I don't know why you mention the Purkinje effect for this star since it is not red, actually is colorless and indeed looks pretty white compared with the 47 and 48 comp stars which are orange.
There is no linear trend from night to night at the same site and same time of the night, I think that the nova is varying continuously up and down so that may be accounting for a fraction of the 0.5 mag. scatter you see. And the rest, I agree it must be related to differential extinction. There may be people observing from Northern latitudes where the nova is never at a reasonable altitude for estimates.
I have estimated it slightly brighter than 4.9 last night and tonight.
Thanks David & Sebastian. I looked at the nova through a small telescope at low powers the other night and it definitely looked red (orange) to me, while that was not obvious either naked-eye or in binoculars. Photographically it's quite red when properly exposed (unfiltered DSLR). I mentioned the Purkinje effect because I was seeing some quite bright estimates from similar latitudes and wondered whether it might be a influence. My measure last night was very close to yours Sebastian (~0.1 difference) but I don't think the brighter estimates I was getting through staring were an illusion. Must admit I didn't notice the 'effect' naked-eye.
I certainly agree with Rob that there is some sort of effect causing the visual to diverge from the photometry. None of the photmetric measurements went near as bright as 4.5 in the most recent rebrightening (which has already reversed to a decline).
Talking about the colour of the nova is a little tricky. Most recent B-V (mine, last night) was 0.14. So quite blue. But of course the most recent V-R (three weeks ago!) was about 0.7 - very red! Not surprising given the contribution Ha and Hb are making to the overall visual flux. Given the relative brightness of Ha:Hb in the spectrum I took last night, my hunch would be V-R would be even larger now.
How this plays out with reference to the Purkinje effect, I don't know. The fact that the recent visual estimates are up to .5 mag brighter than V could be real, and a consequence of the Ha spiking in the last few days. We really need someone who owns an Rc filter to take some measurements. I may just have to order one myself. In the name of scientific necessity ;-)
Getting back to Purkinje - I've definitely noticed (or imagined) over the last week that as I get more dark adapted, the brightness of the nova increases relative to the comp stars.
It's worth pointing out that very few photometric observations for the last peak have so far been submitted, so can we really conclude anything much about visual vs photometric observations for that time period?
The attached plot shows visual, V, and Unfiltered with V Zeropoint for the nova.
Hi Rob, all
I wrote a blog post about observing the nova from Stockport Observatory last night.
Rob, I used both an unaided and binocular approach with short observations of the nova and I'm confident of the 4.5 observation.
As soon as I heard about this pink hue to the star, it cast my mind back to another nova I monitored for a brief time at maximum many years ago.
Nova Muscae 1991 also appeared to have a very strong 'rose' tint to it - something I'd never seen before or since. I wonder if any of our visual observers here had a similar experience with bright novae at or near maximum?
And thanks everyone for the discussion on this. I've followed it with interest.
HMm. On 13 Jan Arne and I got a timeseries over the same period. It shows the same decline, but mine are consistently around 0.05 brighter in V. Up till know I have been very pleased with the fit of my observations with the BSM South data - we've been observing at almost identical times.
I only used 2 comp stars in my ensemble on 13 Jan. I might redo the reducation using my usual four and see what happens. It might mean I capture less of the timeseries; two of the comp stars had drifted out of the usable part of the field by the end of the run, which is why I didn't use them. But if it improves the accuracy that would be better.
There seems to be an interesting increase in the rate of oscillation. SOme of the observations, especially visual, are leaping around a fair bit over the last few days, but using the polynomial fit tool in VStar shows a plausible pattern - oscialltions of a period around 1 to 1.5 days. See the attached light curve.
I've used V, visual and unfiltered with V zero point data here. Very approximate, but the data is still so thin that I needed to use all three.