Determining Carbon Star Magnitude - Direct v. Averted

American Association of Variable Star Observers (AAVSO)
Tue, 06/19/2012 - 19:25

I have been working on the AL Carbon Star certificate and have had a bit of trouble determining magnitude with the red color shift. I generally try to look at an area equidisant to the target star and nearby stars to get an averted vision take on the whole group. I know averted vision adds brightness so this seemed logical (I hope, been observing for a few years but new to the variable scene).

This has yielded some pretty decent results when the stars were roughly the same color. However once you get into to 10+ region with some significant color the carbon star won't come into it's own until you stare directly at it for a couple seconds. The star will literally appear to brighten in the process. Try doing that to a standard white star and you get the exact oppisite effect.

A really good example of this from last night was V CRB. Direct vision put it in the 9.5 region (max) but averted suggested something more like 11.5 (min). I checked the light curve this afternoon and it showed roughly the same ambiguity.

I know our eyes take a bit for the red photons to process and stars themselves vary in peak spectra. Maybe the averted/direct technique could be a biological light filter? That would be kind of cool if the case!

Anyway, just curious how the more seasoned observers out there tackle the problem. Thanks in advanced!

Clear skies,

Cody Lawson

American Association of Variable Star Observers (AAVSO)
Carbon star visual estimates

Very very interesting topic.
In fact, carbon stars are the most difficult ones to estimate visually.
You have described a problem variable star observers live with.
The Purkinje effect which is the transition between cone vision and red vision. Cones being colour sensitive and rods being color blind.

Although averted vision is very useful for deep sky hunting, it is not so in variable star estimates, unless we are talking about detection only. If you work at your limiting magnitude and need to see if you can detect a faint CV in outburst, you use averted vision. If the star is very faint, you use averted vision. But doing so you should be careful with color differences.
Averted vision may work well in cases where all stars (comps and variable) have the same color. Actually if the stars have the same color, it doesn't matter how you look at them as long as you do it the same way.

In the case of very red stars like the carbon stars, if you look at them avertedly you may even fail to detect them (rods being color blind) so you should use direct vision.
The V filter peak is closer to the cone vision sensitivity so my advice is always to use direct vision BUT without staring. You already know what happens when you stare at a very red star. Take a look at it, long enough to see it clearly but not so long as to start seeing it grow in brightness.
If you use averted vision the result with be VERY different from V and will depend strongly on the comparison star colours.

Anyway, even though you take all precautions you may imagine, the errors in carbon star estimates will be larger and the results will be more difficult to compare with V. The amplitude is usually larger in the visual because when the star is at maximum, it overstimulates the cones but when it is at minimum it gets difficult to detect.
Personally I apply a correction to my visual observations depending on the star's magnitude. It becomes positive when the star is very bright (e.g. +0.1-0.2 mag. for a 6th mag. star through 7x50 binoculars [talking about V Hya and R Lep here]) and negative when it is faint (e.g. -0.2 mag. for 8th mag. stars with the same binoculars). It varies depending on colour. I've made those corrections based on contemporaneous V data found from the literature (e.g. ASAS-3).


Bright Red Stars

Great topic on ways to observe Red stars, like the caarbon stars. I remember Sebastian and I discussing this at the 100th anniversary meeting in Boston. Sebastian, I remember you sharing your technique by applying a correction.

My method is called the "out of focus" method. It has been long used by visual observers. When you are observing the star, defocus the image a bit so all stars in the field loose their color and of a bit fuzzy. This helps avoid the Purkije Effect that Sebastian talks about. A good example of this would be observing some real red stars when they are bright, such as WZ Cas, R Leo, R Lep, W Ori, and TX Psc. 

This works much better for me than the quick glance method. Some folks like the quick glance better. I find that very uncomfortable on the eye.


Chris Stephan

Robert Clyde Observatory

Sebring, Florida  USA

American Association of Variable Star Observers (AAVSO)

There is always a personal preference regarding these matters. As long as you are able to produce accurate results, it should be okay.
The same way as you don't feel comfortable with the quick glance method, I'd never recommend the out of focus method. The star is red and if you defocus it, it will get dimmer in comparison with stars of different color. We need a strong stimullus to activate the cones. I think that the out of focus method will end up increasing the difference between the visual results and V and that is the opposite to what I want to achieve.
Have you make a comparison of your results against V or even between your own results using a defocused and a focus image? That would be a nice test.


American Association of Variable Star Observers (AAVSO)

I do the same thing, as red stars that are focused will appear brighter then red stars that are defoused. Unless the variable is very faint I defocuse all the variables I observe.

Rich Tyson (TYS)

American Association of Variable Star Observers (AAVSO)
Handling Very Red Stars

Well, this is a tough topic! Visual estimates of extremely red stars is probably the hardest job to do accurately.  It might be best to leave it to CCD and its standard filters. But, if you MUST visual estimate, I could suggest one approach.

First, lets take a look at a typical one - R LEP. It has a B-V~+5 which makes Miras look "white" by comparison ;)  This corresponds to a Teff of less than 3000K, which means a peak spectral emission around 1 micron or so! So, visually you are really only catching the "tail" of the star's output curve, which pretty much goes to zero at 500nm and below. But, 500 is the peak response wavelength of our rods. So, using rods/averted vision is not a good plan for these Carbon stars. You will be operating at the extreme low tail of your rod response curve, and response will vary a lot depending on degree of aversion from the center, and your rod density. This will result in a large variance in measures between individuals, how you observe, and also depend on how bright the star is vs. aperture of your instrument. A real mess.

Certainly, to estimate these type extreme red stars, its better to use your cones. That means pretty much looking directly at it, and the star needs to be at least 5 magnitudes above your instrumental limit, to start activating your cones. And, preferably even brighter than that, because the cones red response is several magnitudes weaker in the extreme red than at its peak of 550nm. So, basically, you need to use the largest aperture instrument possible.

For R Lep, at its minimum around mag 11, you would need at least a 20" scope for it to appear luminous enough to start activating your cones. But at maximum, even binoculars would be sufficient. So, I think part of the problem with the scatter in the data for Carbon stars that vary a lot, is observers use the same instrument, which becomes inadequate at the star's minimum. 

Of course, this is not a practical solution for many observers, to get a large telescope just to observe Carbon stars at minimum! So, here is my idea.

Since the response of the rods and cones becomes EQUAL at wavelengths redder than about 650nm (even though both are responding weakly) how about use a sharp cutoff red filter at around that wavelength? A Photographic Wratten 29 (deep red) filter should work really well, since its a complete cutoff below 610nm, with a sharp rise into the deep red and IR. This would guarantee equal response of your rods or cones, and allow either direct or averted vision to be used.

This red filter will also make the normal colored comp stars appear fainter, so it will raise the magnitude estimate of the carbon star, counteracting the "too faint" visual estimate problem as well.

I haven't yet tried this approach, to see if it works as well in practice as it appears it would from theory! But at least its an idea which might work to help solve the extreme red visual estimate problem.

Mike LMK

American Association of Variable Star Observers (AAVSO)
Estimating Carbon Stars


Appropriately- red means caution.

I only use averted version when at my instrumental limit. I use it when otherwise I wouldn't detect it. Make sure you use the code "V" when submitting this type of observation. It signifies that you just glimpsed the star 

When using direct vision I only take quick glances. If the red is very intense I will defocus the image or go to a smaller aperture- i.e. finder or binocs. Either way- I look away, "Reset" my eyes, and take several estimates. 

One of the most important things is to not to outsmart yourself. I have tried to "Calibrate" my eyes to make corrections  as others have mentioned but at least in my case the results were inconsistent so I don't do it.  Unless you are very confident in your calibration I wouldn't try it. Just do the best you can. You will get better with experience

In addition don't withhold your estimates because you think they are off. Unless you know why they are off you are better off submitting them and leave the vetting for the folks at AAVSO who are used to that sort of thing.

American Association of Variable Star Observers (AAVSO)
stars to observe - and not

A good solution is to only observe stars that aren't over-bright in your scope. If you like to observe 'red stars' (by this I mean M-type and later, irrespective of whether they actually appear red to you) there are a lot of them out there, and many are under-observed. You don't say what scope you are using, or what power, but (to give a personal example) as a rule I wouldn't observe any red star brighter than about magnitude 10ish since these appear too bright in a 36cm aperture. I'd simply leave it alone in the hope that an observer with a smaller aperture was looking at it. Otherwise I'd be contributing to a skewed light-curve.

I recently started on the 'neglected Miras' programme, and had a look at SV And on Monday. It was about magnitude 13 so I logged that observation. If it had been (say) at magnitude 8 I simply wouldn't have reported it.

V CrB is very red, and because of that it's good for small apertures. I used to observe S Cep (even redder) with a 6" refractor but streered clear of it when it was at maximum!

American Association of Variable Star Observers (AAVSO)
Carbon Stars

I use two methods to observe red stars:

If the star is bright, I defocus to compare the star with comparison stars.

If the star is faint, I use averted vision, or, if the redness is not aparent, I use stardard observing techniques.