Globular clusters and nonlinear modes in LodestarLive

[Martin Meredith]

Globulars are one area where I've yet to see EAA matching the eyepiece view in a big scope, perhaps because of the relatively low spatial sampling frequency of the Lodestar camera (big pixels, undersampling), so I occasionally experiment to see how close I can get to the visual view. I was out viewing a few weird faint things last night but trained the scope on M5 to end the session. Since it was still on display in the LodestarLive app this morning I took the opportunity to experiment with the linear and non-linear modes at a more leisurely pace.

My favoured approach at the moment with bright globulars is to stack short exposures. 

I attempted to get the best representation of the cluster in each mode (linear, arc sinh and square root) while setting the sky darkness to a similar level in each case. In the linear mode it turned out (as expected) to be nigh on impossible to get a non-blown-out core and still preserve the true extent of the globular. The only way to get close to the full size was to increase the contrast (i.e., widen the histogram) and set the white level at a lower value. This results in core burn-out. The top two panels show these two linear approaches. 

For both non-linear modes (lower panels) the problem disappears, with stars resolved right to the core yet showing more or less the full extent of the globular. Actually, I used another criterion too. The reason the cluster is off to one side is to permit a few faint galaxies to be shown, as indicated. The brightest is mag 15.9 IC 4537 while the others are fainter, as marked. You can judge for yourselves how well these are represented in the two non-linear modes. For the high contrast linear case the two brighter galaxies are visible. I'm not sure about the mag 18.1 galaxy. 

This is a really clear example of how the non-linear modes in LL are essential in getting somewhere close to a visual view of a bright globular cluster. 

Note that although I ended up with a large stack of 10s subs, this is simply because I left it stacking: the view doesn't alter much after the first few as this 4 x 10s shot below indicates:

Oh, and never mind M13's propellor. Has anyone noticed the smiley in M5? :-)

cheers

Martin

[Martin Meredith]

BTW I should add that this I'm just following in Don's footsteps from a month or two ago:

http://stargazerslounge.com/topic/241729-night-of-the-globs/

Martin

[Dragon Man]

That's a good result Martin.

You've got the stars looking quite small and sharp.

Yes the pixels at around 8um and larger in our cameras don't help with Globs, but you have done well.

This is the best I have got stars in a Glob so far with the NTSC Xtreme but when I get my PAL Xterminator it will help as the pixels are slightly smaller and square.

Keep going with Globs, there's plenty of them out there 

[RobertI]

That's a really good representation of what can be seen in terms of stars visible in core and outer regions, very pleasing to the eye.

[Dom543]

Martin,

Thank you for doing the experimentation and sharing the results. Your globs look very nice.

I have only a very limited experience with the non-linear DR mapping options, as I am addicted to colours. But it seems to be a must for several brighter galaxies.

Clear Skies!

--Dom

[Martin Meredith]

Thanks Ken, Robert and Dom. I was quite surprised at how sharp the stars are. Some of that is due to the excellent live stacking, I think. I've a nagging feeling the scope isn't quite collimated properly too, so things might get even better (some nights I just want to get out and observe rather than tweak the secondary… actually, I'd like to get rid of the secondary altogether, but that's another project). 

Dom, while I didn't mind too much handling the colour controls, since focusing on the mono Lodestar I find the whole process more relaxing. But we are entering that time of year when colour is crying out to be captured, so perhaps I'll switch back.

One thing I've noticed with the nonlinear modes is that there is a clear sensitive region which I imagine corresponds to the place where the slope of the nonlinearity is steepest. By sweeping the brightness control upwards slowly I find that at a certain point small changes in the slider create large shifts in the histogram. It seems important to avoid setting the brightness lower than this. I think the effect would be to compress the levels into too small a range, resulting in posterisation. Paul: I wonder if it might be useful to show the nonlinearity superimposed on the histogram somehow?

The other thing I've found is that the 'best' contrast setting is nearly always quite low, say in the bottom 25% of its range. Any higher and the histogram becomes quantised (i.e., instead of a smooth peak one sees a lot of individual peaks). So another suggestion would be to reduce the range of contrast and thereby give finer control.

But these are second order tweaks: the controls at the moment are demonstrably sufficient to significantly reduce burnout, improve details in larger galaxies, and help stellar/galactic discrimination for smaller galaxies.

Martin

[arkosg]

Thanks for the explanation Martin! I'll have to give it a go next time using your hints.

Cheers,

Greg A

[HiloDon]

Excellent report and images, Martin. Your experience is the same as mine. I find in the linear mode I have to increase contrast more and it is more stable in the histogram. In the non linear mode I usually find it best to just leave the contrast in default and move the histogram with the brightness, then move the black and white levels toward the peak to get the best result. I usually move the black level very close to the peak to give a nice dark background. Once you do it a few times, the process gets really easy.

I was out getting some galaxies the other night and saw M3. I like these globs far more with Paul's non linear scaling. Thank's, Paul.

Don

[Martin Meredith]

Thanks Greg and Don. I agree about the stability of the linear mode. The other modes are a bit like driving a fast car (I imagine; or say a nice light road bike) -- very responsive! Nice work on M3 by the way. Short subs definitely seem the way to go, and make even fewer demands for those of us who haven't yet graduated from alt-az!

Martin

[Dragon Man]

Great result Don!

Nice sharp stars.

[nytecam]

No camera has the dynamic range of the eye on bright DSOs like globulars so the camera is always playing catch-up!  However for gx, where you can see little more than a blur,  the magic of the camera comes into play

With globulars, in the attempt to reveal the faint outer stellar members, any stretching routine enlarges the stars so they may look soft but this usually can be restored with 'smoother/sharpening' rountine.    I like your globular image Ken but, for me, it's oversharp with a grid-like patterning and square star. 

[Dragon Man]

. . .    I like your globular image Ken but, for me, it's oversharp with a grid-like patterning and square star. 

yeah, I zapped the sharpness up to try make it look better but it killed it 

What do I care, if it looks like rubbish the viewers say so and I adjust to suit, until they say 'that's better'.

That's what's good about 'Live' broadcasting. It doesn't matter if it's not 'Imaging' quality, and changes can be made instantly, and the viewers can have input.

Bumped mounts, moving views from wind, star trails, etc. All part of the fun and the experience. 

Gotta love it! 

That's also why I said that is the Best I have gotten with the NTSC.

The PAL will be different.