Very inspiring work. You must be much sought after by the media.
I'm just a beginner, but do work with digital imaging. could you explain the subtraction concept a bit more?
when you overlay images of alpha and betapictoris, does the visible area around alpha pictoris extend over the dust disc of betapictoris or is it just the black area in the middle of your image?
what I'm curious to understand is whether the dust disc became visible as lobes only after subtraction and if so how there are no subtracted or black areas as a result of the subtraction outside your central subtracted area?
would you be posting both images before subtraction?
can the same be done with a dslr?
Just taken a look at your telescope truss design with ultra thin spiders and no mirror clips which is inspired thoughtful improvement as I am getting sick of the clip shadows around bright stars. Beta Pict gives a clear clean image as a result of this.
Brilliant bit of work, Rolf. You deserve all the accolades and them some
I was thinking...you know that the GStar mono camera has a very good IR response. I wonder how it would go at attempting such an image?. You might want to have a little look into trying one out and see if it works. A mono Flea would also be good. Speaking of IR pass filters, you could try Astrodon's IR pass filters. They seem to do the trick.
Cracking job there Rolf. You always seem to image the way out of left field stuff and it is always of special interest. When people ask me about imaging I always say there are people out there that image things that most people would not think of; and you always come to mind. te point being that pretty pictures is not always the goal of the astro imager. Fantastic achievement and it deserves the accolades you have obtained so far.
Hi All, thanks again for your comments. I have completed a new image with new data and created another thread for it here: http://www.iceinspace.com.au/forum/s...590#post793590
It's much better than this one so please have a look if you like. I have also written up a lot more details so some of the questions I have been asked should be covered over there I think.
Quote:
Originally Posted by NorthernLight
Hi Rolf,
I´ve seen it right after X2 on 3 last night and was stunned by your result.
Your dedication is truly inspiring and I wonder what else you are going to unlock from the domains of the MOA and the like.
Are you a member of the Auckland Astro Society?
Thanks very much Max, I'm not a member but I would like to be, need to look into that
I might get involved in microlensing research, it's very interesting and it would be great to contribute to science in that way.
Quote:
Originally Posted by alistairsam
Rolf,
Very inspiring work. You must be much sought after by the media.
I'm just a beginner, but do work with digital imaging. could you explain the subtraction concept a bit more?
when you overlay images of alpha and betapictoris, does the visible area around alpha pictoris extend over the dust disc of betapictoris or is it just the black area in the middle of your image?
what I'm curious to understand is whether the dust disc became visible as lobes only after subtraction and if so how there are no subtracted or black areas as a result of the subtraction outside your central subtracted area?
would you be posting both images before subtraction?
can the same be done with a dslr?
Thank you Alistair The black area in the centre corresponds to the saturated are where the signal reaches the well depth of the chip. Even after subtraction there will be no information hidden there. I'm not sure of the scale of the ESO image since there are no reference stars visible, but my guess is it's about the same size as the light blue circle.
Yes the dust disc is only visible after subtraction. In the new thread I'm showing the difference images etc so you can get a better idea about what I did.
Quote:
Originally Posted by John Hothersall
Just taken a look at your telescope truss design with ultra thin spiders and no mirror clips which is inspired thoughtful improvement as I am getting sick of the clip shadows around bright stars. Beta Pict gives a clear clean image as a result of this.
John.
Thanks John, yes the wire spider made a huge improvement to my star profiles. I don't know if a wire spider is absolutely necessary to see the dust disc, I wouldn't think so, but it certainly helps. It all depends on the combination of scope and camera I suppose. The ToUCam is fairly good at picking up IR light too.
Quote:
Originally Posted by renormalised
Brilliant bit of work, Rolf. You deserve all the accolades and them some
I was thinking...you know that the GStar mono camera has a very good IR response. I wonder how it would go at attempting such an image?. You might want to have a little look into trying one out and see if it works. A mono Flea would also be good. Speaking of IR pass filters, you could try Astrodon's IR pass filters. They seem to do the trick.
Thank you Carl, glad you enjoyed it. Yes I believe other cameras can do this as well, it's probably just that nobody thought to try before. I'll definitely get an IR pass filter at some stage, probably together with a new CCD camera.
Quote:
Originally Posted by Paul Haese
Cracking job there Rolf. You always seem to image the way out of left field stuff and it is always of special interest. When people ask me about imaging I always say there are people out there that image things that most people would not think of; and you always come to mind. te point being that pretty pictures is not always the goal of the astro imager. Fantastic achievement and it deserves the accolades you have obtained so far.
Thank you very much Paul I have had to make it my mission to image the more obscure stuff because my little chip cannot compete with the modern ones. But I have learned a lot from it and enjoyed it immensely. When I buy a better camera then I'm sure I'll continue in this direction, there are lots of strange things to image
I've done a few more sets on a few more nights, pretty much with no change in my imaging train.
Seeing has been better on two occasions to my initial sets. Still no result for me.
The moon is in the way now for a few weeks so I hope to use the down time to try a few things with the secondary/ spider.
You may be right about the spider vanes.
It's certainly not the primary obscured by clips, and I feel it's also not due to collimation, or collimation
change rather, due to OTA flex when slewing from Alpha to Beta.
See the attached Airy disks of out of focus stars near Sirius (low declination) and stars near the SCP (high
declination).
Almost no change.
I really want to get to the bottom of this and hopefully get a positive result from my rig.
I'm reluctant to turn the entire OTA 10 or so degrees to take the diffraction spikes further away from the PP Disk PA.
And I probably won't if the tests with an alternative spider show no
improvement.
I might even give one of the SC1.5 webcams a go too!
Also, it has been remiss of me not to apologise.
I did raise doubts, I followed them up with my own observations, and
until I get to the bottom of it, I really need to apologise for firstly
hijacking the thread, and secondly doubting your results.
I never doubted your abilities BTW Rolf, only your results.
It's also interesting the list Dennis has attached to your better result thread.
I noted the PA, inclination and angular size of a few other PP disks.
It looks like we could have some far easier targets for amateurs, based
on your result with beta Pict.
I might select one from that list and give it a go with the Mewlon 180 if the weather ever clears in Brisbane. I haven’t had much success with Miranda, one of the fainter moons of Uranus, because of the distinct diffraction spikes from my 3-vane spider (6 spikes) and the plane of Miranda’s orbit always seems to place it on one of the vanes!
Hi Steve and Dennis, I received the following advise from Karl Stapelfeldt, chief of NASA's Exoplanets and Stellar Astrophysics Laboratory:
Quote:
Originally Posted by Karl Stapelfeldt
Let me advise you a bit on which disks to try in the catalog. Many of them are only detected at radio or infrared wavelengths, have been tried with Hubble in visible light and not seen. You'd only be frustrated with those, so let me steer you away from them. For your equipment you need disks that are bright and large. I see that you are in New Zealand. The best targets to try for these criteria are AU Mic and HD 15115. Both are a good bit fainter than beta Pic though.
BetaPictoris is by far the brightest one. I don't think any of the other discs are going to be easy but they are certainly worth a try.
I got a significantly better image when using shorter exposures of only 7 and 4 seconds for Beta and Alpha respectively. I chose 7 seconds because with less exposure I didn't see any noticeable decrease in the area of saturation unless I went all the way down to something like 0.5s where I'd imagine the disk is definitely not visible. But that depends on the camera's well depth, it may be different for you. It might be worth trying 5 and 3 seconds when you have a more sensitive camera than mine.
Just remember it still has to fit with the factor of 0.59 for Alpha, so depending on how fine you can adjust the exposure times you may choose another number. I can only adjust exposure in 0.5s increments, so I had to choose something that was compatible with that
Quote:
Originally Posted by kinetic
Rolf,
I've done a few more sets on a few more nights, pretty much with no change in my imaging train.
Seeing has been better on two occasions to my initial sets. Still no result for me.
The moon is in the way now for a few weeks so I hope to use the down time to try a few things with the secondary/ spider.
You may be right about the spider vanes.
It's certainly not the primary obscured by clips, and I feel it's also not due to collimation, or collimation
change rather, due to OTA flex when slewing from Alpha to Beta.
See the attached Airy disks of out of focus stars near Sirius (low declination) and stars near the SCP (high
declination).
Almost no change.
I really want to get to the bottom of this and hopefully get a positive result from my rig.
I'm reluctant to turn the entire OTA 10 or so degrees to take the diffraction spikes further away from the PP Disk PA.
And I probably won't if the tests with an alternative spider show no
improvement.
I might even give one of the SC1.5 webcams a go too!
Also, it has been remiss of me not to apologise.
I did raise doubts, I followed them up with my own observations, and
until I get to the bottom of it, I really need to apologise for firstly
hijacking the thread, and secondly doubting your results.
I never doubted your abilities BTW Rolf, only your results.
It's also interesting the list Dennis has attached to your better result thread.
I noted the PA, inclination and angular size of a few other PP disks.
It looks like we could have some far easier targets for amateurs, based
on your result with beta Pict.
regards,
Steve
Quote:
Originally Posted by Dennis
Hi Steve
I might select one from that list and give it a go with the Mewlon 180 if the weather ever clears in Brisbane. I haven’t had much success with Miranda, one of the fainter moons of Uranus, because of the distinct diffraction spikes from my 3-vane spider (6 spikes) and the plane of Miranda’s orbit always seems to place it on one of the vanes!
Here is comparison of BetaPictoris and AU Microscopii: http://astro.berkeley.edu/~kalas/dis.../aumicvis.html
The image has the same scale for both stars, so it seems AU Mic might be quite a lot more difficult than BetaPictoris. It will most likely require a lot of experimentation to catch it since the sweet spot between oversaturating the star and picking out the disc will likely be different again in this case.
here is a report on yet another attempt to get this very difficult object,
if only to show what still has not worked for me and maybe help others
with their attempts..
A bit of background info first:
I have changed out the spider vanes to thinner vanes yet again.
I have rotated the OTA so that the diffraction spikes are now East/West.
I have captured many sets over four good nights now, last night being
the best recently for seeing. FWHM was reading 1.76 at one point on the
live readout!
Collimation is as good as I can get it.
Now, I have also refined a way to exactly match my darks to lights
as far as CCD temp goes with a better temperature control than I
previously had.
I can keep the CCD at an accuracy of 0.1C whereas previously it was
swinging around a delta of 2C. This much change shows up in results.
Results are below,
I don't think I can say I got anything.
Note the similar diffration patterns in the pre-layered Alpha and Beta raws. This I think, shows that
swinging from Beta over to Alpha has not significantly changed any collimation.
Layer alignment was sub pixel, I was working with the actual 4x resized raws, not these 1x jpgs.
The next thing I'm going to try Rolf is an obcuring disk on the CCD itself
and I will let you know the results. A difficult bugger to be sure!
Yeah it doesn't look like you got it there.
My thoughts are as follows: I can see you used ~70 frames, I think that's too few. I tried a couple of weeks ago using a NIR filter but like you I came up empty handed (more on that below). I did use 500 frames on each though and it really makes a difference for the background noise. My conclusion so far is that the disc is not very bright in IR at all, contrary to what I've always thought. At least it's not bright in the NIR range, but maybe at longer wavelengths.
I did some further reading on it and as far as I can gather the disc is optically neutral, meaning that it simply reflects the same colour of light that hits it. So there wouldn't be any wavelength that offers a particular advantage in contrast between the star and the disc.
I think the best option is to just use no filter at all, so for me I'll need to have another go witout the NIR filter.
Apart from number of subframes you do seem to have some star trailing (the bright double star in the 2 o'clock position looks close to an 8), but I'm not sure if or how that amount of blurring will impact the detectability of the disk.
Also, the disk gets exponentially brighter closer to the star, and the saturated area is quire large. Maybe try shorter subframes and many more of them.
The diffraction pattern is also visibly different between the two stars, though only slightly. But maybe still enough to overpower the faint signal from the disk in the difference image? I think more subframes will help here too. I don't think it's due to collimation changes but it's my experience that all the small diffraction spikes slightly change shape and size under atmospheric influence so it does need to be averaged out over a large number of frames to get a true image of the PSF of your system.
That's what I can think of at this stage. I'll probably try myself again once I get a break in the clouds that coincides with a new Moon.
My thoughts are as follows: I can see you used ~70 frames, I think that's too few. I tried a couple of weeks ago using a NIR filter but like you I came up empty handed (more on that below). I did use 500 frames on each though and it really makes a difference for the background noise. My conclusion so far is that the disc is not very bright in IR at all, contrary to what I've always thought. At least it's not bright in the NIR range, but maybe at longer wavelengths.
I did some further reading on it and as far as I can gather the disc is optically neutral, meaning that it simply reflects the same colour of light that hits it. So there wouldn't be any wavelength that offers a particular advantage in contrast between the star and the disc.
I think the best option is to just use no filter at all, so for me I'll need to have another go witout the NIR filter.
Apart from number of subframes you do seem to have some star trailing (the bright double star in the 2 o'clock position looks close to an 8), but I'm not sure if or how that amount of blurring will impact the detectability of the disk.
Also, the disk gets exponentially brighter closer to the star, and the saturated area is quire large. Maybe try shorter subframes and many more of them.
The diffraction pattern is also visibly different between the two stars, though only slightly. But maybe still enough to overpower the faint signal from the disk in the difference image? I think more subframes will help here too. I don't think it's due to collimation changes but it's my experience that all the small diffraction spikes slightly change shape and size under atmospheric influence so it does need to be averaged out over a large number of frames to get a true image of the PSF of your system.
That's what I can think of at this stage. I'll probably try myself again once I get a break in the clouds that coincides with a new Moon.
Thanks for the comments Rolf,
Here is a result from 27 Nov.
details: exposures for Beta: 5 sec, exposures for Alpha 3 sec.
(that is still a a ratio of 0.6, same as 10sec vs 6 sec for previous attempts)
You state that 0.591 is ideal and your successful attempt was 0.571 with
the critical Beta exposures at 7 secs.
So I'm happy that I'm close to the ideal. Maybe closer to the ideal than you were
Difference align was done with 4x resize raws
I've also shown an inset with a deliberate misalign using the same raws.
You can see the same resultant flare pattern after a curve manipulation shows up with both alignments.
Conclusions:
I might have something resembling the PP disk in the correct P.A of the disk.
However, curve adjustment of the difference layer should show anything
'different' in a 'brighter' layer as a brighter feature.
My artifact is darker.
Shorter exposures than 10s/6s and other ratios have shown
a different flare pattern.
Now what has also changed?
I remade the spider with 0.25mm vanes, and made a secondary vane holder and screw adjustment assembly on the lathe.
Laser collimation was spot on. Collimator was also collimated in a vee block as previous.
Also, 27 Nov was a brilliant night, way superior to many recent nights.
You can see how good it was by the 'star' curve I have applied to
the close double star, inset.
Another plus was that all subs were obtained with the CCD sitting at 10.3C for over two hours, both lights and darks.
That is very stable imaging.
I'm confident that this is a fairly tight set of data to use as a benchmark
for future nights.
a bit of an oops moment Rolf !
I realised I had been stacking crops!! of the main result raws instead of full frames for the latest attachment (27 Nov)
This only clicked when I puzzled over why the 'eclipse' Alpha/Beta dot was so stark and different in size in this set!
Attached is a layer difference with usual curve with the intended 1:1 ratio
of beta Pict raw and alpha Pict raw.
The dark stars are from the Beta frame, white stars from the Alpha frame.
So, by definition, I'm guessing the bright(dark) artifact in the almost correct P.A.
of the P.P. disk is a bright feature (once curves made it stand out) on
the Beta raw.
I post the following further results in his thread with kind permission from Rolf,
27 Nov results were redone right from scratch.
This is a very good data set. I had let a bad sub slip through
which shows as an elongated blob almost N-S on the central star(s).
I resized all dark calibrated/ BPM subs to 6x and restacked.
Difference layer was performed on the full dynamic range FITs of
Alpha and Beta this time, not a TIF or PNG crop.
Curves were performed on a TIF result.
A second 12th December set was also done.
This was , again , an exceptional night of data.
Same processing sequence was used.
Conclusions:
There again seems to be a promising bright artifact at the approximate P.A. of the protoplanetary disk.
While I am not even close to assuming I have nailed it, it really does
look promising.
I think a lot of follow up imaging needs to be done.
Rolf agrees that it would be great if others could try this technique.
It really is not such an onerous task.
1500mm F.L. imaging with 5-10 sec subs is not that hard.
Cooling a CCD doesnt seem to be a big deal either (the 10C result shows
the same artifact as 19C results).
What does matter is what Rolf has been saying all along:
The spider vanes do need to be thin and well aligned.
The night has to be exceptional seeing, otherwise the PP disk will smear
over more pixels.
IR filters don't seem to matter.
The use of a Baader IR filter has still picked up this artifact.
The timing of Alpha to Beta subs is critical, also as Rolf has been saying.
You can see by brutal manipulation of the difference layer that stars
have rings around them, most notably the central dot of the main stars.
If the timing had been exact, the central star would be black and not reappear after a severe stretch.
That means resolving the PP disk could get better if timing was tweaked
more.
On a personal note, nothing would give me more pleasure after many decades of imaging and observing under the stars, than knowing I
might have captured a protoplanetary disk around a star from a backyard
scope setup. For that possibility I have Rolf to thank for the inspiration.
Congratulations to you and thanks for your kind words. I have no doubt that you've captured the disk this time. Particularly the 5th image is very convincing - is it a 3D plot? The contours of the anomaly in your image has a strong resemblance to the 1984 discovery image (image is upside down): http://astro.berkeley.edu/~kalas/dis...bpic84full.jpg
So, now that imaging the circumstellar disk around BetaPictoris is apparently common routine among (two) amateurs, I wonder what's next...
Actually I strongly suggest you create a new thread with your latest post because you've got a significant result, and it's also a scientific confirmation that this can indeed be done routinely by amateurs.
You'd probably also get more response from other forumites than when it's buried in here in the old thread
A result from very , very steady skies last night attached.
The plan was to place two finepoint texta dots on the CCD and
try to obscure the central part of Beta.
This was always going to be a bit hit and miss because the dots
weren't exactly opaque, not hard edged, and the optimum place for any obscuring
plate or dot is on the CCD wafer itself, not on the cover glass.
Because of this they were just fuzzy semi opaque dust bunnies.
I did a quick preview on Beta, saw that it was hopeless and because
I didn't want to strip down the camera and clean them off, I proceeded
to just do the normal subtract/ difference method with the dust bunnies
off to the side, out of harms way.
So these full frame results show the smear of the two obscuring texta dots as tracking PE made them oval.
Conditions were brilliant, and I did a few other objects the same night.
As far as the anomaly , I think the diffraction type artifact spiking off at about 4:30 on the clockface is pure diffraction but the two lobes arrowed
are definitely part of the consistent anomaly being picked up, matching the 1984 discovery image you referred to.
.
