Ngc3621

[peter_4059 (Peter)]

Despite the moon I managed to get some imaging in over the past two nights. Last night I was experimenting with dew control but managed to capture 25x2min subs using the DSI II and Newt. All the usual problems including a horror gradient and noise but it's good to finally have some clear nights!

[allan gould]

Nice one Peter, I resisted imaging this as well when you told me what you were up to. Came out really well but the stars look just a little "eggy" or is that just me? Nice just to get out and do some imaging but from the forcast that looks about it for a while.

[kinetic (Steve)]

Peter, that's a lovely smooth result.
Well done. Stars are pretty good for that exposure
length and FL.

Here's my one and only result from ages ago with
the DSI II at 1524mm and only 5 sec subs.
Yours is way smoother and far more detail.
Also, check out the same mag 18 stars in your image.

My image on left, yours on right.

Steve

[peter_4059 (Peter)]

Allan - it wouldn't surprise me if the stars are a bit off as I'm still trying to track down the flexure and have a bit of drift in the 2 min subs. I'm planning an experiment with a differnt guide scope mounted directly on top of the newt to see if that fixes the problem.

Steve - thanks for the comparison image. It's amazing what you can get out of entry level equipment from a light polluted location and a near full moon! I've been discussing your approach (lots of short subs)with Troy and Allan - I think the DSI really benefits from a lot of subs and may revert to more shorter subs myself - especially if I can't isolate the source of flexure.

Peter

[kinetic (Steve)]

Originally Posted by peter_4059

Steve - thanks for the comparison image. It's amazing what you can get out of entry level equipment from a light polluted location and a near full moon! I've been discussing your approach (lots of short subs)with Troy and Allan - I think the DSI really benefits from a lot of subs and may revert to more shorter subs myself - especially if I can't isolate the source of flexure.

Peter

Sure Peter,
The only benefit I get is the stars are sharper and I sort of beat the seeing.
However, in this case, my image is certainly a bad
example of the method. Your subs are 2 mins, mine 5 secs. Look
at your stars. They are awesome.
On exceptional seeing nights, my method really works well but the
down side is I have to go longer to get extended source faint stuff.
There is no point me ignoring that.
It's hardly my method btw, it's just a carry over of the ten odd years
where I squeezed every last ounce out of 8 bit webcam images by
hundreds of subs and extreme processing.
As you say, it works extremely well for the DSI II.
I will never part with this camera, I am looking forward to winter with
it's many more good nights.
I love the image scale of these also with fast newts. Without even
reading your imaging details I knew the result was from your DSI II
and 10"

ps: If you're interested, I can pm you the processing routine(s) I use
for good seeing nights/ average seeing nights for DSI Mono work.

Steve

[peter_4059 (Peter)]

Thanks Steve - I'd like to see your processing steps. My image has had some pretty severe processing to deal with the noise and gradient. I focused by eye on Friday night rather than with the Bahtinov mask and the stars were a little soft so I've used some mild deconvolution to tighten them up a bit.

I'm the same with the DSI - I really enjoy using it and have been eyeing off the one for sale with the outback cooler!

[troypiggo (Troy)]

Good one mate!

[RobF (Rob)]

Nice image Peter. I'll have to check this one out. I see there's a stack of galaxies to explore in Glen C's "All Sky Messier Catalogue" I'll have to get into, including this one.

Just a thought about your "flexure" - are you sure its not just mount and guiding software manifestations. I've spent about 15hrs this week getting my head around guiding with Maxim to see what it offered. After 3 nights, I turned it loose on a series of 10min subs on Eta C only to find pretty bad smearing mainly in the RA direction. However, when I push Guidemaster (which I usually use) to 10 mins and beyond the majority of my smearing is in the DEC direction. Phd is never quite as good as GM even when I spend hours optimising it.

I know my HEQ5 is hardly a Paramount, but its interesting how different settings on different software can manifest in such different ways.

Anyhow - even though you're on an EQ6, just suggesting lofting the Newt around may require some specific tweaks in your guiding software. (Actually, I think your stars look pretty damned good in that image too anyway)

[peter_4059 (Peter)]

Rob,

I'm yet to get to the bottom of it but I'm seeing drift from frame to frame on the image yet the guiding keeps the guide star in the reticle - I can only explain this as differential movement between the guide camera and the imaging camera. My drift is typically in RA although it depends where the scope is pointed.

[RobF (Rob)]

Perhaps try programming PEC?

Agreed though if your star is stable (and you have sufficient magnification on the star) more likely to be mechanical.

[peter_4059 (Peter)]

Originally Posted by RobF

Perhaps try programming PEC?

Have you got this working with EQMOD?

[RobF (Rob)]

Only this week after lots of fiddling around. Too early to say if it helps. In theory should allow less aggressive guiding settings.

Used Toucam + PECRecorder (freeware)
Then fed about 5 cycles into PECPrep, and smoothed to get curve. Unfortunately with EQ mounts you have to park religiously after each session as the mount encoders don't give a readout of their position (other than relative to parked position). I can usually park successfully for many sessions in a row, but I'm sure I"ll have a crash now PEC is running

My understanding of the theory is the PEC should help adjust out RA inaccuracies in the mount, and if you're very well polar aligned that leaves autoguiding to adjust DEC (and any remaining RA error). PEC can really at best offer an average of the RA worm error - there will always be cycle to cycle variations, but overall max deviation should be less.

I did have it running in my initial Maxim efforts where majority of error appeared to be in RA mind you - as I said, insufficient data to prove benefit here as yet. Would be happy to try and help if you give it a spin and get stuck on something Peter (I suspect there will be many others more knowledgeable on the subject anyway).

[peter_4059 (Peter)]

I've generated the PEC curve after hypertuning my mount to check my remaining PE however the thing that has been putting me off using it is as you say the religious parking requirement. I assume it is still ok to release the RA clutch as this only releases the last element in the drive train. How big was your PE?

[Rastas]

Realise just how frustrating this can be trying to sort out the flexture problem - but just in time for Easter -"eggy" stars

[allan gould]

Peter
You have cursed me as I never had significant flexture until now, and boy do I have it now. After one and half hours there are significant trails from hot pixels in my shots with the DSI but DSS gracefully removes them whereas the darks dont do that good a job. Ive found if I dont really tighten up my guidescope to the Losmandy rail then its worse and I think this is where my faults lies although it appears as solid as enything.
When the QSI comes and I can guide through the imaging scope Ill be interested to see if this "cures" my problem.
Ill bring the camera around and we can trial it on your scope to see if it improves yours singnificantly.
Allan

[RobF (Rob)]

Originally Posted by peter_4059

I've generated the PEC curve after hypertuning my mount to check my remaining PE however the thing that has been putting me off using it is as you say the religious parking requirement. I assume it is still ok to release the RA clutch as this only releases the last element in the drive train. How big was your PE?

I really hadn't looked too closely as it was very late on the night I configured this. Have just fired up PECPrep again on the same data and run the autofilter settings.

Before smoothing (but with Linear Regression which I'm guessing allows for any DEC drift) for 5 cycles (each about 10mins on HEQ5):
Peak PE about +/- 15 arcsecs
RMS 5.22 arcsecs

After Autofilter:
"Peak" down to about +/-8 and RMS 3.3 arssecs

I've attached some pics in case you understand it better than I do (which wouldn't be hard). I'm pretty sure clutch releases won't affect the Parked position of the steppers. One of these days when I have time (and the guts) I'll try and tweak down the DEC backlash. Hypertune sounds nice, but only EQ6 I believe alas.

Raw data
Data treated with Auto Filter button settings
PEC page after Auto Filter (PEC bottom right)

Rob

[Jen]

nice Peter