Clipped/noisy subs - Possible FITS pedestal/black-level regression: Page 6

Clipped/noisy subs - Possible FITS pedestal/black-level regression

wvreeven

zjW8Hx8Y for the fact that they finally take our complaints seriously and are investigating again. Only complaining doesn't get us nowhere. A little appreciation can work miracles.

Rob603

To better illustrate the clipped data problem that has existed since the 3.0.0 update, attached are two screenshots from Siril showing the histograms of two 20 second subs with LP filter of the Pac-Man Nebula. The first sub was taken before the 3.0.0 update and second taken after. Both subs were taken on perfectly clear nights, no clouds. Notice on the second sub that the data spike is jammed over on the far left side of the histogram. This histogram shows that the data has been clipped and is missing data that would have contained the fainter details

of the nebula.

Nasa-HSIA

Rob603 It appears that the images processed in the new version were processed using an incorrect Dark image, resulting in over-subtraction of these images.

Support@Seestar

wVU1j6V6
Rob603
zjW8Hx8Y
Please contact us at this email address：support@seestar.com; we would like your help in analyzing this. Thank you.

Rob603

Nasa-HSIA
Yes, it does look like a dark subtraction issue during calibration. The fact that it’s lingered for over a month through multiple updates suggests a more complex root cause for the bad subtraction. It may involve how the app handles dynamic scaling for temperature variations, interactions with the LP filter’s reduced signal, or an even broader regression in the dark frame matching logic. If there are a lot of factors, it may be difficult to isolate the problem quickly.

YF5I3f19

The problem seems pretty clear now, I wonder why it takes so much time to ZWO to fix it, given that it is a glitch that impacts so heavily on Seestar usability and should be ranked as critical.

mbits

zjW8Hx8Y I'm at 6/7 Bortle.

hlc

Support@Seestar I am also seeing this problem, and include my evidence here:
Comparison of two jpegs of individual subs, before and after the firmware update (both using LP filter):

Comparison of the .fit files opened in Siril - stretched, with histogram view, to show the clipped pixels:

I can upload the .fit files, and provide more examples, if that would help.

As others have said, I hope a fix for this is being worked on urgently. Thanks

tom-dd132

Support@Seestar

Support@Seestar We understand that the matter is being handled and resolved.

But I understand that the actual problem of this whole thread has neither been understood nor resolved by ZWO so far?

zjW8Hx8Y

Support@Seestar
Exactly one month has passed. What is the current status?
Has the root cause been identified, and will a fix be included in an upcoming update?

Seestar_S50

There are so many issues with this version v3.0.2 please ZWO let's us know when they will be fixed so that the end user can start using the Seestars as intended and not as paperweight.

Nasa-HSIA

Support@Seestar Please tell us clearly when these issues can be resolved.

Support@Seestar

Nasa-HSIA When this problem occurs, please take a photo of Seestar and the surrounding environment with your mobile phone and send it to us for troubleshooting, thank you.

Tim1343

Support@Seestar

Sorry am I missing something here, are you asking Nasa-HSIA to take a picture of their Seestar in operation? And you now want your customers to send in data to help you identify the problem, after a month of repeated reports about this fault? This is getting increasingly ridiculous and bizarre. ZWO are looking extremely amateurish right now. As others have suggested, why don’t you roll back to the previous software version immediately so we can all get on with using our smart scopes, while you try to fix what you have broken!

hlc

Support@Seestar this is not an environment issue. It started happening after the firmware update to different users all over the world, in different conditions. Please roll back the change so we can use our Seestars again!

Michael875

wVU1j6V6

I also tried maximum stretching.

M31 has black pixels all over the picture except for the center of the galaxy.

M45 has black pixels everywhere except for the center of the seven sisters.

Both pictures were taken using alt/az mode with 10 sec subs.

Histogram also shows that the data is clipped

Attached the fits files in case those help to resolve the problem:
M45
M31

My guess is that the Seestar-internal post processing causes the issue. Any progress on resolving this?
Unfortunately, I do not have any comparison pics from before the update - got my Seestar after the update was released and immediately applied the update.

Nasa-HSIA

Michael875 ZWo asked us to send an email to them for analyzing the issue.

email address：support@seestar.com

Michael875

Nasa-HSIA
done

6gYT4MFO

Tim1343 its time for an dwarf...

mbits

This is not related to owners individual local environment. It's an update bug. I've had my s30 nearly a year, before the December update this clipping never happened, now it's every time.

jgood

I’m a new S50 owner that just read this thread for the first time. I’m actually relieved to discover this problem is out there! I’ve been struggling with really noisy images coming from my S50 that are terrible if processed in Siril. They all look like everyone’s “bad” examples above. My S50/App has only been 3.0 and above since I received it. Now I’m going to follow this situation closely.

« Previous Page Next Page »