Vertical streaks

[Stub Mandrel]

While I'm pleased with the star shape and colour in this image of Merope in the Pleiades, there are some very strong vertical bands. I'm guessing these are caused by the camera and could be reduced by using flats (I only had bias & darks for this shot), except the same bands (but fainter and narrower) appear when I use flats.

Am I right about the cause or could it be something else?

[Herra Kuulapaa]

How many lights you have there?

[michael.h.f.wilkinson]

Cameras can show streaking artefacts. Darks and flats can sort them

[Thalestris24]

Hi

Looks like sensor-dependent banding (read) noise to me. I'm not an expert but - probably best reduced by using a low iso, and not stretching the image too much. It may not calibrate out - check your master dark, bias and flat frames. I think the problem can be temperature dependent - lower = better. Maybe try to match temperatures of lights and calibration frames. StarTools has a utility to reduce horizontal or vertical banding.

Louise

[Stub Mandrel]

That was 20 30s subs, grabbed near the end of the lunar eclipse - I just noticed the Pleiades were high and bright so I swung the scope round before finishing up...

I suppose the sensor would have been quite warm after spending an hour and a half taking 255 shots of the moon back to back,the latter half all 15 second exposures...

A bit nervous as my Andromeda shows them (but much less noticeably) and that was 64 frames with ~ 20 each of darks, flats and bias.

I've just looked at stretched versions of my darks, flats and bias. They aren't on the darks at all, but very faint on the flats and quite noticeable on the bias frames. Perhaps this is because the bias frames get stretched the most. They look finer, and I wonder if the effect is worst when a relatively small number of frames is stacked with small movement between them?

I will bear in mind the advice about flats at the same temperature - I tend to do mine in the morning.

[Thalestris24]

That was 20 30s subs, grabbed near the end of the lunar eclipse - I just noticed the Pleiades were high and bright so I swung the scope round before finishing up...

I suppose the sensor would have been quite warm after spending an hour and a half taking 255 shots of the moon back to back,the latter half all 15 second exposures...

A bit nervous as my Andromeda shows them (but much less noticeably) and that was 64 frames with ~ 20 each of darks, flats and bias.

I've just looked at stretched versions of my darks, flats and bias. They aren't on the darks at all, but very faint on the flats and quite noticeable on the bias frames. Perhaps this is because the bias frames get stretched the most. They look finer, and I wonder if the effect is worst when a relatively small number of frames is stacked with small movement between them?

I will bear in mind the advice about flats at the same temperature - I tend to do mine in the morning.

Hi

What ISO? And which camera was it?

Louise

[Davesellars]

Looks like a mismatch somewhere in calibration frames.   Only darks need to be at the same temperature it's not necessary for other calibration frames.  It is very important the darks are the same temperature, ISO and exposure as the lights.

Try stacking the lights them without any calibration frames and then add in the flats, then bias and finally darks to get to the problem.  For Flats it is important the focus and orientation is the same as the lights.   If you have any difference in temperature from your dark frames to the lights you'll just be introducing noise and you may get some strange artifacts if the flats were't taken with the same focus and orientation   Bias you can take any time and re-use.

[glowingturnip]

are the streaks exactly vertical in the frame btw, or have you cropped/rotated the above image at all ?

it is normal for sensors to have a vertical or sometimes horizontal pattern showing - i had a look at my own bias frames in post 13 in this thread - http://stargazerslounge.com/topic/252361-darks-and-temperature/?hl=%2Bsuperbias#entry2752496 Your streask look a little wider-spaced than mine though.

The same pattern should be present in all frames you take, so the bias should subtract it from your flats, and the darks should subtract it from your lights, so long as you're taking all your calibn frames properly ?

Amp-glow/sensor heating could be an issue though - I don't know your camera, but my old Nikon D80 (now retired from active AP duty) had terrible problems with this.  I reckon that patterns coming from sensor heating are non-linear so difficult to calibrate out.

[cjdawson]

That looks like camera noise to me. I'm guessing the ISO was quite high.  So yeah, I think this is a case for the full exposure set - subs, dark, dark bias, flat, flat bias.  Take 20 of each, but you can always take more subs ;-)  The Dark frames are the same exposure as the subs, but the rest are fast so won't take long to get the complete set.  (I think it takes me about 10 mins to get the full set of flat, dark bias and flat bias frames without rushing)

[Stub Mandrel]

This is a hoary old Canon D10 I got for £30, the shots in question were at ISO1600, which is high for it.

I've jpegged a stretched master bias which may or may not show the vertical banding (which is a hard to differentiate from the bayer pattern at some levels of zoom), but if it does they are similar to GlowingTurnip's.

cj - I had about 20 of each control frame for the andromeda pic and still got the banding.

I will try stacking with different combinations of the control frames and see if any combos reduce the noise.

I suspect that it will just be something to live with until I get a lower-noise camera.

[Leveye]

NIK Define removes vertical and horizontal banding in Jpegs quite well.

[D4N]

I don't think you are supposed to use darks and bias without flats.

I don't know what software you are using so don't know how it will respond to that but the read noise is contained in both the bias and the darks, you don't want to double subtract.

/Dan

Sent from my iPad using Tapatalk

[Stub Mandrel]

I use DSS and it glows red if you only use lights, calms down to orange if you use one or two of darks, flats and bias, and settles at green when you use all three. I have discovered it's better to leave out any one of them than use one with mismatched settings (other than a small difference in exposure for darks (e.g. 60s instead of 80s still works pretty well). Using the wrong flat has awful consequences....

[Stub Mandrel]

NIK Define removes vertical and horizontal banding in Jpegs quite well.

I use RAWs and TIFF for processing.

The horizontal bands in that bias frame don't seem to bleed through to the images.

[lensman57]

This is a sensor problem. If you could post a link to the original single Bias, stacked Bias and Darks this could be verified. Next time try and image at a much lower ISO setting and see if it makes a difference.

A.G

[Uranium235]

As well as the calibration frames mentioned above, you can also use noels actions to remove vertical banding (it has a filter for for vertical or horizontal), its very effective too. But no replacement for not having them there in the first place

[Stub Mandrel]

I've purchased Noel's Actions, fits my budget a bit better!

I ran the remove vertical lines macro twice, it isn't 100% perfect but it makes a significant improvement. Here's the result:

I'm even more chuffed now - one application completely removed the bands in my M31 Magnum Opus. Hopefully this jpg won't be to badly chewed: