Something started happening in the recent betas that causes auto-stretch not to stretch correctly and it places the max value of the range too close to the min value, and the image is over-stretched and appears white. I have tried different values for the stretch and they don’t seem to have an effect. This is when taking images with the frame and focus control. The effect is so strong that it’s hard to believe others are not affected - so it may be specific to my camera or something in my settings. One thing I am suspicious about is the high speed download setting - which I may have changed.
I do have logs I can provide but I don’t have my equipment currently and can’t do additional testing. The logs didn’t seem to have details on the histogram scalings, however.
I know this all used to work for me in the product version - and I think the problem started happening in one of the betas. If no one else has this problem and there are no obvious leads - I will try to study it in more detail in a few weeks when I have my equipment.
Look at your picture and see if there is a column of dark pixels anywhere on it. When my SBIG driver was messed up, it would give me a line of block pixels on the first column… SGP would see that and it would cause it to generate a completely white picture. If you manually stretch it, no issues.
Hi. It’s possible some new dark pixels appeared in the camera output - but the stretch routine should really be going by the histogram itself, and it shouldn’t be misled by a small number of hot or cold pixels. So I don’t think the problem is due to a bad column - but either way I think something is wrong so that a good range is not being chosen. My last resort was to use manual stretching - but during autofocus runs it would use a different exposure setup and I would end up with the wrong stretch and either a black or white screen.
If other people are forced to use a manual stretch for some reason, then the problem may be more common. I think that the product version was working ok for me - but something else may have changed in my setup.
I’m afraid it doesn’t happen when I load an image - it seems to happen during manual frame/focus exposures, or during focus runs. It may be because the fits I have are longer exposures - as opposed to short exposures for frame and focus - I’m not sure. Changing the level of autostretch appears to have no effect. The log does show high speed download being turned on and off - so I’m suspicious it is somehow related. I also occasionally see brief images with a bright glow in the upper left that is very strange. But the images that appear white when auto-stretched do look normal once I manually stretch them less. The histograms for short exposures are all on the far left of the range - but the problem is that the max value is nearly on top of the min value.
Sorry I can’t do more at this point because I am away from my equipment but I can look in more detail in a few weeks.
On a related note, when I do load a fits image and change the auto-stretch amount - I would like to see the image updated with the new stretch. Currently I don’t see the image being refreshed and I’m not sure how to have the new range have effect without reloading the image.
When you get a chance right click a frame and focus image and save it as a fit for us to look at.
As for the stretch you can also right click an image and adjust the stretch. The auto stretch setting on the hostogram is only applied to images as they come in. Otherwise changing that setting could apply to multiple images at once and would be confusing.
Thanks - Unfortunately I won’t be able to try anything for a few weeks - but I did just realize I could load an old sequence from long ago and view the settings in it - and I confirmed that in the past I did not use High Speed Download - and at some point recently I turned it on - and it may have been correlated with the oddness I see.
So Chris do you have that set to ‘on’ currently? If not - could you try turning it on sometime?
Thanks for trying this. That sounds like what I am seeing. Yes - the image data is fine and you can manually set the high and low range - and it seems like the auto-stretch algorithm needs tweaking to be able to handle the case presented by these images. Fast download is a useful feature but not if you can’t have an automatic stretch.
This isn’t a priority and I can’t revisit it for a few weeks - and for Atik users it has a workaround not to use fast download - but it sounds like the stretch algorithm needs refinement and it may improve its behavior in general. In the histogram view the high caret is right on top of the low caret and you just need to pull the high caret to the right manually and you get a nice stretch. But it shouldn’t have to be manual.
Hi,
just wondered if this was resolved, as I’m having the same issue with frame &I focus and also via the plate solving capture frames with the Atik One.
Normal capture is fine.