From TaareWiki

Site: CameraIssuesForAstrophotography

In the yahoogroup

http://groups.yahoo.com/groups/Astronomy_Activities_2009

The discussions on issues of cameras for astrophotography started with some off group exchanges between Dr. Suresh Mohan and Ajay Talwar, and went on to many useful exchanges in the group on camera issues:-

Canon 450D vs 40D

Weight: The 40D is 1.6 lb; the 450D is 1.0 lb. Before the 450D, I imaged with the Canon 10D and 300D. I found the much lighter 300D more suited to astro imaging from the standpoint of being lighter in the focuser of the scopes I use for imaging. The 450D is even smaller than the 300D and so it has that advantage in the focuser. The 450D is also smaller than the 40D.

Camera Use: The 40D has a magnesium shell as opposed to the plastic body of the 450D. When I modified my 450D, I thought the plastic body was thinner and not as solid as the plastic body of the 300D; maybe because the 300D is larger. For non-astro use, I think the more solid 40D would have an advantage of durability.

Viewfinder: The viewfinder of the 40D may be a tad brighter with a slightly higher magnification, but the live view feature is what would be used for astro imaging and both have a 3 inch diagonal display.

For daytime photography, the 40D has a higher frame rate than the 450D; 6.5 fps compared to 3.5 fps.

The 40D is presently being heavily discounted in price so the price difference would be a consideration.

Both cameras can be modified. More astro imaging and noise results with both cameras over time may point to one over the other for astro imaging. The 40D has 5.7 micron pixels and the 450D hase 5.2 micron pixels. Smaller pixels may or may not mean higher noise....only comparison testing will tell.

With a new 127mm scope (952.5 mm focal length) I am using, the 450D is providing 1.1 arcsec per pixel which is just about optimum for the seeing in my area.

From Gary Honis

Posted by Suresh Mohan May 21, 2008 5:08 am

Hi Suresh,

I was looking forward for this information., I was a little confused about canon 400D and 450D. Is Nikon model good for astrophotography ? if so which one ?

I'm planning to Canon 450D sometime this week. your suggestions is appreciated :)

~Naveen May 21, 2008 9:11 am

The only model in Nikon that tempts me is the Nikon D300 which can shoot upto 6400 ISO with low noise ! Which translates to almost a 2 minute data of ISO 1600 to a 30 sec in Nikon( this may not apply to faint nebulae as the data collected may not be enough ) however at USD 1800 is pricey. 5yrs down the line all may be possible, but If u have a good mount it wont matter, secondly the FASTAR system from starizona converts almost all f 10 scopes to f 2 ! I gues we will have many choices.

"Suresh Mohan" May 21, 2008 5:33 pm

Suresh, on this point, a question: does a very high iso rating really make a difference? isn't a RAW mode sufficient to play around with post-processing? reason i'm asking is that high isos (far as film goes, at least) contribute to detail yes, but also add lots of noise. from a digital perspective, what would this do, considering that all real data that can ever be captured will already be there?

forgive me if this question sounds a repetition, but i somehow can't seem to wrap my brain around the high iso concept...

cheers, Sridhar May 21, 2008 5:44 pm

Nikon have developed a new technology that even at high ISO s the noise is very low, your question is correct but their technology is new.U know that ISO 800 x I minute sounds equal to ISO 1600 x 30 secs but in reality it is not if u carry to extreme.For eg I shoot horsehead nebula at ISO 800 x 3minutes instead if I shoot 180 one sec shots there would be no data collected on the sensor and no horsehead nebula when u stack it. That is why I said this rule will not apply to faint nebula but when u change a f 10 scope to f 2 your light gathering power increases and u can shorten exposures.

For Ulhas - all DSLR s come with a built in IR cut filter .This is required for normal daylight photography but while shooting nebulae we require extra sensitivity to h alpha to capture the reds .There are many options,one is to remove the internal IR filter and make the camera useless for normal daylight photography as the focal point is changed and u cant focus as the light will focus behind the sensor ! next option is to replace the IR filter with clear glass having the same optical quality and focal length and same thickness but requires an external IR filter for daylight photography(as is in my camera). To get a clearer picture here is a summary

While stock DSLR's are fine for capturing galaxies, reflection nebulae and star clusters, they are woefully lacking in sensitivity to hydrogen-alpha energy. This is the light from glowing hydrogen that gives emission nebulae their characteristic red color. However, that sensitivity is already built into your camera...you only need to release it. All digital cameras have a filter in front of the imaging chip which greatly attenuates infrared energy which skews the natural visual color balance and does not focus at the same point as visible light...and DSLR's are no different. Unfortunately, the camera manufacturers didn't have astronomy in mind when they designed these filters since they greatly attenuate hydrogen-alpha light which falls in the spectrum between deep red visible light and infrared. In the Canon DSLR's, the response to hydrogen-alpha is down to about 20% of the sensitivity to other colors. This causes astronomical photographs of emission nebulae to appear washed out. As in the picture from the unmodified camera above, the object of interest may be almost completely absent! Unfortunately, it is virtually impossible to make up the difference by simply exposing longer or stacking more images since the signal-to-noise ratio at the critical hydrogen-alpha wavelength is so low.

The solution is to remove the camera's IR filter. This opens up the imaging chip's full sensitivity to hydrogen-alpha and also makes the entire spectrum almost one f-stop more sensitive. However, as in all good things, there is a price to pay. Since removing the filter removes a section of glass in the light path that is behind the lens but in front of the imaging chip, the auto-focus mechanism becomes un-calibrated as well as the optical viewfinder. The back focus also changes preventing normal camera lenses from focusing to infinity, and since the balance of colors arriving at the imaging chip has changed, the auto-color balance is no longer calibrated. However, for a camera that will be dedicated to astronomical photography through a telescope and focused by software (DSLRFocus?, ImagesPlus?, MaximDSLR?, AstroArt?, etc.), these issues are of little concern and the simple filter removal provides the most economical path to a sensitive astronomical camera.

Suresh May 21, 2008 6:17 pm

Looks like the 450D has many benefits including the main reason that it is cheaper! Now to locate the best place to get the IR filter removed. Suresh, What benefit would a H-Alpha filter have as compared to no filter?

Ajay Talwar May 23, 2008 2:05 am

Hello all, I dont think i get what was said in the prev mail . As far as i understand, the ISO for a film camera is actualy manupulated bythe grain size and hence grain is a problem for conventional film. For DSLR the ISO is changed by changing the gain of the read out amplifier. Every thing now depends critically on the read out noise of the liniear amplifier before the A/D convertor. Now technology can improve it to a limit- the operating temperature. That is why we need to cool the CCD and the analog amp. beyond that it becomes the digital world and noise is no longer a worry. Well not exactly i suppose. The the A/D convertor for most DSLRs? is > 12 bit systems and has a dynamic range of about 72 db. Now if we increase the gain we only limit the dynamic range. But then the stretched image now shows more for only our EYE , but at the loss of the dynamic range and increased nose. Increase in gain does not actually mean the sensor can now detect beyond its quantum effeciency. The gain comes into play after the detector has detected the photons so it has nothing to do with increased detection. All the gain does is report detection with an amplification. If the detection is zero no amount of amplification can give you a signal. But in the real world even if the CCD reports a zero detection the high gain wil report a signal which is basicaly the noise of the electronics. So it is safer to save the RAW image which inherently has a lower noise. Now the RAW image can be re adjusted to what the image would have looked like if you had preset it to a higher gain/ISO. This should remain true at all times and hence when you are unsure as to which ISO to use of RAW Format will help. Of course this assumes that each sensor has the minimum hs the minimum number of photos to detect defined by its quantum efficiency.

I read in the post about the comparison of a Shot taken at 1600 ISO at 30 sec and 800 ISO at 60 sec. The noise will be less in the 60 sec exposure. Not because it has lower ISO but because it has a higher integeration time. This should not be compared to a stacking 60 i sec exposure at 800 ISO. This will definetely result in a nill or a very poor image as the inherent noise in each exposure is way too high or in other words there is no signal. Stacking will improve the SNR only when the singal is above a min threshold for detection but this has got nothing to do with ISO. The noise stastics is the same as a poisson expectation of root N where N is the actual physical count. + the noise of the amplifier. If we assume no cooling (finite bracket of amp noise) then maximizing the number of counts in each ccd well will improve SNR. Increasing ISO only improves the immediate visible results but storing the RAW is any time more benifitial. So the take home summary is this. If you have the RAW format use it. If you dont setting the ISO will help.

I think Nikon has possibly some of the lowest noise amplifiers and this is possibly their market niche, but this does not mean we should shoot at higher ISO. Using a RAW format definetely has a higher likely hood of getting better images. I have not known any professional Imaging system have any ISO system if apapranetly it has any benifits in digital astrophotography. Gains are used but with an understanding to limit the exposure to realistic ranges and knowing that it wil lead to noise.

The removal of the IR filter definetely improves ensitivity and makes a yes no difference to shoot nebulaes. But it will change the white balance which again is not much of a concern if we have a RAW frame and a flat white field. There are are many sites that offer to make the mods to the camera and i thing the world wide astronomers group can answer who gives the best bang for the buck!.

I believe this post does not agree in totality with the prev post in this thread. Woudl be glad if some one could guide me to a better undstanding

Clear Skies Sanath May 23, 2008 3:45 am

PS: for digtal photography it is now the Pixel Scale that becomes the most relevant issue to deal with.

Sanath,

I will try to explain what little I know thro examples already available: for Nikon s ISO 25000 ! yes read ISO 25000 in Nikon D3 here http://www.luminous-landscape.com/essays/comparison.shtml but pls note im a canon man not Nikon due to other problems

And remember the rule the longer the chip is activated the greater the noise and in warm countries like ours what is recommended is high ISO less duration as in jerry lodriguss guide to astrophotography,he says anything above 80 F shoot at ISO 1600 and above anyway if you like the math behind SNR

http://www.starrywonders.com/snr.html this will definetley please Akarsh and the likes .Anyway what is the solution for people in the subcontinent is to shoot at higher ISO keep the chip cooler by using shorter sub exposures and stack many images as here http://www.samirkharusi.net/sub-exposures.html

also u should take a look at how DSLR s work from this site http://www.astropix.com/GADC/SAMPLE3/SAMPLE3.HTM in reality if u shoot a 1600 ISO for 3 minute u must rest the camera an equal time for the chip to reach the original temp but we don't as we are short of time for the number of nights we get to go out so higher ISO and shorter duration BUT MORE SUB EXPOSURES IS THE WAY TO GO in India. Though u could enhance the ISO to 25600 in Nikon u don't have dark skies to prevent sky fog Hope this helps

For Ajay http://www.hapg.org/camera%20mods.htm

Ive chosen option 2 for my camera so I can do pure IR photography when required , if u choose option 3 then it becomes purely an astronomical camera and normal daylight photography with custom white balance.(no external IR filter required as in mine) however u still need custom white balance using a grey card

Suresh May 23, 2008 5:54 am

Hey Suresh Thanks for that insight of operational temperature which explains a some of my misunderstandings. But i still do not agree with many stuff (in the name of science ;-). The ISO has nothing to do with the actual detection. and hence the shoot, save RAW format and then image manupulation is close to same as maunpulating ISO at shooting.

So shooting in RAW just records every thing the sensor sensed and not subjected to any further degredation/ enhancements http://www.dxo.com/intl/image_quality/benchmarks/understanding_measures gives an overview.

So in short it does not matter at what ISO you shoot as long as you save RAW you are ok. And yeah the temperature does put a limit to the useful maximum exposure but this does not mean that subexposure is the way to go. The link you provided does not tell you that lower exposures and stacking wil be better. It only says that there is a funtional limiting SNR for longer Exposures. meaning you do not gain SNR beyond a certain exposure. But this does not mean lowering exposure will increase SNR.

Intutivey i would say that the camera does a lot of calcumations and processing when reads and writes a file and hence common sense tells me to do taht the least to avoiod increasing the temperature of the sensor. These days most sensors are CMOS means they dont need to ploat a current to read and hence heat less!! So Ideally a sensor shoudl not HEAT UP during long exposures!! and on the contarary the whole system shuld heat up more when a number of shsort exposures at highe ISO are made. THis? is so very couter intutive.

Any way thanks for getting me curious into these topics. I think it is time for me to try out some experiments and satisfy directly rather than ponder my head over this. Any way the only way i can be wrong is if ISO is not controlled by changing teh gain of teh linear amp!!!. I shal check into that some time soon but am sure therre wil be some DSLR group discussing these. One thing i find is that there is a lot of amateur (NON TECH) comments on this aspects and find very few articles that has a strong science when it comes to explain sensor characteristics.

off to kite boarding now

Sanath May 23, 2008 1:31 pm

Retrieved from http://www.nehruplanetarium.org/taarewiki/pmwiki.php?n=Site.CameraIssuesForAstrophotography
Page last modified on March 16, 2010, at 06:53 AM EST