STL11k-Class2 Vertical banding

[sadia]

Recently I tried to do some narrowband imaging on Helix Nebula. Initially when I tried to push the data I found it to be quite noisy. I knew it’s a faint object but for 7hrs of data the noise level was quite unacceptable. But Once I tried to reprocess the data and paid some more attention I found that my STL is giving me some sort of vertical banding on each of my sub exposures in areas where s/n ratio is low which is the case for helix over Focal Length 1120 where except the central area rest is pretty much empty.

Exposures taken was 30mins . Calibrated against 100 bias, 30 dark, 30 flat. After doing few more test following are my finding

1. Vertical banding only appears when s/n ration is low. SII 3nm performs the worst.
2. Object occupying only part of the Chip shows the banding quite easily for rest of the area where s/n is low. Eg. Helix nebula at @FL1120
3. This doesn’t happen for objects which fills up the FOV giving higher s/n ratio across FOV. Eg. Horsehead nebula @ FL 1120 doesn’t show banding.
4. dos not happen for LRGB or I can't see it in LRGB
5. I don't see the this pattern in my master dark/ master flat/ master bias. Dark and Bias have their own pattern so it will be hard to isolate banding in them where Flat will be too bright to notice it.

Initially I thought I was doing something wrong so I retested this issue after making following changes but the issue persisted
1. Keeping STL power cable separate from other cables
2. No dew heater
3. Test was done @ -30C temp


I know some of you are using STL or have used it previously. Have any of you face anything like this? I will post some image tonight to show what I mean

Ok I am back home, and i can attach some pics. It's hard to see in isolated section but quite obvious once you minimize the image size.

[sadia]

I just came accross this post http://astrosurf.com/heidemann/suppression%20des%20barres%20STL11K/Barres%20verticales.htm

This looks very similar. Hmmmm :( i am not getting a very good feeling about this. I have emailed Peter on this

[Moon]

I hope the camera is ok!
Did you try to process without bias frames?
I found my STL bias frame somewhat unrelable and I had cases where they introduced errors into the final image (probably operator error of course).
So I just use the basic technique (darks + flats) without bias frames. You can still get good results without bias frames on the STL.
James

[sadia]

Hi James,
Only flats are bias substracted when creating master Flat.
Light frames were Dark and flat substracted from master dark and master flat.

The issue persisted accross all following scenerio
- No reduction
- Only Dark reduction
- Dark and flat reduction

I will have to go back home tonight to upload few pics and give some additional info on background ADU levels at which this happens.

Regards
sad

[Moon]

Perhaps the problem is in the flats then? Try some new flats created without bias frame.
I do hope you find an easy solution ....

James

[Moon]

By the way, I only go to -20.
What percentage was the cooler to get to -40?

[Alchemy]

Hopefully there is a simple software fix, that can be implemented somewhere to rectify it.

Not overly sure of what a class 2 allows in the way of defects.

As a point of interest as far as consumer affairs goes..... If you bought this new in this country and it has a defect, then it is the RETAILERS responsibility to rectify it, you are not obliged to go to a third party.
I found this out having a problem with a product I bought and the electrics in it were faulty and the retailer sent me to the motor manufacturer.... At the time I was doing work for a QC and I got some free legal advice from which I informed the retailer they were responsible !!!!!

A good reason to use local suppliers and agents.

I hope it sorts out for you.

[sadia]

i only got to -30, was at 70%(late night)-85%(early night).

[gregbradley]

It looks like it may be pattern noise in the chip.

I don't recall ever see anything like that in the 2 STL's I had.

There was an issue with STL11's that got fixed by a firmware update. But that was more a central banding I thought.
I'd check to make sure you have the latest firmware as a starting point.

It sounds like a pedestal issue? That may be able to be adjusted.

As you have done it is standard practice to subtract a flat dark from a flat.

Greg.

[sadia]

I had my first response from SBIG on this which reads as

"HI Raki,

This artifact is associated with the silicon substrate matrix in the CCD, the physical construction of the sensor. Low signal levels, binning, and low temperature exaggerate this effect.
Techniques you may use to minimize this effect.
Capture images 1X1 or 2X2, dither the mount between images, and set the cooling no lower than -20C.
What is the camera serial number?"


To be honest I quite didn't get all the technical details so I have replied him with additional questions. Since my last post I have tried few additinal things suggested by few members and none of the following made a diffrence:


1. change usb cable
2. detach remote guide head
3. I have tried CCDOps as suggested in this email. the process seems to detect the banding but end result looks quite the same!


I will try setting the temp to -20 next to see if that makes any diffrence or not.

[gregbradley]

Haha that's pretty funny.

Mike Sidonio had a Proline 11002 for a while that he ran at -35C typically.

I doubt very much he had any vertical banding but I'll leave it to him to comment.

Generally defects are lessened by increased cooling rather than worsened.

Why cooling harder would exaggerate it sounds odd although I am sure they know their own camera.

The STL11 starts to sing around -35C on those cold nights when you can get it there. Being limited to -20C would mean noisier images.

Should've bought a FLI mate. I'd sell it and get a Microline 11002 if you particularly want that chip. Better still get the 16803.

Greg.

[ptc]

I think SBIG is making things up whole cloth and that they really don't know what they are talking about. I refer you to this document

http://www.kodak.com/ek/US/en/VerticalTimingOptimizatioInterline. htm

see the illustration and tell me if that matches what is observed in the defective STL11002

the FLI 11002 cameras do not exhibit this anomaly and they can run significantly colder than the recommendation from the guys that were sold to the singaporean burnin board manufacturer who now refers to the astronomy business of SBIG as "and Other Industries"....

any wonder they are no longer an independent entity when they can't even read a simple applications note or if they did to tell you a cock and bull story?

[sadia]

I know mate, I was eying a 16803 on astromart even though I can't afford spending on it right now. I had a second STL (much older model) which i sold to one of my mate few months back. I am really interested to give that a go to see what sort of result I get.

Been facing few bad luck with equipment lately. Latest is TEC160 focuser needs a fix as well! Wayne asked me to send it back!

Yep i am depressed.

[sadia]

Thanks for the link Richard.

[sadia]

It does look quite similar :shrug:

[gregbradley]

Bugger. What happened to the focuser?

Greg.

[sadia]

while focusing with the microtouch each refocus slips by about 7-10 step on average. I think this was effecting my V curve as well.

Wayne initially sent me a document on how to fix it by taking off all of nuts and bolts of the FT focuser, but yeh I am not that good with my pliers. So it looks be on its way back for a repair.

Just when I thought my setup is complete its all falling apart :p

[Bassnut]

Well, I dont know if this helps, but I remember haveing banding problems like this on a cam a long time ago and it was due to a faulty power supply smoothing capacitor allowing 60hz ripple to occur. If your game, open the back of the cam and see if any of the larger caps (round al tubes) are bulging at the top. Thats a sure sign of a failed cap. Easy and cheap to replace if your a tech.....

[ptc]

Was your sensor a KAI series, Fred?

what I saw in the image from the original poster is what KAIs do when the camera in which they are used is poorly designed.

I'll wager that this is simply a design problem with the camera. In my assessment their design implementation leaves a lot to be desired
1) USB 2.0 not supported
2) poor cooling
3) dreadfully slow downloads (makes twilght flats with broadband filters very difficult to take a full set in one session
4) high readout noise


they have had the "jailbar pattern" show up in many of their STLs. Kodak has an APP note on an aspect of how the problem is solved, but I can tell you that there's more to the solution than what Kodak disclosed.

I can also tell you that the ML11002 and the PL11002 simply do not have this issue nor do they have the other four issues I cited above. Additionally they aren't limited to "brighter objects" and "warmer operating temperatures". They are just robust: period. No excuses, in other words.

You get what you pay for at best and you never get more than you pay for unless it is trouble.

Caveat Emptor

[Bassnut]

Actually,I remember now, the cap was in the (external) power supply module.

[midnight]

My 40D had vertical banding a few months ago. I did some calculations based on read out time etc and came up with about 100Hz. A full wave rectified AC signal (which is 50Hz) has rippling at 100Hz. This is a characteristic of FW rectification. I noticed this only when I was on the AC adapter for the Canon. For comparison, I went back to battery power and the banding was gone completely. An LC snubber cct was built and has reduced the effects but not eliminated them. The internal camera amplifier must be very prone to such interference.

Darrin...

[ptc]

while that may be true for the 40D let's not confuse your result with a well known anomaly that is common in improperly designed cameras based on the KAI series.

the jailbar patterns are readily observed in KAI based cameras that don't do the clocking correctly as is obvious with the STL11K

Kodak is aware of this anomaly and that is why they put out the application note explaining how to change the clocking

please note that they do not blame it on operating temperature and there is an existence proof that temperature is not to blame: FLI's ML and PL series DO NOT exhibit this problem, but many many STLs do....

it is a weak design in my opinion and I personally think that unless you get one really really cheap they aren't worth wasting your time and money to use when there are cameras using the same sensors that do not have the list of "excuses" one has to accept when using SBIG.

Sorry to bust your bubble if I did that, but sometimes the truth is difficult to accept. but the sooner you do the sooner you can put forth a corrective action plan: ie ASTROMART!!!

:lol:

[Peter Ward]

I would not normally respond to commercial subjects such as this on IIS....but...

Even a white pointer wouldn't swallow this.... Richard have you forgotten an early post you made elsewhere using a FLI 39000 series image that had more vertical bars than Long Bay gaol ????

There is more to this that simply clocking. The vast majority of STL11k's have zero problems.

Raki. I have been in touch with SBIG and will contact you with options when I get back to Sydney later this week.

Rest assured you will be looked after !