This forum thread has been created for the b Persei observing campaign being announced 26 February 2016 in AAVSO Alert Notice 537. Please see the Alert Notice for details. Discussion and questions are welcomed here by the PIs and the AAVSO.
Elizabeth Waagen, AAVSO HQ
I'll observe b Per in the next two weeks from France, weather permitting. The forecast is somewhat pessimistic, and the last couple of months have been cloudy and wet here. So wait and see.
In fact, I was planning to observe this possible secondary eclipse since last year - your observations with the NPOI and prediction comes just very near mine, I was off by some 24 hours -, and it confort it.
I've been monitoring b Per for ten days around Christmast (it was just half period) so as to ascertain the ellipsoidal period ; the variation was very well defined.
I positively observed the eclipse last year.
This is a project I would love to be in. My actual setup is not suitable for such a bright star. But, I have access to something smaller, which I think will fit. I will do some testing in the coming days and let you know.
In an off-year, 2014, I did extensive observations of b Persei to map the ellipsoidal variation using a 135 mm FL ordinary telephoto lens that I was able to mount onto my CCD camera with a T-thread. The aperture was f/5.5. It gave a 3 deg FOV on the CCD's long axis and good photometry results.
I currently use an Orion 80 mm aperture guide scope - stopped down to 43 mm. It is gave good results for the eclipse observed in Jan, 2015. It is also important to de-focus somewhat to avoid under-sampling in a short focal length system. The Orioon 80 gives about a 1 deg FOV on the long axis. Finder scopes with CCD should also work well.
I've attached a plot of data from my set-up for b Persei for February, 2016. Lots of bad weather, but I observed last night. Last night's observations appear at phase 0.1 and 1.1 (repeat). The dotted line is the ellipsoidal variation fitted from extensive observations two years ago in 2014.
We should be observing b Persei from now (March 1, 2016) every opportunity for the next 2 weeks as we anticipate the first detectable secondary eclipse as the third star passes behind the inner two stars. The eclipse is predicted to begin March 7, 2016 UT.
For check stars please use any of the AAVSO sequence stars. If your field of view is too small, you may use HIP20296 as a check star. This star is different from the close star I had indicated in the alert notice. Sebastian Otero of the VSX team has learned of the small, long-term variability of HIP20370 - the non-sequence star originally proposed as a check star. Star "55" = HIP20156=000-BLL-386 (5.456V) from AAVSO chart X15836JD should be used as the comp star. See the attached chart with HIP20296 indicated.
Please send time-series reports both to the AAVSO and to me (firstname.lastname@example.org).
I have uploaded a number of observations for February and early March, 2016. The latest was last night (mar 5.0 , 2016 UT. Noisy data due to clouds. No eclipse yet. Keep watching and keep those pesky clouds away!
I'm on the target since more than a couple of hours (from 05/03/16 18h30 UT) ; some clouds, but the sky should stay clear until local midnight.
I'll post here if something happens.
last night I observed b Per. Like in the past campaign, I used an Orion Starshoot Miniguider (50/162 mm) and an Atik 314L+ with Johnson-V filter. This one provides me a very broad field (close to 3º). A pity I can't use the ST-8 with this combination because the precission would be higher.
There was no eclipse, fortunately, because there was some disturbing clouds and the night was, photometrically speaking, mediocre. But I noticed an incongruence in the magnitude of 55 Per: in the alert notice it is noted as 5.456 V magnitude, but in the web page
it is noted as 5.492 V magnitude. So, what is the correct one?
The correct identification of this star is HR 1330. 55 Per is something different.
Don's chart shows the Tycho2 VT magnitude of 5.492. The correct Johnson V magnitude (5.456) is reported in the Alert Notice. VizieR reports the Jonson V magnitude from multiple sources as being right around 5.46; how you combine them is up to the researcher, but for the purposes of a monitoring campaign like this, the important thing is that everyone either uses the same comparison star or reports the magnitude they used for whatever comparison star they chose. The researcher can then offset the submitted photometry to place everyone on the same scale.
Good luck on your observing!
Thank you Fran for pointing out this discrepancy in the comp star magnitude. Thanks to Arne for pointing out the source of the confusion. As Arne says for the campaign we should all agree and use star '55' (HR1330) V = 5.456) as the comp star in our reports. This is also the value for Alert notice for the 2016 campaign. This star was only recently added to the AAVSO sequence about 14 months ago. The older value used the Hipparcos value since this star was not available in the AAVSO sequence.
Weather looks good here in N. Carolina for tonight, March 6 EST (March 7 UT). We need all continents watching for the next week.
Don, Arne, thank you for the information about the comparison star. Here I have been observing this night until clouds covered the sky. My last points, obtained in March 7, 20 UT, shows that eclipse has not begun yet.
I have taken Data on b Per on two nights 5/5 UT and 5/6 UT from my location in central Texas. The first night I did not have an aperture mask. Therefore I must average a bunch of one-second EXPTIME images to minimize scintillation. That seems to be workable OOT but requires about 3-minutes elapsed time for each averaged image. I don’t think that achieves the desired temporal resolution during transit. Therefore last night, 5/6 UT, I used an aperture mask with three 1.25 inch apertures symmetrically located around the center of the with 120 degree separation. I used three apertures rather than on larger one with the same total area because I though the separation between the apertures would help minimize scintillation.
I make flats with and without the mask in place. It seems obvious to me that I need to use the flats with the mask in place, BUT those flats seem to exhibit pattern mottling that looks like an interference phenomenon. I have measured statistics over essentially the entire CCD surface (3000 x 2000 pixels centered at the center of the 3072 x 3048 pixel chip) and the standard deviation of counts for the masked flat is about one-quarter that of the one without the mask, and the Max-Min ranges as a percentage of the image means are comparable. The full set of statistics is attached and the sample flats from which they were derived can be downloaded from the following one-time WeTransfer URL:
Flat FITS images
My question is the following: Do I need to use sky flats (dusk/dawn) instead of EL screen flats? My initial conclusion based on the statistics is that do not. However, there may be considerations that are not evident from the statistics. I am looking for specific insights as to whether I have to replace the EL flats with sky flats for images taken with the mask in place.
Please I am trying to avoid getting into a theoretical discussion about the flat technique that is best in general. I am looking for specific information that supports or refutes the need to replace the EL flats with sky flats in this situation.
Thanks for any insight you have.
Brad Walter, WBY
I would guess that the smaller std dev with the masks is due to less vignetting than with the full aperture. I assume that the EL means an electroluminescent film. If you have a short focal length and wide fov, I think the electroluminescent film gives a better uniform flat than the sky. I've been using sky flats, but difficult to get without stars contaminating the flats. You reminded me that I can go back to the EL film since I have a small aperture system (stopped down to 43 mm) that the EL flat a point blank range should be good. Also try only one aperture opening. If its a refractor, cut the hole in the dead center. The mask is important to lengthen integration times to average out the sintillations.
I observed for 4 1/2 hours tonight until 11:30 pm EST (04:30 UT on Mar.7 UT). No eclipse yet.
Keep observing! Thanks for participating!
It's a reflector - Mewlon 250. I agree that the difference in standard deviation is the reduction in vignetting. that clearly shows up comparing the standard and masked flat images.
I Guess I should try only one opening. That would eliminate the apparent interference pattern - very "pincushion"- like appearance. I should have thought about interference patterns before making a mask with multiple apertures. The obvious is sometimes elusive before the fact. I guess my consolation will be that the comparison between the averaged short exposures, multiple aperture mask and single aperture mask will be interesting even if some of the data turns out to be useless.
Unfortunately, the weather last night and the forecast for the next 3 days gives me little hope for observing the secondary eclipse itself. For me this may all turn out to be only an interesting technical exercise.
I've thought more about the apparent "pincushion" effect. It may be due to the partial focusing of the three holes in the aperture plate on the front of a Newtonian. Of course the apertures are blurry, but each hole would cause its own vignetting pattern. More so at the front of a Newtonian tube than point-blank in front of an objective lens or mirror. Maybe other users have experienced this by trying Hartman masks. I would think that whatever the vignetting, proper flat calibration should take care of it.
If you're clouded out the rest of this week, the next b Persei event will be the "primary" eclipse in December 2016. Not a long wait. We'll need lots of observers for that.
Your comments mirror my initial conclusion when thinking about using multiple apertures in the mask. However, when I saw the flat, my reaction was, "oh s__t. Is this interference pattern going to foul things up, because the EL screen doesn't present collimated light but starlight is well collimated and the two may have very different patterns. I will be reducing images later today to see how the measurements derived from averaging about 15 full aperture images compare with the multi-aperture mask. and how both of thes compare to a single aperture mask images after the secondary eclipse.
Brad Walter, WBY
I had submitted some cloudy data on b Persei for JD 2457452 (March 3, 2016 UT). I recently removed the cloudy data. The excessive noise from the clouds gives a false impression of an eclipse. There has been no sign of an eclipse yet as of JD 2457454.69 (last night). Keep watching!
An observer sent me data from last night that shows a possibility of b Persei just coming out of an eclipse. We expect several ingress and egress events for the duration, and a single eclipse could last less than 24 hours. Wide world coverage is needed. Keep shooting this week even if it means shooting between the clouds on a partly cloudy night! The attached plot shows the possible eclipse egress at the left end of the plot as open circles. The beginning level is about 1.5 sigma below the other b Persei in the plot.
If we confirm the start of eclipses we'll give credit to the observer who hasn't posted results yet.
I have clear weather but won't be able to begin observing until after 0200 UT tonight (9 PM EST) so keep on shooting!
Many thanks for the data that is coming in to AAVSO on b Persei.
The light curve from the last 10 days which many have contributed to (attached) clearly shows a 3-day cycle. That is a beat between the approximate sinusoidal period (1.53/2 d = 0.765 d) and the approximate time between samples (1 day). The beat frequency arithmetic predicts about 3 days/beat cycle.
This means that we see the ellipsoidal variation of the light output but no eclipse yet where we expect to see the magnitude dip by at least 0.1 magnitude. Please keep shooting - even between the clouds on partly clouday nights. The detection of the first eclipse in 2013 was about a week after the NPOI predictions.
Yeh! I, Too recorded a dip in bPer brightness. 10 minutes ago at 2100 EST (0200 UT Mar 10) the b Per magnitude appeared to bottom out at ~4.82 V. Then the clouds rolled in. Normally it oscillates between 4.54 and 4.64 in the ellipsoidal. Congratulations to Paul for shooting between the clouds.
Will post my light curves on the forum later. Everyone who has gaps in the clouds, keep shooting!
I've attached the light curve of b Persei that shows the drastic dip in brightness tonight. Looks like its still going down at 0300 UT. The downward drop is observed by Paul Benni and I (CDK) by shooting between the clouds. I'm almost totally socked in and b Per is setting behind my building.
Hope the western observers can take it up from here.
Keep on shooting, we should have a couple of more days of in-and-out of eclipse!
I attach a graphic with my latest observations. This night I think weather will be good and I'll continue monitorizing.
I'm going to upload all my observations.
P.S.: I am currently observing. At 18:50 UT (March, 10), b Per is still in a deep minimum: mag. 4,94.
Just starting observation at twilight here in Dijon, France. First data point at 4.907 on 18 41 UT.
In past days it was oscillating between 4.58 and 4.63 or so, the depth is obvious now.
Sky is clean now, I continue observations up to clouds are back. Will report all my observations by tomorrow.
Clear Skies !
Since the announcement of the campaign, a group of students (including me) in Tel Aviv University have been observing b Per for a few days, about an hour per day. Attached are some preliminary results (not submitted yet) - for the first 3 days the weather was very nice, and on March 7th we had very frustrating thin clouds. This morning I was very glad to hear that we have not missed the eclipse! The weather here is clear and I'm observing again now. Hope to submit the data next week.
Paul, Don, congrats for your observations and prediction !
The first evening data for 2016 03 10, from 18h32 to 19h10 UT, suggest - as previously noted by Fran and Roger (Salut Roger !) - that b Per is still under deep eclipse, possibly slowly raising from 4.865, under not totally dark sky, to 4.857 at night.
Attached is the LC obtained in late 2015 december (to the left) and LC from 2016 03 05-10. Please note this is not exactly the same temporal scale in both LC.
Data were obtained from Valence, France, with a DSLR (TriG) on a FS102/f6 ref stopped down to 50 mm, 150 to 180 sec exp at 100 iso. Processing with Iris ; analysis using AstroImageJ.
The sky is clear now and data acquisition in progress. More on this tomorrow.
All data will be uploaded next week end, as soon as the eclipse will be over.
PS : the eclipse is sooo deep that I would suggest this one to be remaned as primary eclipse ;-)
It is really good to see all the eclipse data roll in and that several observers have more that they haven't posted to AAVSO yet! Especially from Europe today.
I had done a simulation last December predicting a light curve for the "secondary" eclipse where star C passes behind the AB components. Using the same luminosity parameters as the model that I fitted to match the 2015 data, the secondary eclipse is predicted to be deep (V ~ 5.0) and with a "flat bottom" . Each eclipse of the same C star by the different A or B components of course will be the same depth. More to come later when we get more data.
Keep looking for holes in those clouds!
Hello, It's a terrible night with lots of clouds in N. Carolina. Amazingly, we get some holes and good images. The magnitude of b Persei is still about 4.9 V at UT Mar. 11, 0000 to 0100. Shoot away!
I got some data tonight Mar 10, 0000UT to about 0245UT between clouds. The brightness is slowly increasing. The slow increase is probably simultaneously coming out from behind one star and entering behind the other inner star. The data is interspersed with clouds. I capture most everything and ignore the cloudy ones.
I understand there are a number of other observers on the sky this evening. Keep up the good work!
I wouldn't think it as really possible to caracterize a 0.3-0.4 mag eclipse drop directly to the naked eye, but it is achievable, indeed.
I had some clear sky for a couple of days just before the eclipse begun. My observing location is surburbia just outside a 50000 inhabitants town, where the limiting magnitude usually is around 4.5-5.0 at best. So, by 06/03 and 08/03 evenings, the sky was clear, after one week of rain. I could easily see b Per to the naked eye, if not directly, at least with slightly averted vision. By 10/03, as the eclipse was underway somewhere near 4.9 mag, I couldn't see b Per any longer, under similar sky. The star had vanished !
The 4 stars Lam Per (4.25), Mu Per (4.12), C Per (3.96) and D Per (4.80) were all well defined (the latter also with averted vision), but HD27084 and HD27192 (5.456 and 5.55) were not seen.
I encourage people to try the reverse observation before/after the star reappears. If someone succeed, please let me know ! I'm curious about someone experiment.
Attached is my updated LC. A minima of b Per is pretty well defined near JD 7458.42 at 4.87 mag . Coverage was from 10/03 18h30 to 11/03 02h00 UT, with target only 12deg high at the end of observation.
Tonigth should be clear.
The sky was not so bad here in Dijon, FR, last night, the first three hours were ok, next thick cloud come in and blocked the view for an hour, then I have been able to image through some holes and next the sky returned to clear, maybe some few haze. Observation had covered from 18:30 to 2:30 UT. The b Per mag has been essentially constant with some brightening at end. Start about 4.914, ending about 4.833.
The star was at very low elevation at end and I have to review the processing to check for some extinction differential at end, maybe also some secondary chromatic effect, reason for which I am not ready to report to the database. I will be on trip next days and not able to observe. Tonight should be possible if a big cloud on zenith decides to move !
As soon I am ok with the processing I will report. I have also to recover images that have been shot in between the clouds. The attached curve is the "raw" result from yesterday.
Clear Skies !
Three observations from Roger, Christophe, and Francisco from France and Spain show the minimum during the eclipse at ~JD7458.4. Nice work!
This is probably the second minimum in the eclipse judging from simulations and the structure of the 2015 eclipse. We observed then that we have multiple ingress and egress due to the third star passing either in front of or behind the rotating inner pair. The eclipse also may completely egress out of eclipse and re-ingress before all the activity is finished. I expect the last dip to end sometime Mar 13 UT, but we should definitely observe through Mon Mar 14 UT. The weather report looks very poor for my location so we need all scopes running on less-than-perfect skies!
The sky was not good last evening, with dense low altitude clouds which slowly vanished after 11pm local time ; the transparency was bad, with limiting mag near 3.0 at best. Nevertheless, I managed to catch a few data points from 22h13 to 23h45 UT. At this time, the brightness was between 4.6 and 4.55, possibly just under it's normal state (should have been near 4.54-53 – in my TriG system – at this phase of the ellipsoidal variation ? but my measurements may be off due to the poor sky).
Here is the updated LC.
The forecast for tonight is pessimistic.
Congratulations to everyone who has contributed to this campaign. We have observed the first ever "secondary" eclipse of b Persei by the third star. The eclipse just observed was deeper than the "primary" eclipses observed in winter 2013 and 2015. Some of us are wondering which is "primary" and which is "secondary". I suggest that we call the "primary" eclipse the eclipse in which star C passes in front of the A-B pair. Further analysis and matching with simulations should reveal which eclipse is which.
Please keep observing tonight - March 12 to13 UT in case my calculations are wrong. It is also important to make sure we are fully out of eclipse. Finally everyone should get a good night's time series out-of-eclipse so we can calibrate each observer's observations for the offsets as the inner pair continue their ellipsoidal variations.
Are there any observers in western N. America, Hawaii, east, central, and western Asia? We would love to have your data to fill-in the time gaps.
Please see the attached file for plots and more discussion.
Very interesting update.
I've send all the data to your email adress ; If you don't received them, please let me know. I will send them to AAVSO tomorrow.
By the way, I have data up to 58.58 ; maybe a bit off your needings .. just in case.
Just uploaded 9 observing sessions, each data point is in general the mean of 10 images from 10 sec to 30 sec depending condition. In several cases the air mass was very high at end of session, b Per should not be affected significantly but the check ( AAVSO 74 ) is. I will make a readjustment of it later.
The instrument was a Canon EOS M and a 400 mm tele-lens at F/8 or F/5.6 depending condition. A large defocusing was used but no significant blending should be involved as known from UCAC4.
Don, if you would like more details I can send you all the processing data, flux per image, flux stat, condition, other stars of the field, mag processing using VSF, mag stat, B-V
Nice work Christophe, sky was not that good in France those days, but ok we got it !
Clear Skies !
I wish to thank all observers who battled the clouds and the weather to shoot a successful recording of the 3rd detected eclipse of b Persei involving the third star. Let me emphasize that it is important that everyone record at least one good, cloud-free, time series run of b Persei post eclipse in the next week or two before it falls too low in the sky. These are needed to calibrate all of our photometry to a common zero point offset to fit good agreement to the ellipsoidal light curve out of eclipse. Please use the same comparison star for all your data and same optics and detector. Send all data to AAVSO. I will take things from there - applying box-car averaging to some noisy data, adjusting the offsets for alignment out of eclipse and putting it all together and sharing the results at this forum. I also will attempt to simulate the eclipse with the simple simulation I developed for last year's eclipse. And we'll see if we can determine primary eclipse (where the third star passes in front and blocks the inner stars, or whether the third star passes behind the inner star and is blocked by the inner stars.
For the December 2016 eclipse we need to recruit observers from Western N. America, the Pacific, Asia, and the Middle East. I believe there will be data coming from those places to help fill-in some gaps.
I'll be on a trip for a week beginning Mar. 16, but will have e-mail and will be visiting the forum.
Thanks again! Clear skies!
I just submitted to the AAVSO some b Per data, B and V, from March 11 UT. My process is not very automated, hence the delay. I have 6 more data points from March 11, which I should complete by tomorrow. I also have 3 additional points from March 12, but there were thin clouds.
John Centala, CQJ
I have finally processed the approximately 5000 observations that folks have submitted to the AAVSO database on b Persei for the March 2016 eclipse. A high-resolution processed graph is shown in the attachment as a pdf file showing 10 days on a 30 inch long format with the ellipsoidal variation superimposed. I have "boxcar" aberaged the time-series of observations that were quite short integration times in order to smooth the atmospheric scintillations. I have also offset each observers' observations such that the out-of-eclipse matched the ellipsoidal variation. In a few days I hope to have a model of the occulation light curves plotted that I used for the 2015 eclipse presented as well.
As can be plainly seen (also from the AAVSO light curve generator) is that this March "secondary" eclipse is deeper than the Jan, 2015 "primary" eclipse, which was somewhat unexpected. However this result is consistent with the model parameters that I used for the Jan 2015 eclipse. In that model for 2015 the third star C passed in front of the A-B pair. Star A was assumed to be the brightest and largest of the three stars. When the smaller C star partially occults the larger A star, a smaller fraction of total light is blocked than when the smaller C star is completely eclipsed by the the larger A star. This interpretation is also consistent with the apparently flat bottom of hte eclipse at day 458 seen by the European observers.
Please check your observations using the legend that is placed on the graph and let me know if there are any errors or omissions. I omitted the observers that did not have out-of-eclipse observations, or there were huge discrepancies with any other data.
The previously observed eclipse summaries and light curves are easily read at my web page: http://inside.warren-wilson.edu/~dcollins/bPersei/
There are more features to discuss such as the differing ingress and egress slopes, the apparent duration of the March 2016 eclipse from the Jan 2015 eclipse (attributed to missing observations from the lack of Pacific and Asian observers).
Again, many thanks to all of you who observed this event, obtaining out-of-eclipse and excellent coverage during the recorded ingress and egress as well as the flat minimum. Another "primary" eclipse is coming in mid-December, 2016.