2nd daft question of the Week - Flipped Images

[Marky1973]

I'm set up with Artemis and my 414ex to do the RGB images to add to the lum i took earlier in the week. Set up in the same way, PA, 2 star alignment - CAPELLA and then DUBHE, slewed to M81....start looping in ARTEMIS and the image is flipped along the vertical axis....but I can't figure out why. I realise I can flip the images before stacking ect, but that will be a pain, so put me out of my misery - why are they slipped like that - everything is the same....

The image below has a sub from the LUM from earlier in the week, the second is tonight's test and then the last is tonights flipped vertically....

Help!

Thanks,

Mark

[Davey-T]

There is a setting to rotate 180 degrees and correct aspect ratio,not sure what they apply to though.

Dave

[Marky1973]

I've been clicking everything and can't figure it out, Dave...it probably isn't the end of the world, just one of those things that has me scratching my head! :-) 

Will check settings again.

[Uranium235]

Yep, check the rotate 180 function in artemis. If you have that on it can upset stacking and render your flats next to useless (if you dont rotate them back).

Hmmm... although, if it were rotated then the image would not be just flipped vertically, try opening the images with another package.

[Marky1973]

Hi Rob - played with the rotate 180 and that does just that, but doesn't affect the vertical flip. I have no idea what is going on and, now you mentioned flats etc, which I have spent all day doing in preparation, I have no idea if they are in the right orientation as well. I am sure it is something simple, but it winds me up no end. I can't bring myself to carry on now as I have no idea what orientation my flats were in and I don't want to waste time in the cold.... think I might try and find another hobby! :-)

Oh, I only seem to be able to open the FITs files in PI, but they are saved straight from ARTEMIS with no manipulation.

I was wondering if the scope could have slewed "the other way around" but this wouldn't make any difference to the vertical flip surely? And I have checked the original subs against the stacked LUM in PI and there is no change in orientation.

[Marky1973]

Thus isn't something that could be caused by the filters is it? I use an IDAS LP filter for LUM and then baader RGB filters....presumed they would all work in the same way, but you never know....

[Thalestris24]

Hi

I don't know Artemis but with APT you can flip the preview images displayed on the capture screen but it doesn't affect the saved image. Not sure if that helps.

Louise

[Marky1973]

Hi Louise. I had hoped it was something like this; something I had done in the settings or in the processing, but the saved images were coming out in the same orientation as the preview in ARTEMIS! It has me stumped... so much so, I've given up in a huff to drink rum! :-)

We need an ARTEMIS expert... do we know any.....**COUGHOLLYCOUGH** :-)

[Thalestris24]

I plugged it in google and it came back with this old thread amongst other references

[cjdawson]

Looking at the images, I don't think it's a rotation issue, if it was, the triangle of the bright stars to the left of the DSO would be on the right in one of the images.  The asterism is always on the same side.

As for the image being flipped vertically.  The only thing that I can think of is that a star diagonal might have been added to the optical train.  That will have the effect of flipping the image.

It would surprise me if it was something that obvious.

[geordie85]

Opposite sides of the meridian maybe?

[Marky1973]

10 hours ago, Thalestris24 said:

I plugged it in google and it came back with this old thread amongst other references

Thanks Louise, I must have missed that one! I will have to look for that setting, as I could only see the rotate tool, which wouldn't help, time to get the manual out if I can remember where I put it! ;-)

10 hours ago, cjdawson said:

Looking at the images, I don't think it's a rotation issue, if it was, the triangle of the bright stars to the left of the DSO would be on the right in one of the images.  The asterism is always on the same side.

As for the image being flipped vertically.  The only thing that I can think of is that a star diagonal might have been added to the optical train.  That will have the effect of flipping the image.

It would surprise me if it was something that obvious.

Definitely flipped CJ, there is no diagonal, but that was kind of why I was wondering if there could be a similar effect caused by differently manufactured filters in the light path?

10 hours ago, geordie85 said:

Opposite sides of the meridian maybe?

I wondered if the Meridian might be a problem, but both sets of images where taken after 8pm this week - M81 was well past the Meridian at this point... just one of the those really annoying, probably very simple, things! :-)

[Thalestris24]

Hi

It won't be filter - only a lens can vertically flip an image. A setting is the most likely cause...

[Marky1973]

Good to know.....just been looking and can't see anything obvious in Artemis to flip the image, only rotate. Guess it will just be a case of trying again next clear night and seeing what happens....just worried about the orientation of flats etc, not knowing WHEN the flip may have occured and whether they will work....mind you, no-one said it would be easy..... :-)

[Thalestris24]

I don't know Artemis, but I think there's something that allows you tell it where the origin is (bottom left or top right) and that determines how the image is written

hth

Louise

[Marky1973]

Thanks Louise - can't find anything, but hopefully someone will point me in the right direction - thanks for your help

[Thalestris24]

Hi

It's to do with the 'origin' stored in the fits file header. If you look at the headers (fits liberator) you should be able to see if the x,y values have changed between flipped and not flipped. Apparently, different software interprets the origin values differently or maybe ignores them.

[Marky1973]

Just opened one of the original FITs from the LUMINANCE set and FITS LIBERATOR opened it with "FLIP IMAGE" already checked, opening it in the same orientation as the RED Channel image - so I guess I could take the images and just re-save unflipped if needed.....just still doesn't help me understand why the images were changed from one night to the next in ARTEMIS.... :-) anyway, my head hurts now, so I'm off to the cinema to take the kids to see MOANA! :-)

Try, Try and Try again!

[Marky1973]

Here's a thought, could the flip be caused, perhaps, by viewing TIF files and FITS files because of the different information in the file headers?

[Thalestris24]

1 hour ago, Marky1973 said:

Here's a thought, could the flip be caused, perhaps, by viewing TIF files and FITS files because of the different information in the file headers?

Certainly could!

[Marky1973]

I think it is that. The subs from earlier in the week are different in orientation to the final stacked image in PI. I think I may have realise this quicker if the markers I had saved (to "plate solve") in Artemis had saved (or, at least, I had actually saved them) and then I would have seen that the preview orientation in Artemis last night was the same as the previous session! So I think that is solved and there should be no issue with the flats....woohooo!

[ollypenrice]

The 'rotate camera' function in Artemis is handy but needs to be handled with care and I try to avoid using it in the dead of night with only 30% of normal brain funcion available! In my opinion it is a software rotation and will, if used, require rotated darks, bias and flats. I use it only to reframe after a meridian flip. I don't plate solve so I use the screen markers, but once I'm re-aligned I go back to the 'not rotated' setting. Or I do if I don't forget, which is why I don't really like using it! For the same reason I never use the preview download speed in Artemis either. Sure as eggs, I know I'll forget to re-set it if I use it!

We did once end up with a mysteriously bad set of images and this turned out to be because the 'rotate camera' button had been inadvertently applied by the user. Once correctly calibrated the data was fine.

Glad you're sorted.

Olly

[Marky1973]

Thanks Olly - I generally don't touch any other settings than those I really need - so not gone anywhere near the rotation settings etc. and will leave well alone in future. I am pretty certain it is just a flip in PI when the FITS files are opened in the software, because of that "checked" setting in FITS liberator. The preview looked like it was round the wrong way because I was referring back to the preview in ARTEMIS for the second session. If only my markers had saved properly I would have shown myself that the image preview was fine. I am pretty certain now that that original preview when taking the LUM would also have been the reverse of the loaded subs, I just didn't remember.....so 30% is probably a bit optimistic for me....the most frustrating thing is that the clouds appear to have set in for the week and the moon is, of course, getting bigger!