Sun, 05/07/2017 - 19:25
Beginner here and interested in starting DSLR photometry but do not currently have a camera or lenses. I see a lot of recommendations for the Canon 1100D in previous years, but what about current models? Any opinions about the Canon T6 or T6i vs. the Nikon D5500? The Nikon uses Sony sensors which are said to have very low read noise, but are there features on these newer Canon/Nikon cameras that present problems for photometry? What focal lengh fixed lens is best to start with, 100mm? Could good quality older manual lenses be used (with mount adapters if required)? I will be using on a tripod to start and then later piggyback a C8 on a goto GEM mount. Recommendations and advice sincerely appreciated.
Hi,
I would not promote any particular brand but just discuss some points that seem important to me:
1) read noise: not important for photometry, the present CMOS level is just a couple of electrons, in photometry the shot noise is inevitably much higher and far dominant, even with a low signal of 1000 e- the shot is 33 e- (sqrt), then the read noise has very little impact on the SNR.
2) the ADC offset: it is used to leave some coding space for the negative part of the noise signal in images that are in any case coded as unsigned integer. Without it there are cases it's difficult / uncertain to calculate the background level. Not all brands use it.
3) software: some brand have a full set of PC software delivered (free) with the camera. some others you should buy it in addition, sometime very expensive. Some do not have any... Not all softwares have the possibility to make series of long exposures, or even any series at all, depending brands and now, camera model in a given brand. With the wifi fashion there are now some camera that doesn't even have PC connection. In general the wifi phone/tablet apps are not ok for us, at time being, could evolve...
4) special pixels: present sensors include a set of "phase detector" pixels for autofocus. Those pixels have not exactly the same optical behavior than others and this is not properly corrected by the flat process. In cameras I use it's not too critical but it's better to avoid them. One new solution are the sensors with "dual-pixels": all the pixels are phase detectors, that should solve the pproblem ! At time being such cameras are about $800, not the entry level...
5) manual settings: it's clear all shall be accessible, including focus, elimination of noise reduction: both long exposure and high iso (don't use too high iso, any... ). Some cameras impose the taking of dark at each long exposure image: improper to us... It's usefull to have the possibility to set short series (usually 10) by the retarder, low light level/red of the live view, 10x live-view zoom (focus)... There are also live-views that shows the stars much better than others.
6) I don't use any more DSLR but the said "hybrids" instead, the reflex chamber, the back of the mirror, have some impact on the accuracy of the flat (light scatter). The hybrids doesn't have such problem and the short distance between the lens flange and the sensor can help a lot. Those cameras have the same electronics than the DSLR, the same performance level. They are small and low weight, short throw to the flange, all that helps the attachement, balance, to telescopes/mounts.
7) Lens: it's very depending what you are interested in. It's possible to work on mag 3 with a 100 mm lens f/4 and a large defocus, mag 6~8 at f/2.8 and light defocus, better at mag 8 with a 200 mm F/2.8, then you get an high SNR at mag 11 with a 200/800 newton... Some make massive wide field photometry with a 50 mm lens of large aperture.
There are two key aspects: the size (F/N mm) of the aperture that shall provide enough photons for a given magnitude and the field of view that enables to access the proper comparison stars. More luminous are the stars farest are the comparisons. There are tables of exemples in the DSLR manual. https://www.aavso.org/dslr-observing-manual
The lens shall also provides an accurate manual focus control. Some SLR old lenses sold for a few are very good for us, there are often adaptable to present DSLR with the proper adaptaion rings.
Welcome to photometry, it's fun science !
Clear Skies !
Roger
Thanks for the info. After some consideration I'm getting a Nikon D5500. I can always use it for an everyday camera if it doesn't work out.
Hi!
> After some consideration I'm getting a Nikon D5500
Sorry for being late to the discussion, but I would like to throw in one advantage of the Canons vs the Nikons that I find important:
As Roger had mentioned:
> Some SLR old lenses sold for a few are very good for us, there are often adaptable to present DSLR with the proper adaptation rings.
However, there can only be an adaptor ring available if the distance of the sensor to the mounting plane is not shorter for the old lens' system (flange focal distance) compared to that for you new DSLR! The Nikon-F DSLR mount has a longer flange focal distance than many (most?) camera models of past and present, so there are a lot of legacy lenses that you just cannot use with a Nikon.
To give you an example, I got several very fast and very good legacy Pentax-K lenses from eBay for next to no money compared to modern lenses with the same aperture and focal length. It can save you hundreds of $ ! But the Pentax-K flange focal distance is shorter than the Nikon-F's. It's actually only a mm or so, but it's enough to make a simple adaptor ring adaptation impossible.
CS
HBE
Sorry to be joining the thread so late, but I'd like to weigh in for the benefit of future queries as well.
When speaking of software, one really needs to think in terms both of processing- and control software. Control software is what you use to take the exposures, and within that realm there is nothing that I know of that beats Back Yard EOS, a version of Canon's EOS Utility program that is specifically designed for astronomical imaging. There is a free version, I think, but I have the pro version which cost all of $50. I cannot imagine doing camera-based photometry without it.
Also, to attest to Roger's advice, I would recommend that anyone getting started consider alternatives to the DSLR, which can offer equivalent performance at much reduced weight as well as lower price. Make sure you can turn the display off; otherwise the requirements are the same as for a DSLR, as documented in the AAVSO DSLR photometry manual.
Control software is an excellent point!
Astro Photography Tool (APT) is another good choice, it supports many Canon EOS models as well as ASCOM "compliant" CCD/CMOS cameras.
http://www.ideiki.com/astro/
I'm not implying that there isn't anything comparable for Nikon cameras, tho. I just don't know.
CS
HB
Thanks for all the additonal input. I am now looking for a gently used Canon T3i/600D body. I can take what I save by not getting the Nikon D5500 and put that towards a 200mm f2.8L lens.
While I am not advocating one brand over another (until they pay me to do so ;-} ) I use a T3i, which is supported by BYE. I have nothing bad to say about it.
Hi!
Good choice IMHO.
> put that towards a 200mm f2.8L lens.
If you can do without auto-focus, image stabilization, perfect color correction and other modern goodies (and for photometry you can do without all that), make sure to keep an eye on the legacy lens market.
For example, look at this one:
https://www.pentaxforums.com/lensreviews/SMC-Pentax-K-300mm-F4-Lens.html
whichcurrently sells, used, for around 200 -250 US$ (see the price tracker on the bottom of that article). Sensors have improved a lot in recent years but lenses as such were hard-to-improve already many decades ago. And they age quite well, you just need to make sure no fungus has made it into the lens assembly.
CS HBE
Comment: In regards to SFS's post, there is also a BackyardNikon program that is the equivalent of BackyardEOS.
My question is related to Roger's post near the top of this thread: In your point #1, you mention that the read noise isn't a big problem in photometry. In reading the DSLR photometry manual, it recommends using a low ISO to retain as much dynamic range as possible. But in my astrophotography, my Canon 1100D usually needs to be using an ISO of 800 or 1600 due to the lower read noise at those levels for that specific camera (of course I am chasing very dim nebula, etc. in that endeavor). So is it correct to assume that I don't need those high ISO's for photometry?
Thanks!
Hi Scott, in general the astrams use much too high iso compare to the present DSLR technology status. Reasoning should always be done in electron count, not ADUs. Using high iso means much less dynamic, saturation at lower electron count. The shot noise is sqrt(flux e-) that means the "potential" SNR is higher at low iso. We are in "shot noise regime".
But ok it's the limit of the potential SNR, if your signal is anyhow very very low the Nyquist-Johnson noise of the sensor could have an impact (on long exposure at high temperature the shot noise of the mean dark currents is also involved, but it's not iso depending). Today the NJ noise of the CMOS sensors is very low 1~2 e- is common, then if you make photometry of faint star providing only 1000 e- you have 32 e- of shot noise, the NJ noise of your DSLR has no impact if your measuring aperture is about 30 pixels or less (combinaison is quadratic ).
ISO 250 to 400 is the good range for photometry ( NJ about 2~4 e-) with recent sensors. At 250 we have a calibration of 1 e- per ADU that's enough. In case of bright star 100 ISO is right even if we have more than one e- per ADU, generally using large defocus, then SNR of more than 1000 is possible. For medium / faint stars a moderate defocus is also recommended depending the optic being used and compromise against SNR.
But don't focus too much on such theoretical SNR, for us it's not usually the worst contributor, the sky is the problem ! We are not all observing at Paranal...
Clear Skies !
Roger
With your Canon 1100D, I doubt that you will be using anything higher than ISO 400 for photometry. The ISO is equivalent to the gain. Ideally, you are looking for an ISO setting that will give you a gain of close to 1 electron/ADU. This is part of the camera testing that you should do when starting DSLR photometry, determine the gain for each ISO setting. In Appendix A in the DSLR manual, the example uses a Canon 1100D. It's ideal ISO setting would be ISO 400 which gives a gain of 0.83 e/ADU. I operate my Canon 70D at ISO 200 because at ISO 200 it gives me a gain of 1.07. I will operate at ISO 100 for very bright stars because ISO 100 gives me a gain of 2.24 e/ADU. So I decrease the risk of saturation at this lower ISO.
Barbara