First light of self modded 1100D

[Rowland]

Hi Marco.

I hope this helps?

Added my dither, no-dither comparison jpeg

I had a look at the data in your first link. Dithering could be more aggressive. For DSLRs 10 - 15 pixels between frames is a good setting. Otherwise, I would use the highest dithering setting available.

However, you might be able to rescue this set by blinking your files to find the best quality sub-frame - and use this as your reference image. It might help to turn up pixel rejection as well.

Unless the data set is beyond help the method of calibration described in my previous post and selecting the best sub-frame as the reference is the best way to go in the future. It took a few years for me to work all this out.

The data in the second link looks like a problem with the flats.

Rowland.

[MarcoS]

Hi Ivo & Rowland,

Ivo, I agree, I do not discern the streaks in the light only stack, I am also completely at loss about what is going on.

Rowland, unfortunately DSS will not let me do what you propose as stacking method...
The data in the second link was a set without Flats...

I use DSS nearly one year, and I use it always with the same settings for stacking, and never had this problem.
I attach a sample of an image DSS stacked and processed in ST. This image was taken previous to the camera Mod.

M-42-small.jpg (103.45 KiB) Viewed 5571 times

This was easy to process and did not display any anomalies.

The first set of light I took after modification display lots of problems, could it be that those problems are related to the mod?
The mod was LPF#1 filter removed. LPF#2 filter replaced with Baader filter.
sensor shimmed -0.3mm to compensate for focus.

The flats bias and darks used are all "freshly" taken after the mod.

I will try other staking combination ans see if I can found out some clues....

And if I got some clear skies I will take an other set of lights and see how do they look.....

[Rowland]

Marco. I have had similar experience with good and bad results using the same software package and identical processing. Sometimes, and depending on variables in the data, errors in calibration are masked and not self evident. And this is all dependent on the quality of the sum of the data out of the camera.

EDIT: Other things that might affect your results - temperature consistency across the whole data set. It is not unusual for the lines to be absent in an uncalibrated stack. This doesn't mean you should not calibrate, the lines may be present because the calibration frames, most likely the darks are not temperature or time matched very well to your lights. Dithering is intended to hide the unavoidable, but hopefully minimal errors in DSLR RAW data calibration.

For uncooled cameras it's best to keep imaging with minimal time between exposures to keep the sensor temp as consistent as possible. Do the same with darks in the same session preferably.

Astroart does let you process your data as described in my post and the results are more consistent. Pixinsight the same.

My apologies, I misunderstood your post with the no flats applied.

[MarcoS]

...... Other things that might affect your results - temperature consistency across the whole data set. It is not unusual for the lines to be absent in an uncalibrated stack. This doesn't mean you should not calibrate, the lines may be present because the calibration frames, most likely the darks are not temperature or time matched very well to your lights. Dithering is intended to hide the unavoidable, but hopefully minimal errors in DSLR RAW data calibration.

For uncooled cameras it's best to keep imaging with minimal time between exposures to keep the sensor temp as consistent as possible. Do the same with darks in the same session preferably.

Astroart does let you process your data as described in my post and the results are more consistent. Pixinsight the same.
........

Hi Rowland,
Thanks for all the time and explanations, I fully agree with the above, sometimes due to a combination of factors the results are not as expected.
Although in this set the darks were freshly taken on a very similar temperature of the lights, (within 5 degree) the bias where also fresh, as where the Flats.

I have done a bit of investigating of the histograms of the flats , and I think that I have pinpointed the biggest culprit of the failure of this set.... to the Flats...

I usually take the flats with a self made Led flat box, this box has always worked fine until I modded the camera...
Attached screenshots of the histogram of the flats made before, and after the modding with the Led Light Box, and the histogram taken with the white T shirt method after the modding.

Histograms.jpg (93.87 KiB) Viewed 5554 times

Those histograms speaks by themselves, I think that the light of the led box produce is not suitable for a modded camera.
The histogram taken with the T-shirt method has the 3 colors much better positioned not so far from eachother.

I had yesterday e 30 minutes clear skies before started to rain, was a bit windy as well, but anyway, I took 10 x 180sec ISO 1600 lights of the horse head nebula. The moon was not up yet.
I have made a stack with flats (led flat box) darks and bias and a stack with flats (T-shirt method) darks and bias

I have processed the two the same way (autodev, crop, bin, autodev, wipe, develop (same develop settings on both stacks)

Here the results:

Horse-head-Led flat-box-medhod.JPG (204.51 KiB) Viewed 5554 times

Horse-head-t-shirt-flat-method.JPG (209.62 KiB) Viewed 5554 times

The stack with the flats made with the T-shirt method produce better results.

I think I can state that the biggest troubles makers where the flats....

Any way, I will give a try at your suggested method with Astroart.
I did try a while back a demo of Pixinsight but I did not like the program much.

Any way, thanks again for all the help:)

[Rowland]

Thats good news Marco. Always good to track down the culprit. I see what you mean about the flats AstroArt is virtually set up for the style of calibration. But now that I have worked out Pixinsight calibration I can't justify purchasing AstroArt. Having used the demo, it would be my first choice of convenience. There's a bunch of tick boxes for various parameters and is easy to set up.

[MarcoS]

Good day,

A little update, I have modified my self made Led Light box by adding on the back of the light box reflecting side a blue plastic film coming from a plastic office maps, you know, the one's to put A4 documents inside?

The one on right most side is the histogram of the resulting Flat, all the colors are nicely balanced in the center.....

Histograms.jpg (127.88 KiB) Viewed 5537 times

Have a nice weekend

[Rowland]

That looks much better. Not sure where you have the camera histogram from those screen shots. Around ~70 - 75% works quite well for me.