An exploration of star colours in live captures

[Martin Meredith]

Following on from a discussion on an earlier thread about star colours in live RGB imaging, I thought it would an interesting way to spend an evening looking at this in a more controlled way by choosing stars of comparable magnitude and situated at a similar altitude but which differed in colour index. I'm using the common B-V colour index, which corresponds to the difference in magnitude measured between standard 'B' and 'V' filters. Everything else being equal, stars with a low or negative value will be blue-ish, and those with a high value will be red, passing through white, yellow and orange on the way. 

I chose 14 stars, in the main equally-spaced along the scale, but I was also keen to capture some extremes. The Hipparcos catalogue (which is reasonably complete to mag 11 or so) lists B-V values ranging from -0.4 to 5.5. Since I was constrained to choose stars that were both of similar magnitude and visible at the same altitude on a given night, I couldn't necessarily capture the entire range. However, I managed to find a set with the lowest at -0.29 and the highest at +4.4. They were all in the range 65-70 degrees of altitude at the point of capture (a logistical nightmare!) and all have magnitudes in the narrow range 10-10.25. However, some of the most intense red stars are variables.  

This is the result. All are based on 4x10s subs in each of RGB Baader filters through a Skywatcher Quattro 8" f/4 reflector, using a Lodestar X2 mono camera and StarlightLive v3 to do the combination. They are all captured/RGB combined live at the scope with no post-manipulation other than resizing in Gimp and annotation with star designation and colour index. All StarlightLive controls were held constant throughout. The hue control (lower slider of the colour modifiers) was set to zero, and saturation (upper slider) set to 1. I also captured each star at saturations of 0.75 and 1.5 but haven't assembled them yet. I allowed 10 minutes per star (I needed to know this to 'schedule' the star to reach the right altitude). Luckily this was about right. GOTOs based on inputting RA/Dec were not always very accurate as the scope was swinging around the sky a lot, so a lot of field-checking was needed...the whole thing had the air of a mini-astro-marathon i.e. stressful but fun! 

Unfortunately, the scope started to dew up for the final 5 captures, which are those in positions 1, 3, 5, 8 and 12 above. The effect on star size is obvious, but the colour still comes through.

One outlier (at position 7 from left) is HIP 79412 which has a greenish hue. Looking at the whole 'plate', some of the other stars have a green tinge, so I suspect this is a case of either me selecting the green channel more than once, or some colour misalignment. In any case, I don't think it is a reliable result.

It looks to me that even at mag 10, the stars are saturating at 10s exposure, but fortunately the colours are coming through on the 'skirts' of the stellar image. I'd like to repeat this with shorter exposures.

All star fields were confirmed using Simbad after the event (I picked up one mistake but fortunately the correct star was also in shot...)

Some individual notes

(1) HIP 86986, also known as BD+39 3226, is a type sdO hot sub-dwarf which has lost most of its hydrogen and is now burning helium; there are very few of these brighter than mag 12; discovered as a hot sub-dwarf as recently as 1978.

(3) HIP 105693 (BD+35 5412) is a confirmed member of the NGC 7063 open cluster

(4) HD 337512 is a double but isn't mentioned in any catalogue that I can find so is either optical and considered uninteresting, or just hasn't been spotted. Intriguing. The peachy colour of the other component provides a nice contrast. I'd estimate separation at around 5-6 arcseconds.

(7) colour issue mentioned above

(8) HIP 103067 (BD+42 3895) is the binary star COU 2542 whose separation is just over 0.1 arcsec.

(11) HIP 98123 (BD+58 2015A) is a nearby high proper motion star lying at a distance of 103 light years

(12) VX Cyg is a delta Cepheid variable which varies from 9.5 to 10.5 over a period of 20 days

(13) WX Cyg is a Mira type variable with range 8.8-13.2 over 410 days.

(14) V CrB is also a Mira type variable carbon star with the range 6.9-12.6 over 358 days, first referred to in 1899. It looks to be near maximum in this shot. This is definitely worth seeking out optically at the moment! It lies almost on the meeting point of Hercules, Bootes and Corona Borealis. 

I conclude that keeping the standard hue and saturation settings in StarlightLive does a great job of representing the entire range of star colours, for a mono camera and RGB filters, and that's what I intend to do from now on (two fewer variables to manipulate live is always a good thing!).

cheers

Martin

 

BTW in case anyone is interesting in tracking these objects down, this is my observing plan showing the locations, other Hipparcos data, and order of observation. The breaks didn't materialise...

[DHEB]

Very interesting post. Thanks!

[aparker]

Great work as always, Martin.  I have asked for an RGB filter set for my birthday, which fortunately comes along before next galaxy season...  This is good information to have as I transition off of summer (narrowband) imaging.

[RobertI]

A fantastic piece of work Martin, very interesting and informative.  The green is an odd one. I think I have convinced myself that there are no green stars in reality (because a blackbody follows the planckian locus, which misses green, as it changes temperature) and that any green stars observed are illusions caused by effects such as observing a white star close to a red star. So I guess as you say the green must be due to something else. But its good to know that the standard colour settings in SL give a faithful rendition of colour, as you say, makes things simpler. Thanks for all your efforts to produce this.

[Martin Meredith]

Thanks Cinco, Alex and Rob. 

I'm fairly sure it is some error on my part due to the other greenish stars in the image. I'll revisit when I get a chance (maybe tonight).

I've been thinking more about the colour settings and what matters really is the relative hues across the BV range; saturation is more of a personal choice. As an example, here's a wider field for the sub-dwarf very blue HIP 86986 (top right) with saturation values of 0.75, 1 and 1.5 from left to right. I've expanded the field to include a rather nice orange/yellow double.

 

This is the same for the reddest star V CrB

After that evening's work I'm now more convinced than ever that for EAA a motorised filter wheel is the way to go; 14 stars = 28 filter changes (going RGB in one direction then BGR in the other to save one change) is a lot of opportunities for inadvertent field shifts or slight focus changes...

Martin

[RobertI]

I've been pleased with my motorised filter wheel and I certainly couldn't do without it as my scope is some 30 feet away. It's a five position Xagyl filter wheel and is not compatible with SL (at least I haven't tried it) so I use the ASCOM test program which is fine. There is still some possibility for error, in particular having to manually stop after X exposures every time, but so far so good.

[Ain Soph Aur]

Very interesting post.

The SX USB filter wheel has spoiled me. I briefly owned an Orion Nautilus USB filter wheel, but the SX USB wheel is in another league. Highly recommend!

[Astrojedi]

I second Brandon. The SX filterwheel works very well for real time imaging with SLL. I have the mini USB one with OAG.

[nytecam]

Martin - well researched and present - well done  

 

Nytecam

 

 

[Martin Meredith]

Thanks Nytecam. It was a lot of fun and something I intend to repeat (with a different selection of stars), perhaps under more scientific conditions. I now have the motorised filter wheel from SX so it will be a lot easier next time too. I didn't save the FITs unfortunately and it is likely that the colours/saturation could be improved by different settings, so this is another reason to repeat the exercise. 

Martin