Flats 😡

[+Stu]

Good morning. These are a thorn in my side. I've tried the APT flats aid and it doesn't reach the required ADU which is set to 25000 with a max exposure of 30s. The highest it will get to with a tablet/laptop screen in front of the scope is 4094. I'm clearly getting something wrong somewhere. Any suggestions?

[+Stu]

* currently using a duo band filter

[Bob Dobber]

My only suggestion, after nearly being reduced to tears with flats and the hateful (IMO, others love it, fair enough) ZWO 294, and other cameras to be fair, is to get a dedicated flat panel. I picked one up for just over £100 second hand, perhaps luckily. But it revolutionised my flat taking, and meant it was one thing that was still a bore and a chore, but at least I could be pretty happy it would work. I use NINA as well, I used to use APT and have never looked back to be honest.

 

This is what I  have...
 

 

 

[+Stu]

@Bob DobberThanks, after some extensive searching the only thing I'm thinking is that 25000 is too high for my sensor. I'm working on a better light source to make things easier. 

[Clive]

Is your sensor 12 bit?  If so the maximum possible ADU out of the sensor will be 4095.  The 25000 refers to 16 bit data so you would be aiming for 25000 / 16 ~ 1500

[Bottletopburly]

30 minutes ago, Clive said:

Is your sensor 12 bit?  If so the maximum possible ADU out of the sensor will be 4095.  The 25000 refers to 16 bit data so you would be aiming for 25000 / 16 ~ 1500

Apt scales too 16 bit 

[+Stu]

33 minutes ago, Clive said:

Is your sensor 12 bit?  If so the maximum possible ADU out of the sensor will be 4095.  The 25000 refers to 16 bit data so you would be aiming for 25000 / 16 ~ 1500

It is 12 bit, that explains why I wasn't getting near it. So do I aim for 4095/12? Or 25000/16?

@Bottletopburly

Edited September 3, 2023 by Stu

[Bottletopburly]

2 minutes ago, Stu said:

It is 12 bit, that explains why I wasn't getting near it. So do I aim for 4095/12?

Altair website says it’s 12bit and the ADC scales too 16 bit 

[+Stu]

13 minutes ago, Bottletopburly said:

Altair website says it’s 12bit and the ADC scales too 16 bit 

Apologies as I'm more confused now. So as it tops out at 4095, should my target be 1500?

[Bottletopburly]

I can’t seem to find what the full well value is for this camera ? Usually half that, here’s what to do , using apt flats aid set it to 22k Adu open histogram your looking for 1/3rds histogram approx , now if your leaving camera on scope then your flats will work for a while , no need to keep shooting flats every time which is why having autofocus unit is good as it locks focus , now the reason you had an issue with focus aid is probably because you light panel too bright or too dim , so put a T shirt over end of scope using elastic band run apt flats aid adjust brightness of screen to suit until you get a flat exposure between 1-5  sec with histogram between 1/3 and below 1/2 , which is why bobs flat panel good as it’s adjustable for brightness, obviously if your using different filters results will vary experiment so you know what brightness to use for each filter you use , flats can be a pain but stick to 1/3rd histogram as a general rule .

[+Stu]

@BottletopburlyCool, thanks. When you say 1/3 histogram, does that mean spikes in the left hand 3rd of it? I'll post a pic of what I did, see if I'm on the right track or passing water into the wind. 

[+Stu]

@Bottletopburly?????

[Bottletopburly]

What’s it like clic L button 

[Bottletopburly]

Did you set Adu to 22k in flats aid ? Your range only showing up to 4094 

[+Stu]

2 minutes ago, Bottletopburly said:

Did you set Adu to 22k in flats aid ? Your range only showing up to 4094 

It won't go any higher than 4094 regardless of exposure or brightness of the screen I'm using. This wasn't taken using flats aid as it never reaches the target ADU of 22k

Edited September 3, 2023 by Stu

[Bottletopburly]

8 minutes ago, Stu said:

It won't go any higher than 4094 regardless of exposure or brightness of the screen I'm using. This wasn't taken using flats aid as it never reaches the target ADU of 22k

Ah so maybe that’s your camera setting try 2000k in flat aid 

[+Stu]

2 minutes ago, Bottletopburly said:

Ah so maybe that’s your camera setting try 2000k in flat aid 

I'll give it a whirl when I get the chance, the imaging train is still assembled. As for the image I posted, it has a fade from top to bottom. This introduces a gradient on my images when I stack them with everything else 🫤

[Bottletopburly]

1 hour ago, Stu said:

I'll give it a whirl when I get the chance, the imaging train is still assembled. As for the image I posted, it has a fade from top to bottom. This introduces a gradient on my images when I stack them with everything else 🫤

Sounds like a light leak maybe between scope and flat panel or t shirt not taught possibly 

[+Stu]

@BottletopburlyI don't own any white t-shirts (messy eater) so I use some plain A4. I had the tablet screen flat against the dew shield, maybe not flat enough. Some experimentation needed I think.

[+RonC]

I use a white T shirt and 2x A4 copier paper and that works with my 183c.

[Clive]

That seems a very strange histogram shape for a flat, is it the average of the 3 colour layers?. As @Bottletopburlystated above, although the camera is only 12 bit, the image data in the FITs file should have been converted to 16 bit in the camera by multipling by 4 bits (ie x16) so the FITs file should be 0 to 65535 rather than 0 to 4095. I don't use APT but any software that displays the 'real' histogram of a FITs file data should show what the ADU range of the FITs file data actually is.

[Bottletopburly]

6 hours ago, Clive said:

That seems a very strange histogram shape for a flat, is it the average of the 3 colour layers?. As @Bottletopburlystated above, although the camera is only 12 bit, the image data in the FITs file should have been converted to 16 bit in the camera by multipling by 4 bits (ie x16) so the FITs file should be 0 to 65535 rather than 0 to 4095. I don't use APT but any software that displays the 'real' histogram of a FITs file data should show what the ADU range of the FITs file data actually is.

When I run flats aid in Apt I get a single spike so you can roughly see when your 1/3 rd on histogram I checked tonight , I can only assume maybe in settings under ccd tab make sure colour fits preview debayer needs to be checked  @Stu, probably worth asking on SGL see if anyone using same camera/sensor and see what settings they use , or drop Altair a message, apt is only reading info from the camera 

[+Stu]

@Bottletopburly @Clive Thanks for the support. Im going have a play later and do some research once eldest is at school, things have been a bit rushed over the holidays. 

[+Stu]

On 03/09/2023 at 09:54, Bottletopburly said:

can’t seem to find what the full well value is for this camera ?

I've been looking at the Orion starshoot G21 which uses the same sensor. That says the full well value is 23k.

[+Stu]

A bit more info on the Orion site. It mentions having to use longer exposures than the 294 which is also available for this camera.