Dig EOS 300D - Astrophotography

[vwestervelt]

Ok, already got some input asking "what" digital EOS I am using. Sorry for the omission

Here's a repost. It's the new 300D.

I've had several requests to post some of the images I'm getting from my new Digital EOS.

Here is a link to an image I've taken of NGC498 - Triangulum Galaxy. (more to come when the weather improves)

The image is not retouched, but is color level adjusted. The CCD array used in this camera is more sensitive in red/green than in blue. With extremely low light photo's such as this, it's very evident in the raw frame.

The image is 50% of actual size.

30 second exposure @ ISO 1600

The lens assembly I'm using on my telescope is designed for a flat field but does have spherical aberration, which is evident. The trade off is a nice flat field imaging area at F2. Telescope w/out this lens assy is an F10- 2032mm. This a a very faint object in the night sky, the F2 lens really helps as you can do shorter exposures and limit telescope tracking errors (Rotation of the earth would not allow this image w/out motorized tracking!)

I'm still looking for a way to get a dark frame with this camera to subtract noise if anyone has any ideas.

Link
http://www.vernwestervelt.com/Galaxy/Triangulum.htm

Please let me know what you think. I'm very happy with this camera so far.

Vern Westervelt

[vwestervelt]

I should note the full size images show a moderate amount of red/green noise caused from target "glow". The camera does an excellent job of processing most noise, but with extreme low-light conditions, the glow from the target is very evident as noise surrounding the target....

I sincerely wish this camera had the ability to turn the processing off to allow the capture of a dark frame to "noise-subtract" from the main image.

Still, for single shot color CCD astrophotography, these are the lowest noise images I've seen. Next clear night I intend to do long exposure with this camera.

I'll post the results.

[RichardtheSane]

If you shoot RAW then the camera should do no processing at all, and you should end up with exactly what the sensor saw.
I may be wrong though :)

[vwestervelt]

Tried that. Unfortunately, the camera sill processes out noise. In most cases I'd say that was good, since you're not adding artifacts, you're subtracting a noise constant from the final image.

In the case of such low-light images, however, the noise caused from target "glow" isn't accounted for in the software. A simple darkframe, subtracted from the final image should almost totally elimiinate noise. This camera still has some of the best low-light capabilities I've seen in a off the shelf ccd array.

A 1024 x 1024 CCD imaging camera made for astrophotography runs in the 2800.00 + range. They are made for low noise, and usually peltier cooled to eliminate heat generated noise. (or some type of cooling)

For a sub 1k camera, with 6.3 mp.....this camera is going to be an excellent low cost entry into astrophotography. It'd just be a lot better if I could get a dark frame. :)

[robertwgross]

Vern, your camera still does not have a CCD array.

It has a CMOS sensor.

---Bob Gross---

[vwestervelt]

Yes, you are correct, it's CMOS.

However, minus the differences in supporting circuitry, the actual imaging elements are very similar, as are the complications of noise either solution presents.

Noise in CMOS is actually worse in raw form than CCD (mainly because the supporting cuircuitry is located so close to the actual element), but the speed enhancements from the differences in design allow cannon to do a noise subtraction "on-the-fly" for CMOS and still whop the data transfer of thier ccd counterpart.

Technically, the semiconductor manufacturing differences are "metal oxide semiconductor" for CCD's and "complementary metal oxide semiconductor" for CMOS...hence the name. The primary differences are in the circuit designs, not in the imaging element itself. The "major" difference is that the supporting circuitry is actually located alongside each light receiving element in CMOS....which, adds noise to the element itself.

I appreciate the correction of terminology.

Still need a dark frame though :)

[jimsloy]

Vern-

Just a thought here, and this may be stupidly primitive so please, shoot me down here if it makes no sense.

Why can't you just create a gient 6MP sized image in PS and fill it with BLACK as 12am, midnite? Won't that do?

[vwestervelt]

Unfortunately no Jim.

Whether your imaging element is CCD, or CMOS, the physical characteristics of the actual sensor is very similar between the two.

Basically, each pixel is a container for a charge. There are 3 primary sources that will transfer to an electron charge in each cell.

1. Thermal. Heat will transfer it's energy into electron charge in the cell. Try taking a low light photo in a very hot climate. Results will vary dramatically from a low light photo in a cold environment.

2. Light. Roughly translated, one photon of light will transfer one electron charge to the pixel element.....depending on the sensor, each has different characteristics of sensitivity. My HP 850 CCD camera is way more sensitive in blue...where the 300D is more sensitive in Red - Green.

3. Electronic Noise. Noise from the power in the curcuit will trickle charge into cell elements. It's actually much worse in the CMOS in raw form as the primary difference for CMOS versus CCD is the supporting circuitry actually being placed in the immediate vacinity of each pixel sensor cell. The speed enhancements in this arrangement allows the software to process out noise on a sampling technique which still leaves you with a faster overall response....and in the case of the 300d and others...a better image without having to do the typical "10 second image - 10 second dark frame subtraction"...they kind of do it on the fly in millisecond increments usually.

The problem is, software really doesn't account for "glow" from a target in very low light conditions. Your target always changes, and the software targets "knowns"....mainly the "thermal" and "electonic" noise. Light from "glow" trickles into neighboring cells depending on the sensitivity of the sensor.

The Dark Frame just allows you to do a low level subtraction that will significantly clean this up.....but this camera won't allow the capture of one since it filters ALL noise out with a lens cap on or telescope cover.

Anyway...probably way more information than you wanted.

Talked with Cannon...their techs say it might be a firmware update.....but they aren't sure.

Vern