Can't get OIII or SII to stack properly

[Lonnie]

OK, I'm really in need of some assistance!!   
 
I have tried pretty much every setting I can think of in Batch Preprocessing but cannot for the life of me get OIII or SII to stack properly!! Ha stacks (same parameters) with no issues and looks fantastic.
I have 5 hours of each channel, 10 x 30min lights and the data looks good. I'm using 20 flats, 200 bias and 50 darks for each one.
My usual stacking parameters are: Combination = Average, Windsorized Sigma Clipping, Sigma low - 3, Sigma high - 4, Lanczos 3. Up until this morning it has worked fantastic for OIII and SII.
I've attached a screen shot of what I get now, with STF applied. As you can see, OIII and SII are almost completely obliterated.
I'm hoping somebody knows why this is happening and can set me straight again.

I have exhausted almost every path to take. I've tried stacking with and without darks, or flats, or bias in almost every conceivable way possible. I've tried using less of each as well. I've tried stacking each filter on their own as well as all together. I've tried with and without optimize dark frames.

Camera is an SBIG STF-8300M. All images were captured with Sequence Generator Pro.
I haven't tried any other stacking programs yet but that is my next approach, just to see.

[pfile]

what does a single SII or OIII sub look like with STF applied?

[Lonnie]

I think they look OK. They're faint but there really isn't too much OIII and SII to pick up on. Maybe I'm wrong.
My last image of NGC281 did not have much data either but it still stacked. I actually just tried it and it stacked perfectly.
I'm at a complete loss.

[Lonnie]

And here's what the master bias and dark look like, STF applied.

[pfile]

i have seen similar before, but the calibration had gone terribly wrong. in that case it was because i was using masters that came out of maxim, rather than making my own in PI. but it sounds like that's not the problem here.

is SGP doing anything funny with the data? that is, does the fits file it produces contain i16 data?

what happens if you subtract a single dark from a single OIII/SII light?

rob

[Lonnie]

Thanks for the help Rob! 

is SGP doing anything funny with the data?

This gives me an idea, I've downloaded PI for Windows and will try stacking in my observatory, just to see. I capture everything on a Windows computer in my observatory then copy to thumb drive and process in my house on a Mac. Possibly something is getting messed up in the transition?

What PI see's when it opens the frames seems normal;
Reading 1 image(s):
/Users/lonnie/Desktop/Elephant Trunk Nebula/SII Light/elephant trunk-1_1800sec_1x1_SII_frame2.fit
Reading FITS: 16-bit integers, 1 channel(s), 3352x2532 pixels: done

Reading 1 image(s):
/Users/lonnie/Desktop/Darks/Dark x 1800_ -20/darks_1800sec_1x1_Dark_frame16.fit
Reading FITS: 16-bit integers, 1 channel(s), 3352x2532 pixels: done

As far as subtracting a single dark from a single light, it does more damage than I thought it would, as long as I did this right. I just used a simple equation in pixel math, light frame - dark frame.

[Lonnie]

Well, it looks like my observatory computer did the trick! I'm not sure why yet but it worked a whole lot better. 

Thanks again Rob for all the help! Without your input I would have never thought of trying my PC. From now on I'll most likely do all the stacking in my observatory.

It may not look like much but here's the quick stack of OIII and SII files using auto settings. I couldn't even get close to this kind of result today. Some minor tweaking and I'm sure it'll be fine.

[pfile]

this is really quite strange and bears more investigation, i think. i also have a hybrid environment - if i use my DSLR i am capturing with backyardEOS running under vmware, which is dumping files into a shared folder. if i use my STT-8300M i use equinox image running on osx. however, astrotortilla is totally fine with the fits files being written by equinox image.

not the same case exactly, but i am writing all the files in question to an SD card formatted with ExFAT and have never had problems with the files.

i know other people use SGP and PI but perhaps they are not working cross-platform...

rob

[Lonnie]

To be quite honest, this is the first time I've had these issues. Now I'm thinking there may have been some data corruption someplace in the transition from my PC to my Mac, possibly with the thumb drive. The last image session I took had zero issues and with this batch, it was only the OIII and SII. The HA data worked perfectly. I'll give it another try tomorrow with a different thumb drive just to see.

Another thing is I have yet to take any updates on my PC for the Batch Preprocessing. I'll take them updates tomorrow and try it again to see how it works.

[Lonnie]

Well, it appears like I just needed more subs. I collected another 8 x 30min subs of OIII last night and viola! I guess the SNR just wasn't good enough?
Chalk it up to inexperience I guess.     I'm calling this one solved.
Now I just need to capture some more SII!