ATIK 383L+? Very dissatisfied!

[lukebl]

I recently acquired a used mono Atik 383L+ to replace my Atik 428ex, as it seems to have rave reviews and I needed a bigger sensor than my 428ex.

However I really miss the awesome 428ex. It was incredibly sensitive, able to pick up really faint objects with relatively short exposures, and had virtually no noise or hot/dead pixels. I didn't even need to take darks.

On the other hand the 383 has thousands of hot pixels and is nowhere near as sensitive as the 428. Is it faulty?

Here's a 2x binned dark frame, heavily stretched to show all the hot pixels. Is this normal? I can live with the column of hot pixels on the right, but DSS can't seem to get rid of the the blizzard of hot pixels.

[Richie092]

Just had a look at a couple of my 600 second darks at -15 and that looks perfectly normal. Don't have PI on this machine so can't stretch and save one but I don't see anything wrong with yours.

 

[steppenwolf]

2 hours ago, lukebl said:

On the other hand the 383 has thousands of hot pixels and is nowhere near as sensitive as the 428. Is it faulty?

This is pretty normal for this particular sensor although my 'columns' disappear as soon as I cool my (QSI 683) one to -20 degrees C. Dithering while imaging and ideally using PixInsight's 'Cosmetic Correction' algorithm before stacking will remove these noisy pixels well and you'll get some great images then.

[ollypenrice]

Agreed. That looks fine to me. Once you get the right calibration routine sorted out you should get clean output.  I use AstroArt for stacking and, although my Kodak chips have far more innate noise than my Sony, the end result is excellent. AA gives you the option to correct hot pixels and repair columns. Just mouse over the column to see it's y-axis co-ordinate and plug that into the last page of the pre-processing menu. The old Kodak chips are not as sensitive but have a fantastic track record for producing great real world results.

Olly

[Davey-T]

Looks pretty much like mine Luke, you were spoilt with the technical advances of the 428, a quick blast of hot pixel remover in whatever processing software you use and they'll all disappear like magic 😄

For even more sensor real estate look out for a second hand Atik 4000, similar snowstorm of hot pixels.

Dave

Edited April 15, 2021 by Davey-T

[lukebl]

Thanks for all the reassurance, folks. However, I think I'll always regret selling the 428ex.

I'll have to sort out my dithering. People tell me I'm very good at dithering!

[CraigT82]

Cosmetic correction in PI, or a bad pixel map in APP sorts it out very nicely. 

[Sp@ce_d]

I had "That Moment" of shock when I first got my QSI683 back in 2015 after being so used to the Sony sensors. Mine looks worse than yours! Dithering makes an amazing difference and anything else left is sorted by PI preprocessing for me. I love this camera/chip, it may be old tech now but it still produces dependable results and you know what they say.. if it ain't broke...

[ollypenrice]

OK, I'll say it. I prefer my Kodak chips to my Sony.  If I wanted to take and publish dark frames then I would go for Sony every time. The dark frames are great. But I actually want to take and publish astrophotos. 

The real business begins when you start to post-process a calibrated stack. It is at this point - and only at this point - that you begin to appreciate the Kodak data. I find that it behaves itself. I don't have to fight it. In particular the background sky and the bright stars just sort themselves out. The Sony chip's background sky will not come up to a natural brightness 'naturally.'  We think of noise as being unwanted extra signal. That's not my problem. I have a problem with overly dark pixels in a stretched background sky and a background sky which remains too dark for comfort. (I have tried all the calibration options like darks, no darks, bias-as dark, etc etc. Nothing much changes). Then the stars just aren't what I want. When I added a Sony chip to my line-up, sharp observers like Sara Wager said my stars were not what they used to be. Quite right. What was the 'secret' of my previous stars? An Atik 11000. That's about it.

In post-processing I have to fight my Sony data all the way. It may be lack of well depth but I'll leave that for the theorists.

Call this a minority report! 

Olly

[wxsatuser]

My 383 900sec dark -20

[newbie alert]

I've heard from a very reliable source that the 383 needs a decent power source .. because of its very deep well depth it also likes long subs.. and calibrates out clean as a whistle

[carastro]

I had the same problem when I had an Atik 383L and was fed up with cloning out hot pixels.  I bought a 460EX instead and later a 428EX and never looked back despite the smaller sensors.  

Carole   

Edited April 19, 2021 by carastro

[Magnum]

ok firstly the 2x2 binning sub you posted  will look worse than 1x1 on this camera in my experience, I never use binning anyway.

this camera needs more than 12 volts and cooled to -20C, i had similar amount of hot pixels opwhen i first got my camera but after speaking to Vince at Atik he told me The Kodak Atiks can safely use upto 15v and they are cleaner with more voltage, i did some tests and once i got to 13.8v the warm / hot pixels greatly reduced to a similar lever as my Sony sensor Atik460, i got no improvement above 13.8 and thats quite convenient as most Bench power supplies are in fact 13.8V anyway. So I use my old Maplin bench supply with the 383. The 460 is happy on standard 12v laptop style supply. 
similarly if you use it warmer than -20C you will see a few more hot pixels coming back.

so 13.8v at -20C is the cleanest it gets.

I agree with Olly, I prefer my Atik383 to my 460 for the final images they produce. The only time i dont like the 383 is when focussing and searching for the DSO as it looks noisy and dramatically thin compared to the 460, However the actual 10 - 15 min ha subs look as clean as the 460, and i actually find that the hot pixels  disappear into the stack much quicker in the 383 so much so that I dont really need to dither with the 383, (on the 460 dithering is essential) and  I dont use darks either though I do take bias frames as they fix the left edge of the frame which without them is a little brighter than the rest of the frame.

Sarah wager also found she prefered the final image noise profile of the 8300 chip compared to her previous sony sensor and had QSI swap the sensor in her cam.

I would say the Kodak 8300 is actually the cleanest Kodak sensor you will ever use, if you want to see a snowstorm, Olly might show you what the 11000 looks like in single subs LOL. 

There is one for sale on here at the and im tempted to get another one LOL, so i must be quite a fan. 

Edited April 20, 2021 by Magnum

[Magnum]

On 19/04/2021 at 22:44, newbie alert said:

I've heard from a very reliable source that the 383 needs a decent power source .. because of its very deep well depth it also likes long subs.. and calibrates out clean as a whistle

Yes it needs 13.8V

[Magnum]

Thought I might as well sow the quality of data the 383 produces when used at 13.8volts @ -20C, ive attached screenshots of uncalibrated single 900 sec subs of the Pelican and Heart nebula's both at 100% to show how clean the subs are, and also the final images which are both only 15 x 900 secs each using just bias frames and dithering.

Please don't give up on the camera unless you want to give it to me really cheap LO 😛

Lee

[ollypenrice]

On 20/04/2021 at 00:54, carastro said:

I had the same problem when I had an Atik 383L and was fed up with cloning out hot pixels.  I bought a 460EX instead and later a 428EX and never looked back despite the smaller sensors.  

Carole   

Did you try the hot pixel filtration in AA, Carole. It is adjustable in intensity and gets nearly all of them for me. There should never be any need to clone them individually.

 

7 hours ago, Magnum said:

Olly might show you what the 11000 looks like in single subs LOL. 

 

 

I've lent the camera to Tom for a while for his robotic rig so don't have a dark in stock but snowstorm is the word. Who cares? It's the picture that matters:

https://www.astrobin.com/383965/?nc=user

https://www.astrobin.com/321869/?nc=user

I do hardly any small scale cosmetic cleaning after stacking. Quite often I do literally none at all.

Olly

 

[Magnum]

9 hours ago, ollypenrice said:

 

 

I've lent the camera to Tom for a while for his robotic rig so don't have a dark in stock but snowstorm is the word. Who cares? It's the picture that matters:

 

 

no exactly the point I was trying to make, as your images are world class 🙂

[carastro]

Quote

Did you try the hot pixel filtration in AA, Carole

I have not heard of that option Olly, so no I didn't. 

Carole  

 

[ollypenrice]

5 minutes ago, carastro said:

I have not heard of that option Olly, so no I didn't. 

Carole  

 

Preprocessing, page three. I think it's called 'extras.' It's full of great features. As well as hot pixel removal you have column repair. Since dead columns are just a fact of life in CCD imaging you select Column Repair and type in the number of the dead column on your chip. It will be repaired by giving each pixel in that column the average value of the pixels to its right and left. The dead column then simply disappears. (To find which column it is just open a sub on which the column shows and mouse over it. Put the cursor on the column and look for the x value at the bottom of the screen. If x = 2234 then 2234 is your dead column.) I write it down on a bit of paper and stick it next to my screen. You know what I'm like!! 🤣 I have lots of bits of paper because I use lots of cameras...😜

Olly

[gorann]

Maybe I am swearing in church and this comment is not immediately helpful for the OP but there is the new generation of cooled CMOS. Here is a dark from my ASI2600MC (5 minutes at gain 100), and a stretched frame grab of it from PI. I can forget darks or dithering or bad pixel/column processing or waiting for download😉

2021-04-06-1238_7-CapObj_0014.FIT

Edited April 21, 2021 by gorann

[Magnum]

9 hours ago, gorann said:

Maybe I am swearing in church and this comment is not immediately helpful for the OP but there is the new generation of cooled CMOS. Here is a dark from my ASI2600MC (5 minutes at gain 100), and a stretched frame grab of it from PI. I can forget darks or dithering or bad pixel/column processing or waiting for download😉

2021-04-06-1238_7-CapObj_0014.FIT 49.77 MB · 11 downloads

I have a 533mc with is smaller version of that chip, yes the individual subs are cleaner, but I find I have to stack loads of them to get a good final image, yet the Atik383 gets super smooth in just a handful of 10-15 min subs, I dont use darks with any of my cameras, but do use dithering, actually I find dithering has the most benefit on the cmos or I get walking noise.

If I had to pick one camera to keep out of my 3 it would be the 383 .

[gorann]

12 hours ago, Magnum said:

I have a 533mc with is smaller version of that chip, yes the individual subs are cleaner, but I find I have to stack loads of them to get a good final image, yet the Atik383 gets super smooth in just a handful of 10-15 min subs, I dont use darks with any of my cameras, but do use dithering, actually I find dithering has the most benefit on the cmos or I get walking noise.

If I had to pick one camera to keep out of my 3 it would be the 383 .

That is explained by 14 bit vs 16 bit. The 383 and the 2600 are 16 bit while the 533 is 14 bit so you would need more subs to get to the same depth.

[Magnum]

On 21/04/2021 at 23:14, gorann said:

Maybe I am swearing in church and this comment is not immediately helpful for the OP but there is the new generation of cooled CMOS. Here is a dark from my ASI2600MC (5 minutes at gain 100), and a stretched frame grab of it from PI. I can forget darks or dithering or bad pixel/column processing or waiting for download😉

2021-04-06-1238_7-CapObj_0014.FIT 49.77 MB · 13 downloads

Id agree with most of that, but we can't forgo dithering with cmos, as they all have a faint fixed pattern noise that CCD doesnt have. you can see this become evident if you stack a load of bias or darks from even the latest zero amp glow models, a pattern quickly emerges. In real world imaging with the 533 I too dont use darks or bias, but if I dont dither I get diagonal banding pattern to the noise, ie walking noise which I dont see with my CCD's. luckily dithering completely solves  it and I always dither with either CCD or CMOS, so yes the 533, 2600 & 6200 are probably the closest cmos has got to CCD so far.

Lee

[Magnum]

 

.

Edited May 4, 2021 by Magnum

[fwm891]

I have a 533 and have never taken dark or bias frames. A flat for dust bunnies sometimes if a test frame shows a problem. Otherwise it's DBE in PI to even out a background.