Is this a bad column?

[Jessun]

Hello all,

This is a 10 second dark with a newly delivered SXVR-M25C. Specs state less than 50 hot pixels in 10 seconds which this one passes - just.

But this annoying green line? Is that in fact a 'bad column'? Specs state there should be none for a grade 1 CCD.

The line appears regardless of exposure time...

CCD temp is -10°

Thanks in advance

[andrew s]

Yes it looks that way to me. Contact Terry at Starlight Xpress and I am sure he will help you. Andrew

[Jessun]

Thanks Andrew, I sent the question to the guys, and know they'll be quick to answer as they helped me out before. It's weekend now close to Christmas but it wouldn't surprise me if I had an answer in the morning.

It's pretty clouded here now so I can't do much more yet...

[Bukko]

Hi, Jessun,

that looks just like a defect I got on my SBIG ST8300. Mine was lower and about a third in from the left.

Mine developed at about 6 months old.

I emailed a cropped section of the image to SMIG and they checked it.

It turned out it was a single hot pixel, about 75% on. The column was an artefact from this single pixel (Note the lines does not go to the top)

So no warranty.

I asked if it was anything I had done - cooling, heating up, knocked it ? They had no answer to explain it, just one of those things.

Good luck with SX, maybe they will have a different view of this than SBIG did.

Gordon.

[Jessun]

Thanks Gordon, yeah wait and see. ATIK for example clearly state that they use grade two CCDs unless specified and that the odd bad column can be present BUT that they assume that darks will be used anyways. SX say no bad columns and that no darks necessary for 'most deep sky objects'. As for SBIG... well I saw 9000 dollars for a full frame OSC and quietly left that website...

I do not know for sure if this is a bad column per definition or something else - I am no CCD engineer. My line goes all the way, only dotted and weaker 'up' so takes a bit of effort to see it. Perhaps it falls under single pixel error too.

I have run some trials with Nebulosity's Bad Pixel Mapping which is an very interesting alternative to darks but this does not cure it in the result. Darks do and I have played with flats as well and that seems to cure it too. (Still puzzled by that btw considering I did a 3*3 median blur on the flat and Nebulosity does another blur on top of that...)

I got the SX based on the 'no darks required'.

Anyways, the jury is out on this and I don't mean to discredit SX in any way with this thread. I just seek to get the facts right. Apart from that one gremlin, the camera is an absolute joy to work with - Nebulosity does the best job so far as AstroArt is a hit and miss still if it downloads a useful image or not. That's another story though...

Cheers

Edited December 18, 2011 by Jessun

[Bukko]

Hi, Jesper,

I think SX mean that the electronic noise os so low that darks are not needed. My SBIG uses the Kodak CCD and the noise is a lot higher, so darks have been required from new. I guess this will be true for all the 8300 chip based cameras.

My saviour is the integraded shutter, allowing me to automate the darks and have the software do automatic dark subtractions.

I no longer really notice the problem.

I got a Lodestar last week for guiding and the noise from the Sony CCD is certainly a lot lower than my SBIG. I am sure SX will be pretty straight with you, but will probably want to investigate the hot pixel/column. Possibly their definition of a column defet is different to SBIG's.

Yes, I have also seen the $9k cameras, they will have to wait 'till I retire and cash in my pension plans, I think.

Good luck with SX.

Gordon.

[ollypenrice]

I don't like the sound of SBIG's answer and they have just moved a long way down my wish list. A very long way down. I don't like firms like that.

I get a much shorter line on one of my Atik 4000s and bizarrely it sometimes goes with darks and stacking and sometimes doesn't. I think that dither guiding and drizzle stacking would make it disappear, though. In any event, when I do get it in a final stack it is dead easy to deal with in Ps.

Keep us posted on the SX response. There seem to be quite a lot of issues with SX stuff but the reputation for helpful service is always sustained.

Olly

[steppenwolf]

I have a column fault like this on my my SXVF-M25C. It developed a few weeks after purchase and I was advised that it was a high energy (gamma ray) hit and that this was not covered under warranty. I was also told that these can happen at any time including before use by the purchaser.

I was really not impressed with their response (although I personally hold SX in very high regard) but it was backed up by Ian King who I also trust so I decided it was 'the way things are'. A BIAS frame essentially resolves the issue very well provided you don't have to overstretch the image too far - something I avoid anyway as noise is all that you get at the level required to overcome the good work carried out by the BIAS frame! As a matter of interest, I never take darks with this CCD camera.

[Jessun]

Thanks for the info Steve! As of yet I have heard nothing from SX, so I am not sure of a) what this is - as specified by SX - and if it is covered, so I'll leave it there for now.

I can always return the camera anyways to the dealer if I wish - German consumer law I'm guessing...

I was hoping not to have to take darks so that is good news. Darks in the wrong hands just make matters worse and I need to develop my skills more...

At least a BIAS is quick! It would be great if it was such a simple cure for this CCD too! I'll give it a crack asap.

On another note what size is your book? I am trying to work out some flat packages under the Xmas tree...

[steppenwolf]

On another note what size is your book? I am trying to work out some flat packages under the Xmas tree...

I couldn't possibly tell you that and spoil any possible Christmas surprise (which comes in one of two flavours):-

Flavour 1: The pleasant surprise of discovering that it is my book or

Flavour 2: The equally pleasant surprise of knowing that someone has bought you a better book than mine .....

[harry page]

Hi

A bad colum would be one that is dead i.e fully saturated all the time , is this the case ?

I have several smaller ones on my chip that have never caused me any problems and as said a good bias sorts out most problems , I use a master made out of 100 subs

Also never used a dark with this camera

Harry

[Tim]

I have a similar line to that on my QHY8. I dont take darks for that camera either and you have to look really close to see the effect of the line. Even then it is easy to get rid of in processing.

Keep us updated with the response you get though

[Jessun]

Flavour 1 confirmed Steve! We're a day early on the continent! Excellent!

Harry, this pixel and associated line is not completely hot. It seems to feed a bit on light, but visible to certain degrees in Lights, Darks and Biases. After tinkering around with a few exposures before the clouds drifted in it is pretty clear - as you guys pointed out - that it does not pose a problem. As Olly points out this will even easily dither away with a stack.

Not heard from SX, but if they want to take this camera from me they will have to send a SWAT team to get it! My initial play around is very promising.

I ran the Lodestar through its first run too and it too performed really well. The only problem was to bring it in focus during daytime to get all spacers right. I was down to 0.0001 second exposure in Nebulosity (can it do that??) with centre dust cap off only - and still it over exposed like crazy! Sensitive little thing!

Thanks for all your help with this 'issue'.

[andrew s]

Jasper - I think what you have is a bad pixel that continually generates charge. When you are reading the column out it is adding charge to the other pixels charge in the column as they pass through it (hence it only impacts one side of the pixel. If this was there when delivered then I feel it is a column fault not just a pixel fault. Andrew

[Jessun]

Here's a BIAS stack, 20 of them. The line is dotted up and more solid down. Does that give a clue?

[Jessun]

Ok, the official SX answer is this: a warm pixel in the vertical register, which is a readout issue, and the cure is BIAS as pointed out by you guys. Very common too, and within specs.

I am happy with the reply, and can forget about it now. A few BIAS is no biggie, useful in any case I reckon. I can't wait for some clear nights...

Thanks again for the input on this guys!

[Warhen]

Hi guys, I realise this is an old thread. I'm a big fan of Terry and Michael at SX, for that matter, most of the camera makers are good folks, IMO. Just to back SX up, yes- this is such a common and insignificant issue. I think almost every CCD has, or will have it. Dark lines too. 'Bad columns' are a CCD spec. used with regard to video performance only, and not what we do. Further, as stated, these aren't 'bad columns.' A single warm or hot pixel can spill its charge over in this way. I believe the pixel's 'wall' may be weak as well. To some degree, a master bias may take care of it, but it's more likely that a defect map will, and most all software such as PixInsight, MaxIm DL, etc. will provide it. The white (or black) line will be filled-in with the average value of the surrounding pixels. Bottom line, no worries! Of the millions of pixels on a CCD sensor, this is a very minor and manageable thing. While CMOS has its own issues, they seem pretty immune to either hot pixels or vertical defects.  Thanks!

[ebdons]

Hi all, this isn't that related apart from I'm using the  opticstar ds616cxl+ which has this sensor as well abeit without the superior cooling, not had an issue with any lines but unless opticstar do something different with the way the their cam is put together, I might get this issue in the future? so this might be mostly down to just SX and QHY cams then. like folks say not a big issue with taking bias to sort out. maybe I've been lucky not having the bad column issue?