Affiliation
American Association of Variable Star Observers (AAVSO)
Mon, 03/30/2015 - 20:55
Dear colleagues,
Can anyone explain what is 'Visual from Digital Image' format what is used for, how to make such photometry and what are cons and pros for it?
Thank you in advance!
Hi Andrey,
"Visdig" means to display a digital image on a computer screen, and then make an estimate from the display like you would visually (interpolating between two sequence stars). You can usually get about visual-estimate accuracy in this manner, depending on many factors such as the contrast of your screen. There are a handful of people making such estimates, and the AAVSO added an observation type to accommodate them. It is strongly recommended that you use standard photometric software to make your measurement instead of the computer screen display, as you will gain far greater precision.
Arne
[quote=HQA]
You can usually get about visual-estimate accuracy in this manner, depending on many factors such as the contrast of your screen. There are a handful of people making such estimates, and the AAVSO added an observation type to accommodate them. It is strongly recommended that you use standard photometric software to make your measurement instead of the computer screen display, as you will gain far greater precision.
[/quote]
I've used it a few times from BSM images, mainly when I cannot do visual, usually because too much lunar interference. It is less accurate and less precise than visual. You are typically looking at the size of a "smudge" on just a single digital image, you can't adjust for "flat", etc., so my confidence in estimation is quite a bit lower than with direct visual. Maybe 0.2 or 0.3 magnitude of uncertainty, vs. 0.1 or better by direct visual. So, this method is better reserved for "fainter thans" than actually trying to estimate a magnitude by "eyeballing". Unless this is the only way you can do estimation, for whatever reason.
Mike
I use it almost exclusively for fainter-thans associated with known or potential recurrent novae monitoring, where I take widefield DSLR images. Very occasionally I'll post observations where there might be a paucity of any data (eg early obs of new novae, particularly southern ones). Case in point the recent PNV J17291350-1846120 where I had pre-discovery images taken two days before the discovery.
But that's rare. I don't think there's any point in submitting regular observations of variable stars by this method as, frankly, no one's going to use or be interested in the data - or see them for that matter as who thinks of clicking Tri-G when generating a light curve?
But you can increase the accuracy by firstly ensuring that your exposures aren't over or under saturated. Secondly restrict the method to widefield DSLR images taken with lenses unless you want to go the route of darks, flats, biases etc in which case you might as well go the whole hog and do photometry. Widefields ensure that, in the scale of the image, the comp stars are very close to the VS and in similar areas of vignetting or whatever. Thirdly always take multiple exposures and stack them to even out frame-to-frame differences (and there can be significant differences frame-to-frame!). For my own interest I've done many personal light curves on novae using this method (not submitted to the AAVSO database other than a few several years ago) and have got very nice, consistent results that follow the visual trend well and have little scatter.
But that's neither here nor there. It's good that the AAVSO has allowed for this type of observation nonetheless.
Cheers -
Rob Kaufman, KBJ
Thank you a lot for helpful thoughts! But I found that it is impossible to plot a light curve with just, for example, CV measurements made with CCD, not "visual from screen", because Visual from Digital Image observers also select CV, as filter.
The idea of VFDI is interesting, but I think there are too many uncertainties and nonlinearities, it is practically impossible take into account systematic error of the videoadapter+monitor.
I've been using VfDI for the last 4 years using the Slooh Robotic Observatory on both the Canary Islands and Chili and I admit, at times, it may not be as accurate as visual but it's the only way to estimate fainter nova, super novas, and other variables from my very light polluted site. Or to observe variables in the southern skies that I cannot get from Long Island. I do find it does work better on fainter stars. I still prefer visual observing but on some stars this is just impossible.
Rich (TYS)
Hm,
This ViSDIG option is new for me, although might be useful in some cases. Some of our observers made JPG image of the Nova Sgr 2015#2 using DSLR cameras. They might submit the VisDIG estimates of the nova this way.
But my question: what "filter" code needs to be used this case what they estimate the brightness from a colours DSLR image?
Thanx,
Robert
I take my images and split them into their Red, Blue, and Green components. I make my estimate using the green component. I use the Tri-G filter code.
Rich (TYS)
Thanx, but my question was related to the cases when only jpg image is available (other observers, not me).