Sat, 11/19/2016 - 21:28
Hi all,
I am after some suggestions tips and advice on doing photometry on a bright (mag 6.7V) star using a 12" scope. I see my options are to 1) Stop down the aperture or 2) use stupidly small exposure times or 3) Defocus somewhat.
Option four is to pull apart the obersvatory to replace the 12" with a 3" for the project.
What shortfalls am I going to come up against and what would you suggest.
Kind regards
Simon Lowther
Another possibility is to use an Optec photometer (model SSP-3). Used ones are not expensive, and they can give excellent results. See the AAVSO PEP web page:
https://www.aavso.org/content/aavso-photoelectric-photometry-pep-program
Tom
Simon,
In past years, with my previous 12" scope, I used one of the available 3 hole focus masks that are sold by a number of vendors. I then covered 2 of the three holes giving me about an effective 3" aperture.
While you did not mention the specific target or your FOV, The larger problem became one of being able to find suitable comps within the resulting FOV; the downside being that often they were simply to faint to provide a reasonable SN ratio. Something you should check on before investing in a mask.
If the mask will not work then your options have narrowed to a smaller scope (check on the impact of the smaller scope with your CCD and the resulting image scale as well as the FOV-use the CCD Calculator before making an investment in a smaller scope) or going the PEP route as previously suggested.
Per Ardua Ad Astra,
Tim Crawford
Simon:
I don't think I saw any of the suggestions... about doing photometry (CCD) on a bright (6.7) star using your 12" scope but just to muddy the waters would ask why you chose to do that as opposed to studying the many fainter targets that might better match the 12" scope?
Taking apart your system to change to a 3" scope appears to be a poor choice IMHO, unless bright stars are really what you want to study? Could you mount a 3" scope on your 12" scope, i.e., is the mount big enough to handle it?
You also mentioned "stupidly" small exposures. That is certainly not a very scientific conclusion. I perceive that you have asked the right questions about the options so what do you think are the potential problems with short exposures and how might you mitigate them? One problem is "scin....". I didn't spell out the word since I think you know the word? How might you mitigate this problem after imaging?
You mentioned defocusing so you are thinking about ways to reduce potential saturation of pixels. Good!
I hope this response doesn't just annoy you but get you to think about the issue a bit more. Why not try the simple alternatives first and see what your photometry looks like?
HTH. Ken
I agree with all the previous comments. I think the ultimate problem you would have trying to use a 12" scope on a star that bright was mentioned by Tim. Your FOV will be too small to have any appropriate comparison stars.
I think your best plan would be mounting a DSLR with a telephoto lens on your 12".
... but, if you think your best shot at winning the Nobel Prize is doing high quality photometry on that star, I'd have to agree with Tom. Learn how do PEP with the 12 inch.
Phil
Hi Simon,
You've caused some interest here. Several points: I've always thought the field of view was dependent upon focal length and the size of the detector. In any case, what's wrong with sequential photometry, even if using CCDs?
Everyone needs to be flexible. I've measured Jupiter with a 50cm telescope for a particular project and achieved acceptable accuracy - about 2% in that case. Massive stars are also a coming popular field and most of them are rather bright - eta Carinae at 7500 LYs is ~4.5 or so. So tis topic is important.
A word of warning about the early SSP3 units. They came with two preset focal plane apertures -whether this was set by the detector or an actual aperture I-m not certain - but the 0.3 mm one needs extremely accurate guiding and with larger telescopes produces a rather small preset star diaphragm ~17" with a conventional 300mm f10.
The classical method was to use neutral density filters but I've always worried about some aspects of this. But has anyone other than me used two sheets of sun control film on a glass disc - similar to a solar filter but designed to pass 4% of the incoming light?
Regards, Stan
Using a PEP observing sequence with the CCD might work. You could take a number of short exposures of the comp star, then of the variable star and then the comp, and repeat this several times (at least 3). This would essentially be the observing protocol for a PEP observation, and would be reduced in a similar fashion by extracting the instrumental magnitudes from the images of the comp and variable (after the normal corrections for bias, dark and flats). You would also need to measure extinction if you want to achieve high accuracy. This would depend on the difference in air mass between your check and comp star.(You would also need to account for second order extinction if you are going to work in the B band).
I believe you can do bright star photometry with CCDs as long as your detector does not saturate. What you give up is throughput, and perhaps ease of data reduction, since the CCD reduction programs work with one field of view that includes both the comps and variable.
A comment on bright stars in general. I think this is a sweet spot in the parameter space for AAVSO, as many of the automated surveys saturate at magnitudes below 7.
A comment on PEP. Its a lot of fun, and produces highly accurate photometry of bright stars. We will be publishing a PEP handbook shortly which provides more detail on the PEP observing protocol, some of which would be applicable to bright star photometry with CCDs.
Jim
The SSP-3 was and is available in four aperture sizes: 0.5, 0.75, 1.00 and 2.00mm. The 0.3mm was for the early SSP-4 photometer. About 90% of all SSP-3 photometers were delivered with the 1.00mm aperture.
Jerry Persha
Thanks for the information, it has given me some food for thought and directed the project in the right direction.
Si
Simon,
If you want to use a smaller aperture, why not just mount the smaller OTA right on the 12" scope somehow, as at least one other person suggested ? There may be a way to achieve this without any major surgery to your current setup, and it would presumably have the advantage of stability and pointing accuracy that is typical of larger, presumably fixed, instruments.
There are some very nice camera/filter wheel combinations that would work very well, and aside from a little wire management and counter-balancing work, would leave your current equipment package intact and untouched. I'd suggest a close look at those options. These smaller scopes can also achieve better resolution for a filter wheel mounted diffraction grating or grism, so that's another option.
Good luck and clear skies,
Brad Vietje, VBPA
Newbury, VT
Hello! The two problems would be scintillation and comp stars in the field in order to perform differential photometry, I believe?
When I was stationed in Australia, I wanted to follow R CEN with my 8" LX200 and ST402. The field contained a couple of comps. To minimize the error with scintillation, I took dozens of exposures and averaged them. Since each exposure was 0.04 sec to a couple of seconds depending upon the filter, I could get as many as I wanted during the night's run in order to average out the effects of scintillation. Best regards.
Mike
Hi Mike
Is it better to average the frames before doing the photometry, or do the photometry first then average the magnitudes?
Much thanks
Simon
Depending on your workflow and software, averaging frames first and then doing the photometry might be easier.
BTW.: If I had a 12" scope and a 6.8 mag target, I might want to try low-res spectroscopy with that Star Analyser I got and never found the time and opportunity to try. I wonder if that might be an option for you as well. At the same time you could do photometry wirth a smaller 3" or so scope strapped to the big one.
HBE
Hi Simon,
A distinction can be made between scintillation - a slow excursion of the Airy disk around a more or less fixed point in the field on a time scale of from about 1-3 seconds of time - from seeing, which is a very rapid "twinkling" and fuzzing out of the image on a time scale < 1 seconds, with occassional balloning of the image. Both may be active, or one may dominate. This distinction can be seen more easily with with larger telescopes. A time-lapse video of speckle interferometry images can be, ahem, eye-opening.
Stopping the 10-inch down to 3", as Tim Crawford suggests, will reduce the effects of seeing. I have done experiments with using a CCD camera in "PEP" mode when there are no comparison stars in the FOV and have gotten excellent results, as long as nearly the same pixels are used for variable and comp stars and normal PEP precautions are taken. Changing out the CCD camera and SSP-3 frequently to do faint and bright stars is another possibility, but it means a lot of work.
Choice of instrumentation and observing methods depends on your interests and time available. It's a truism that if you're observing objects that don't grab you on some level, the temptation to stay indoors with a good book can become too tempting to resist!
Whatever you decide, please consider observing bright stars. There are some very interesting and little studied gems among the bright stars, including new discoveries that make radical changes to the prevailing understanding of those stars. And they make super spectroscopic targets as well. For example, there is a sixth magnitude star that was thought to be an Algol, but which turned out to be a multi-periodic very short-period variable that falls in the gap between the delta-Sct and SPB instability strips. This is a significant result. The observations with which that was discovered were rapid-cadence ones made on a 12-inch Bright Star Monitor telescope of the AAVSOnet - which might be another way for you to go. The Bright Star Monitor telescopes are equipped with BVRcIc filter sets, too. (You should check out the AAVSOnet web page, because my experience with it is several years out of date.)
Best of success,
Thom (GTN)
Hi Thom,
Your photo shows you're approaching my vintage.
I was interested in your comments about bright star photometry and wondered what areas you were working in Could you drop me a line at astroman@paradise.net.nz so we can follow this in a little more detail.
Regards,
Stan Walker
Hi all,
Thanks to those whom have contributed especially with there own experiences. I have a clear path in place which will be trying most of what has been suggested in some tests over the next few weeks to see what will work for me, then applying what I have learnt to a real target.
To answer a couple of questions; putting the C90 on top of the 12" unfortunatly is not an option (even though I have all the equipment) as my poor CEM60 mount is already at its limit; I do have another mount (Vixen SXD) for the C90 if needed. Also I do normally stick to more suited targets, but in this case I am going to a) win my Nobel prize and b) learn something during the process. While a PEP does have some appeal about it the reality is that I am too used to the automation of CCD (oh dear what is the world coming too.)
If anybody knows of any useful papers I would be interested to read.
Much thanks, Si.
The tracking quality of the mount for your scope with PEP is not crucial and does not need to be precision astrophotography-grade, since all it needs to do is keep the star near the middle of the reticle in an SSP3 (the diameter in arcseconds depends on focal length). So you ought to be able to mount the C90 on an older equatorial mount that tracks reasonably well, and testing will reveal if it can maintain sufficient accuracy. If you decide to set up the scope and PEP this way, then you may find it to be very portable and quick and easy to set up when you have a photometric night.
Frank
OK, but a C90 has a focal length of ca 1200 mm, right?? That is not so good for bright star photometry, as it will give you a small FOV with fewer comp and check stars to choose from, if you can find any at all. Before investing more thoughts into using the C90 I would definitely check the FOV and whether it is sufficient for your particular project.
For comparison, on the other extreme, I'm currently doing observations on WR 140, V~=6.8mag , with a f=135mm f/2.5 telephoto lens (a vintage Pentax smc lens from eBay for 150 EUR) and an APS-C sensor (DSLR actually). There the FOV is such that you don't even need a driven mount, just use a static tripod, I take batches of 25 exposures a 5 sec and then re-align manually, for the next batch.
P.S.:You are not by chance interested in joining this very same WR 140 campaign? https://www.aavso.org/aavso-alert-notice-546
CS
HBE
HBE,
I agree that a DSLR with a (not too long) telephoto lens is a good solution to the original problem. I had recommended this in my earlier post, but I suggested that it be mounted on the telescope. Although you are happy with the data you're getting with a stationary camera for your project, since a driven telescope is already available there may be an advantage to using it.
For each sub frame of a stacked image a little bit of read noise is added to the final image. Depending on the noise characteristics of the DSLR and the strength of the signal from the star, sacking 25 images could possibly add a significant amount of noise. (I'd be interested in hearing what the read noise is in the uncooled DSLR's.)
I would think that most motor driven equatorial mounts for a 12 inch telescope would be able to produce untrailed images with exposures of at least one or two minutes with a DSLR and a 135mm lens. Using a single 2 minute exposure or two 60s exposures, rather than stacking multiple 5 sec exposures (of equal total exposure time) might give a noticeable improvement in the SNR in final, stacked image.
Picking the right exporsures for the sub-frames would take a little experimenting. You would first want to know the linearity limit of the camera, then pick a sub-frame exposure that stays below that level in the target and all of the comp stars. Stacking fewer of the longer exposures should give a better SNR in the final image.
Phil
Here we are talking about bright star photometry, and this is what motivates exposure times in the few-seconds ballpark. The limiting factor is saturation, isn't it? When you have a star at ca 7 mag, exposing for a minute or even more with the kind of optics (3" or perhaps 2") we are talking about here will cause saturation unless you do very massive defocusing, which is also increasing read-out noise because you now have more pixels that are read-out. Whether you read out pixels (that are near saturation or near linearity limit) sequentially (in many short exposures) or in parallel (long exposure, but massively defocused, so more pixels per target star) should make no difference.
A quick reality check: I think I read somewhere that a star at mag 0 will give you (very roughly, as a nice rule of thumb) 1 million photons in the visual band per second per square cm aperture area. So a 7 mag star gives you ca 1500 photons per cm^2 per sec. For a 3" aperture (if my math is right), this is about 72k photons per second already! IIRC typical CCD or CMOS sensor pixels will saturate after receiving typically << 100k photons(full well capacity). Exposing 60 seconds gives ca 4.3 Mio photons. (This ignores things like quantum efficiency and loss in the optics, but you get the drift)
Read-out noise in modern DSLR sensors is quite ok, on the order of <<10 e- rms, e.g. see http://photonstophotos.net/Charts/Sensor_Characteristics.htm . For a pixel that receives a few thousand photons (for bright star photometry!), this is still small compared to the inevitable Poisson (quantum) noise of sqrt(a few thousand) and will not degrade your Signal-to-noise-ratio significantly, right?
HBE
Greetings, HBE,
I was intrigued by your comments about photons and the numbers. In reality, they are much larger than you think. At Auckland Observatory in the 1970s we used a 50cm Zeiss and 14 stage EMI pm tubes in two custom built photometers. In our collaboration with Auckland University we attracted, amongst others, a student wh was keen to use the EMI tube's photon counting capabilities. So the equipment was upgraded to do this and it was in use until 1990 when a change in policy saw this equipment destroyed in error.
The 14 stages with their internal focussing amplified the signal to a stage where photons could be counted using a standard digital frequency meter. But we got into trouble with SN1987A. The pulse duration was about 10 nanoseconds from memory (it needed high quality quick response resistors, etc,, which were all there) but we were getting too many pulses (about 30 * 10^7) and saturation ensued. Luckily an empirical correction saved the data - in itself this was complex as UBV filters give quite different strength signals). We needed to use a pre-scaler with all measures in order to reduce the numbers to a usable level, even down to about V = 12 or maybe more. Cannot remember the range, but is was large.
I don't know how the CCD chips are fabricated but one would think that they are merely using A/D converters to presentt either a current or voltage as units which then fill up the wells. The numbers seem more in keeping with this and the costs of discriminating each incoming photon would seem to be more than what it's costing to make these chips. But there may well have been fantastic breakthroughs. I don't know anything about the structure of CCDs - I just use them. Can anyone clarify this point?
Regards, Stan
> I don't know how the CCD chips are fabricated but one would think that they are merely using A/D
>converters to presentt either a current or voltage as units which then fill up the wells. [..] Can anyone
>clarify this point?
During exposure, photons kick electrons into a well of limited capacity. Then at read-out time, electrons from the well are allowed to flow back , and an ADC converts the current into a digital value, ideally linear to the number of electrons in the well (modulo noise), where the conversion from e- count to digital value is influenced by the gain (ISO setting). I think that's pretty much it. Electrons can also get kicked into the well via thermal processes ("dark current") instead of by photons, but here we are talking about very short exposure times, so no big worry here.
CS
HBE
PEP photometry can be just as automated as CCD. As I'm typing this response, my trusty 10" Meade with a SSP-5a (automated filter slider option) is taking data of four variables in three colors withou any human present.
Jerry Persha
First all digital cameras use CMOS sensors, no more CCD, but the physics is in any case the photovoltaic one, no more the photoelectric physics of the PM tubes. In visible spectrum any photon forms one electron-hole pair in the Si cristal of a photo-diode and that electron is stored in the capacitance of that inverse polarized diode. The resulting capacitor voltage change is amplified and sent to an ADC generating some ADU count. Each ADU corresponds to a given number of photo-electron. In practical way some photons and some electrons are lost resulting in a given QE, today QE of 50 to 70% are common (filters not included). In usual APS-C digital camera having CMOS sensors, the one electron for one ADU level is set about 250 ISO and the total capacity of the photodiode is about 25000~30000 e-. But that capacity is readable by the ADC only at 100 ISO. Above that ISO the "readable" number of e- is divided by the ISO/100 ratio (the ADC saturates before the photo-diode). On a brigth star at 100 ISO we can easily fill 100 pixels that's about one million electrons (given some distribution across the pixels). This is also an SNR of about 1000. This is typically where we are when observing 3~5 mag stars using a DSLR with a 200mm F/4 telelens and 5~20 seconds exposure. This is well in agreement with Heinz-Bernd estimate.
Clear Skies !
Roger (PROC)
Simon,
I migrated from PEP to CCD photometry about 5 years ago but still had a requirement to measure stars around magnitude 6.5 (M giants with well-measured parallaxes). The problems for measuring such brighter stars with a CCD camera is that you saturate for exposures longer than 1 second even with moderate apertures. I overcame this by buying a telescope that I usually warn parents not to buy their child - a Synta 70 mm f/7 - at a cost of $129! I have successfully undertaken all-sky photometry with this telescope as well as differential photometry. (N.B. Synta telescopes evidently sold under a variety of brands such as Skywatcher, Saxon, Celestron, Orion, etc.)
As broadband photometric filters have a bandwidth ~ 100 nm, the moderate chromatic aberration in such a telescope does not present a problem (I get nice round images that are limited only by the seeing). There is a slight refocus needed when you change filters but this can also occur if your filters are of slightly different thickness.
I mounted my 70 mm achromatic refractor on a standard wedge so I can simply change it over when I want to use it rather than a larger aperture telescope. I also have a ZWO 60 x 280 mm achromatic finder/guider that attaches to the finderscope shoe of a larger telescope. It is an easy matter to attach a small filter wheel (Xagyl or ZWO EFW) to this finder/guider along with a CCD camera. (So you could mount this on your 12-inch telescope.)
Attached is a photo of the set-up I used last year showing 70 mm f/7 refractor with an Orion Nautilus Filter wheel and Starshoot G3 camera attached. Also attached is a graph of 336 measurements of Cousins E-region standards, typically stars in magnitude range 6.5 to 7.5, taken over 20 nights in 2015.
I added a crayford focuser ($49) and electric motor ($90) to my 70 mm refractor this year. Attached is a photo showing this modified telescope with Xagyl FW and ZWO ASI174MM CMOS camera. This makes for a cheap, effective and compact photometric 'collector and detector system' for photometry of brighter stars.
Can provide further information if required.
Tex
Interesting setup. If my math is right, the FOV should be 1.21° x 0.76° , which can still be quite narrow for bright stars with few close-by comp and check stars, tho.
BTW, for estimating photon counts for a given exposure and equipment, I found this nice page from Cornell University, http://topics.sirtf.com/Astro4410/EstimatingPhotons which is much more detailed and precise than the rather crude estimate I mentioned earlier.
Cheers
HB
EDIT: repaired link