Error registering light frames using BPP

sandconp

Well-known member
I am running BPP for the first time and I added my 30 light frames, 1 master dark and flat and no master bias frames.

I get the message that ImageCalibration 31 Succeeded 0 failed and 0 skipped then end calibration of light frames.

Then I get a message begin registration of light frames but I get an error: No stars found error registering light frames.
 
open one of the _c_d (or _c files, if not an OSC) created by BPP and take a look at it. if the calibration is bad then the data in the calibrated sub can be completely destroyed and no stars show up. if you see kind of an image that looks like it only has 1s and 0s in it (meaning the pixels are only completely black or completely white) after applying STF this is probably what happened.

make sure that dark optimization is turned off since you have no bias frames.
 
I have a folder called calibrated then light and I can see 30 files with the _c.xisf extension. I am using a One Shot Color camera.

I don't have a master folder with the stacked image.

What should I be looking for in the _c.xisf files? Do you want me to check the FITS header?
 
1596558273827.png
 
The above image is Comet Neowise. When I do a STF auto Stretch, the back ground turns pink. The comet and stars are in red.
 
I tried it again and unchecked dark optimization since I have no bias frames and this time, I got a different error message.

* Using 4 worker threads.
[2020-08-04 16:56:21] [000] E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_01_c_d_r.xisf
[2020-08-04 16:56:21] [001] E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_02_c_d_r.xisf
[2020-08-04 16:56:21] [002] E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_03_c_d_r.xisf
[2020-08-04 16:56:21] [003] E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_04_c_d_r.xisf
[2020-08-04 16:56:38] E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_04_c_d_r.xisf
[2020-08-04 16:56:38] Loading image: w=5496 h=2754 n=3 RGB Float32
[2020-08-04 16:56:38] 8 image properties
[2020-08-04 16:56:38] 80 FITS keyword(s) extracted.
[2020-08-04 16:56:38] Computing image statistics: done
[2020-08-04 16:56:38] 80 FITS keyword(s) extracted.
[2020-08-04 16:56:38] 80 FITS keyword(s) extracted.
[2020-08-04 16:56:38] 80 FITS keyword(s) extracted.
[2020-08-04 16:56:38] Scale factors : 0.99779 0.00000 0.00000
[2020-08-04 16:56:38] Zero offset : -1.103088e-03 -1.778104e-05 -1.778104e-05
[2020-08-04 16:56:38] Noise estimates : 1.7780e-03 7.6500e-05 7.6500e-05
[2020-08-04 16:56:38]
[2020-08-04 16:56:38]
[2020-08-04 16:56:38] * Waiting for running tasks to terminate...
[2020-08-04 16:56:42] *** Error:
[2020-08-04 16:56:42] *** Error: E:/Sharpcap Captures/2020-07-31/neowise/23_37_59/BPP/registered/frame_04_c_d_r.xisf (channel #1): Unable to compute noise estimate.
[2020-08-04 16:56:42] <* failed *>
[2020-08-04 16:56:42]
[2020-08-04 16:56:42] ************************************************************
[2020-08-04 16:56:42] * End integration of light frames
[2020-08-04 16:56:42] ************************************************************
 
unlock the STF channels when applying the STF to be able to see the image.

clearly calibration is messed up... did you use the same software to acquire the lights and the darks?

rob
 
I removed the master flat and unchecked the CFA Image Global options and I got it to work but I have artifacts because the flat was not applied.

Please Advise.
 
can you post the master calibration frames and a light subexposure?

upload to dropbox, google drive, etc. and post the link here.
 
So I am assuming you mean the output files from BPP?

BPP created four folders calibrated, logs, master and registered.

Within the calibrated frame folder then light folder, I have (30) c.xisf files is this what you want me to upload?

For the light exposure subexposure, do you mean one of the my light frames that is fed into BPP or the one light file that was stacked in the master folder from the BPP output?

Bear with me as I am new to this process.

Thanks.

Liked I stated earlier, I can get this to work if I remove the master flat and just use the master dark and uncheck the CFA Image option since I am using a OSC camera.
 
sort of, i am talking about the calibration masters (dark, bias, flat) which BPP probably created, and also at least one subexposure as it came from the camera - that is, not calibrated - the input lights to BPP.

something is wrong because CFA should be checked. maybe your master flat is bad, but i'd have to see it to make any judgements.

rob
 
all of these files appear to already have been deybayered. all are 3-plane RGB files which although bayered files can be stored that way they should have a bunch of black pixels in each channel and i don't see that here.

did they really come out of sharpcap that way? if so, sharpcap debayered the image. unless it was superpixel debayering, these images can't be calibrated properly. and even if they are superpixel debayered the green channel can't be calibrated properly. you can still do it by treating them as non CFA images but the result is not going to be optimal.

rob
 
Hi Rob,

Thanks for taking a look at my uploaded files. Yes the files did come out of Sharcap that way. I always tell Sharpcap to use 'RGB' and maybe I should be telling it to use 'RAW' Format?

I am able to take the individual frames directly into Sharpcap to stack the images and I don't have any issues with that but it looks like PixInsight won't allow it.

I am going to post a question to Robin over at Sharpcap to see what he says and I will let you know.

Thanks again.
 
Also to get it to work in BPP, not only do I have to un-check the CFA option, but I am not able to user my Master Flat, only the Master Dark works.
 
i didn't check the calibration - i'll have to check it out to see what happens. at any rate yeah you should use raw mode in order to be able to properly calibrate your frames. it's not that RGB mode won't work at all but the results are likely to be incorrect because of the interpolation of pixels and the averaging of the two green pixels.

unchecking CFA is correct for this data, because it is not CFA data. (CFA means "color filter array" and is a synonym for "bayered color data")

rob
 
is "flats_00_49_38" a flat master? if so it looks like it was made in sharpcap... and that is probably the source of the problem. making masters with one program and trying to use them in another to calibrate rarely works.

if the flat was not scaled by sharpcap, then it is pretty overexposed. if it was integrated and normalized in sharpcap then it makes sense that the highest pixel values read out as almost 1.0. the flat subs should really only be exposed to approximately 1/2 the well depth in order to avoid going past the sensor's point of nonlinear response to light. so if you have 16-bit subs, the right side of the histogram should not exceed 32768. in practice most people put the peak of the histogram there since most sensors are still linear past the 1/2 way point and so it is safe to have pixels that are > 1/2 the well depth.
 
Back
Top