience as someone that was 100% against post processing but now now my interest in deep sky objects is way higher.

Regards

David

[CraigG]

Hi David!

Thanks so much for sharing your journey to this point. It really helps! I guess we should never say that we're 100% committed to enjoying this hobby in only one way with all of the options that are available. Since my last post, I've read comments from other imagers suggesting using the PI tool  to remove the banding just as you suggested. Although I'm waiting to hear back from Atik, I'm already moving toward...however reluctantly...to the post-processing camp. I was hoping that the promise of EAA would satisfy my needs for the foreseeable future after a few years of less than satisfying results from processing images from my SBIG 8300c.

To answer your questions, I have been using my Horizon OSC strictly stock with no filters or adding bias/dark frames. The Horizon was a next step up from my Infinity OSC and I'd hoped to use it in the same way.

Anyway, thanks again for your counsel and I may be picking your brain a bit when I start working with PI.

Out of curiosity, would you be able to use the Canon tool on my attached image of Pleiades so I can see the improvement? I would be grateful!

Best wishes,

Craig

[dpaul]

Hi Craig,

The debanding tool is meant to be used on the linear image before histogram stretching. I have however tried on your image and it made 'some' difference (see attached). The tool can add lines if over-used so its a fine balance.

David

[dpaul]

Sorry Craig

Forgot the attachment.

Here it is

David

[bwa]

Hi David!

Thanks so much for sharing your journey to this point. It really helps! I guess we should never say that we're 100% committed to enjoying this hobby in only one way with all of the options that are available. Since my last post, I've read comments from other imagers suggesting using the PI tool  to remove the banding just as you suggested. Although I'm waiting to hear back from Atik, I'm already moving toward...however reluctantly...to the post-processing camp. I was hoping that the promise of EAA would satisfy my needs for the foreseeable future after a few years of less than satisfying results from processing images from my SBIG 8300c.

To answer your questions, I have been using my Horizon OSC strictly stock with no filters or adding bias/dark frames. The Horizon was a next step up from my Infinity OSC and I'd hoped to use it in the same way.

Anyway, thanks again for your counsel and I may be picking your brain a bit when I start working with PI.

Out of curiosity, would you be able to use the Canon tool on my attached image of Pleiades so I can see the improvement? I would be grateful!

Best wishes,

Craig

Craig,
You stated: "As I am trying to stay as close to simple EAA as possible".  This is NOT possible with the Horizon camera / Infinity software at its current state of development, at least from my experience.  In fact it takes additional effort to get rid of the horizontal banding from Horizon images regardless of the postprocessing you undertake!!  And yes, the Canon script in PixInsight does work most of the time (not always); however, a good astronomy camera should not require this additional step.

I've downloaded and installed the most recent version of the Infinity software in hopes it did something WRT the banding problem; no improvement, sadly!

Some people have said this is a problem with all CMOS sensors.  Personally I think it is solely an ATIK problem.  As I've mentioned previously I would have purchased a mono Horizon if the OSC Horizon banding had been resolved.  It wasn't and I purchased a QHY 163M which uses the same CMOS sensor as the Horizon.  I can stretch subs off the QHY 163M to the extreme and get NO banding.  I also image with a fellow using a ASI1600MM Pro camera which has the same Panasonic sensor and have never seen banding in his images, some of which I've processed.  I'm quite sure the banding problem comes right back to what ATIK is doing in hardware/software with the Horizon cameras!?  I suspect ATIK support has not addressed the problem because they have no cure without redesigning the Horizon camera...

The poor man's resolution to the banding problem is 1) don't shoot subs requiring a stretch, 2) use the Canon script in PixInsight to reduce/eliminate the problem.  Not ideal, but at least a partial solution.

bwa

[CraigG]

David...thanks so much for processing that image!

I do see some improvement and I understand that the PI tool is better utilized before stretching. I still have a service ticket pending with Atik so we'll see what they say about the banding. Maybe they are working on something they can fix in Infinity? In the meantime, I'm reading up on PI before taking the plunge!

Best wishes,
Craig