Advice required: how to take flats with an OAG and no rotator

[jazzysam]

Hi Guys,

can anyone advise on how to get the necessary flats when using an OAG?

That is, I'm having to rotate the imaging train to pick up guide star, - if I change target thru the night, I lose the precise position of the imaging train prior to taking dawn flats.

Any thoughts/experiences?

clear skies,

Sam

[redfox1971]

Hi,I'm sure you mean lights?.

[opticalpath]

I'm afraid it's a fact of life: if you change the camera orientation with respect to other optical elements in the path, you will have to take new flats with the camera in the new position.

For example, if the camera orientation changes in relation to a filter wheel, any artifacts caused by dust specks on a filter will appear in a new position in the image frame, so previous flats will not remove the artifact. Also, a vignetting pattern is often not perfectly symmetrical, so again won't be well corrected by flats taken at a different orientation. Rotating the camera also means you have to recalibrate the guider every time. Ideally, try not to change the orientation of your camera too often - certainly not for every object - if you can avoid it.

What camera are you using in the OAG guider, and what guide exposure are you using? Have you tried binning the guide camera 2x2, and experimenting with longer exposures? That would give you more potential guide stars and help to reduce the need to rotate the field. Also, sometimes just a small lateral movement, decentring your subject slightly, will be enough to bring a suitable guide star into view. That could be a compromise worth making to avoid rotating.

For what it's worth, I use a Lodestar camera in my OAG, binned 2x2 and with 2 second guide exposures; this is on a 1000mm focal length scope. I have never had to rotate the camera to find a suitable guide star with that setup.

Adrian

[jazzysam]

Hi Adrian,

thanks for your feedback.

my setup is also a Lodestar on an SX OAG. Like you, generally I have no probs finding a guide star when imaging targets in the Milky Way. The problem is that I'm trying to image targets in Leo (M58, NGC4535, etc), where there are very few background stars. Hence my need to rotate the imaging train to find a guide star as per planetarium program.

I guide binned at 2, on a 1625mm RC, adjusting every 10s. I'm using the Avalon linear mount, so it requires much less guide adjustment. I have tried slight lateral movements in the OAG train, and extending guide time to 20s; but there's nothing much around to capture...

i thought that the situation was not something that I could address without a rotator, or conceivably by taking flats before moving the train by using an artificial light panel. In any case, do you have experience of either?

cheers, and clear skies,

Sam

[opticalpath]

Hi Sam.

I can see that your situation is more difficult than mine, given your longer focal length and narrower FOV; especially so in sparse regions.

You mentioned that you extended the guide time to 10, then to 20, seconds (i.e. the interval between successive guide corrections). I'm assuming that this isn't the same as the *exposure* time on the guide camera and that there is a separate setting in your guider program for that. That's what I meant when I was suggesting increasing the guide camera exposure time - perhaps to several seconds - in order to detect fainter potential guide stars. What guide camera exposure time do you use normally?

Sorry if this is stating the obvious and you already tried all this. If that's the case, your only option may be to take new flats whenever you change the camera orientation. I use an EL panel for flats and that certainly makes life easier. If your telescope is in an observatory, you can even take the flats in daytime with an EL panel or light-box, as long as you can darken the observatory and cool the camera to the same temp. as your imaging run.

The only other thing I can think of that might help is pre-planning your imaging sessions to try to minimize the need to rotate often. Using a planetarium program to display not only the imaging camera frame, but also the offset FOV of your guide camera, superimposed on the star map, would allow you to see in advance what would be framed in different orientations and to decide what angles (or small offsets) would be likely to give you acceptable object framing and suitable guide stars. Then if you're hoping to image, say, two different objects in one session, you may be able to plan in advance one orientation that will work for both. Again, sorry if you're already past this point!

Adrian

[ollypenrice]

You might get away with one set of flats for all images. It depends on how well centered your camera is in the scope's light beam. If it is well centred then the vignetting won't change much after rotation. The dust bunnies will all be in the part of the train that you're rotating so they'll turn with the train. If the light cone isn't hitting the middle of the chip, though, the vignetting will shift and muck up the flats.

There is another possibility, though it might need a bit of setting up. Rotate the entire scope in the tube rings. That way your flats won't change at all. If I were to try this I'd want some little blocks fixed to the tube to stop it slipping down through the rings and messing up the balance.

Personally I hardly ever try two target in a night but that's partly because I have low horizons and partly because I do image very seasonally - ie I start a project when it's low in the east.

Olly

[steppenwolf]

i thought that the situation was not something that I could address without a rotator, or conceivably by taking flats before moving the train by using an artificial light panel. In any case, do you have experience of either?

This is the solution I would go for using an EL panel - quick and reliable.

More than one object per session? - that would be a dream for me but I need all the time I can get on each single object!

[jazzysam]

Hi Guys,

thanks for your responses.

Based on all the feedback, i think an EL Panel is going to be the way to go for me, if I'm forced to change orientation thru the night. I spoke with Ian King about it, and that was his suggestion too (I'm waiting for him to receive a batch so he can send me one).

Just to clarify; the 10-20s timing on guide exposure I was referring to, was the actual guide camera exposure time. I use a planetarium program to identify guide targets with the superimposed FOV of the guide cam, and it correctly tells me where to rotate the imaging train to.

i did attempt to 'save' the imaging position for each target, but found that its impossible to replicate accurately at dawn/any other time to take flats. I'm in an observatory situation, so can replicate conditions when I have to. However, I found that dust bunnies in the RC train rotate relative to the imaging chip - I ended up introducing more errors when applying flats the next day, quite obviously because the train wasn't in precisely the same position as the imaging subs had been ( ie a 1 pix offset is sufficient to cause you to end up with double dust bunnies).

i even tried manually rotating the flat master - I just couldn't get it accurately enough where the pixel math would work...

Ian King also suggested I look at a Pyxis rotator (in theory, it can replicate rotation angles accurately enough). Does anyone have any experience of one?

Cheers and clear skies,

Sam

[opticalpath]

Sam, the Lodestar is pretty sensitive as you know and with 20 sec exposures, binned 2x2, I'm quite surprised you're not finding suitable guide stars more easily - even in these sparse regions. I would have estimated that 20 seconds would give you a decent signal from stars as faint as 12th or 13th magnitude, and there's usually enough of them around even in quite a small FOV. Don't get me wrong, I'm not doubting what you're seeing - I just thought you would be picking up more stars with 20 seconds exposure.

Have you by any chance deliberately defocused the guide camera slightly (for better centroid determination)? I suppose that might fuzz out fainter potential guide stars.

First chance I get, I'll point at one of the targets you mentioned and see what I get with 10-20 secs. on the Lodestar. My FOV is 60% bigger and not directly comparable but will still be interesting to see.

Adrian

[ollypenrice]

I hadn't picked up on the 20 second guide subs. I pick them up at a second at F6.8, almost invariably, though I guide in 4.5 seconds with the Mesu. Not many mounts could guide on 10 to 20 second intervals in my experience, though if yours does then great!

Is the turret far enough down into the light cone? And, yes, is it focused? What's the F ratio and FL of the scope? If you put the turret on the long side of the chip it can go deeper without shadownig the image.

Olly

[opticalpath]

Olly, the OP mentioned that he's imaging with an RC, 1625mm f.l. With the Lodestar, that would translate into a FOV of about 13.5 x 10 arcmin and I would have expected 10-20 sec exposures at 2x2 to be throwing up a LOT of faint stars with enough signal to guide on.

If the OAG prism was offset too far out, could that be the cause of an apparent drop in sensitivity of the guide camera? ISTR you mentioning something about this when you were setting up the OAG on Yves' scope.

Adrian

[ollypenrice]

Olly, the OP mentioned that he's imaging with an RC, 1625mm f.l. With the Lodestar, that would translate into a FOV of about 13.5 x 10 arcmin and I would have expected 10-20 sec exposures at 2x2 to be throwing up a LOT of faint stars with enough signal to guide on.

If the OAG prism was offset too far out, could that be the cause of an apparent drop in sensitivity of the guide camera? ISTR you mentioning something about this when you were setting up the OAG on Yves' scope.

Adrian

Sorry, I missed the scope details.

Yes, if the prism is not in far enough the light cone will be too tenuous. I think this is a likely explanation. In fettling Yves' OAG I kept lowering the prism till it made a slight shadow in the edge of the image then backed it out a bit. The further in you go the more light you pick up. That's why the prism should go into the cone on the long side of an oblong chip. It can go in further without obstructing the working part of the cone.

Olly

[jazzysam]

Adrian/Olly,

apologies for not responding sooner; out of it for health reasons over the last few days...

wrt. Your questions, I have positioned the SX OAG prism as far as it will go; ie on the long side of the SX H694 chip, so that I'm getting as much light as possible from the light cone. I found its easier to do it that way, since you can get Maxim to display the FOV for both the main and guide cameras in the planetarium (to work out where guide targets might be). I have focused both cams as sharp as possible (I know that some people defocus guide cam for better guiding, but I don't do that).

ive found that for m66/m58, I had to rotate the imaging train in order to pick up any guide stars. For M66, I could just about see a guide star without rotation, but the star was very faint; for M58, I had to really hunt around to find any suitable candidates. The only other factor that might be relevant, is that right now, M66/M58 are at approx 45 deg - I guess not ideal from an imaging standpoint.

Adrian, if you do manage to find suitable targets easily, please let me know so that I can pin down what I'm doing wrong...

clear skies,

Sam

[opticalpath]

I'll have a look as soon as we get a glimpse of clear sky, Sam (when, if ... !)

It won't be an exact comparison because my FOV is bigger than yours, but I'll also try to see what magnitude are the faintest stars I can guide on with 5, 10, 15 sec exposures and that should give an indication if you're getting comparable sensitivity at least.

Adrian

[jazzysam]

Thanks Adrian, I appreciate it..

kr,

Sam