Flats and Meridian Flip

[Kaliska]

Just trying to get my head around this, but when I'm taking images before and after the meridian flip, what do I do with my flats as I don't rotate my camera? And I guess it will be the same for my darks as well ?

[Ibbo!]

They should still work.

[michael8554]

Flats are to do with "imperfections" between the front of the OTA and the imaging chip.

You could take the ota and camera (together) off the mount and take your flats indoors (in theory).

Same with Darks, take the camera off the OTA, cap it up, and take your Darks the next day, if you have setpoint temp.

So flip doesn't affect either.

Michael 

[Wiu-Wiu]

You should do your image calibration with these flats and darks, THEN flip them. 

 

As said above: the imperfections aren't flipped, your image is. By correcting before flipping, you will get rid of them before stacking. 

[pipnina]

If you don't rotate the camera relative to the telescope, your flats and darks will still work after the flip. Remember that the stacking process takes all your darks, bias and flat frames, and it stacks them to create a master bias, a master dark and a master flat.

It then applies each master frame to individual light frames before they are merged into the final image. in short, they are still aligned with the way the optics and camera were laid out until after the darks, flats and biases are applied.

No need to worry, insofar as I know at least. (always got to add the disclaimer, in case I'm talking nonsense without realizing it)

[michael8554]

On 15/08/2018 at 22:51, Wiu-Wiu said:

You should do your image calibration with these flats and darks, THEN flip them. 

As said above: the imperfections aren't flipped, your image is. By correcting before flipping, you will get rid of them before stacking

I have a fork mount so I'm finding it hard to visualise a flip.

You're saying that the camera is inverted after the flip?

So if there's a bunny at top left, and a galaxy at bottom right before the flip, then after the flip the bunny and the galaxy will be in the same corner?

So how does DSS handle this?

Michael 

[spillage]

When DSS maps the images it will automatically rotate each image to align with the stars. As mentioned you only need to redo you flats if you move your camera positon in the focuser.

[michael8554]

Mark, I meant how does DSS handle Flats being different orientation to pre and post flip images - that's if the camera is inverting, which you havent confirmed ?

Michael 

[Wiu-Wiu]

Good you mention the bunny and the galax. After you flip, the bunny will be at the lower right. But flats and darks and bias are meant to correct your optical train’s flaws. Which have also rotated. So if there is dust on the right sideof the bunny, the same dust spec will be left of the bunny after the flip - it also doesn’t move in respect of your telescope, camera sensor,... 

 

you can even image a whole other object, or if you wish to make a mosaic: image the ears of the bunny for example, the dust won’t budge, the picture will be completely different, but the flaws - which we want to correct - are still in the same spot relative to the center of the sensor. 

 

When you calibrate your frames, the software will use the flats, darks and bias to make some sort of “master magic cleanup wand” and clean all of your images. 

Only after that, the software will proceed to flip them,align them by stars, and stack them accordingly. 

That’s also what happens when you dither: every frame is taken slightly offset of the previous ones, to even more dilute the effect of noise. 

[michael8554]

Still haven't said whether the camera flips, I've no idea being a fork mount user.

If it does, then the galaxy will move corners on the sensor.

If it does invert, for DSS to stack the Lights it will have to invert some, so the stacked image will have two bunnies.

How will the Flats correct both?

Michael 

[Ibbo!]

If the camera is not moved in the focuser everything will work fine with just one set of flats.

 

The darks and bias will work even if the camera is moved.

[ollypenrice]

Yes, the flip has no effect whatever on either darks or flats. The stacking software knows nothing about how the camera was orientated on the sky, it just knows what it considers to be the top, bottom, right and left sides of the chip. Whether the bottom was physically at the top after the flip doesn't concern it! 

The only way to go wrong here is to do a software image rotation made possible in some capture software so that reframing after the flip is easier. I never use this option because it can lead to misorientated calibration files. Instead I open a pre-flip image for reference, rotate it 180 on the screen and frame to match that. I don't save the 180 rotation when I close the reference image.

Olly

PS Be careful to remember that the mount does a 'flip' in that both axes rotate through 180. But the result on the image is not what image software calls a flip, it is just a 180 rotation.

[Stub Mandrel]

On 16/08/2018 at 23:23, michael8554 said:

I have a fork mount so I'm finding it hard to visualise a flip.

You're saying that the camera is inverted after the flip?

So if there's a bunny at top left, and a galaxy at bottom right before the flip, then after the flip the bunny and the galaxy will be in the same corner?

So how does DSS handle this?

Michael 

DSS will subtract the flat from every frame before stacking so the galaxy only get's the dust bunny removed from the images where it is on top of it.

DSS will automatically rotate the post-flip images after applying control frames and before stacking.

[michael8554]

Thanks Neil, it hadn't clicked that the flats are applied before DSS rotates the post-flip images - doh!

Michael