Sbig St4000xcmi

[jase]

Perhaps its time for some of you to sell the DSLR and get serious...;)

http://www.sbig.com/sbwhtmls/special_production_st4000xcm.htm

[netwolf]

Definetly one of the best things to come out this year. Sbig are really coming to the party and reducing the cost of ownership. Still well outside my budget, but maybe down the road if i win the lotto.

I am also awaiting the launch of there new autoguider, lot of hype around the launch of it.

Not to mention there drop in price on the adaptive optics unit.

Regards
Fahim

[EzyStyles]

ive heard they are limited editions...

[jase]

With summer approaching, the struggle to reduce thermal noise becomes a challenge for the regular DSLR. If cooling mods were cheap and easy, then I think this would be a non event.

http://www.atscope.com.au/sbig/st4000xcmi.html - Same information, but AU pricing - A$4800. Probably comparable to a Ha modded, after marked cooled DSLR.

Possibly a limited edition, but would depend on demand and more importantly CCD chip availability. SBIG, like other high-end camera manufactures (FLI, Apogee, etc) secure a limited number of CCD's from manufactures such as Kodak, Marconi, SiTE, Fairchildimaging, etc. Despite what you may think, these guys don't have cameras sitting on their shelves already assembled waiting to be shipped to a customer. Assembly typically commences when you place an order, hence the sometimes longer then expected lead times. I waited 5 months for the STL11k (and that was quick compared to what I've heard from others). Of course, its worth the wait.

[h0ughy]

well still think my cooled 350 D is good value - 8meg and I can get it to 18 below ambient temp. however, coupled with the adaptive optics, this is an awesome weapon

[Alchemy]

had a look and though i would dearly like a real astroimaging setup (dear unfortunately is the case) its out of my budget too.

i notice this is a one shot colour job, as much of a pain filters are id go for the mono one.

ive got a modded 840k and the dmk 2, both use the same chip, one is mono the other has a bayer, the electrics are probably better in the dmk but theresolution with the mono is soooo much better.

i could probably settle for something in the 2 megapixel range as it would cover most objects such as galaxys, planetarys, not m42 or some of those hugely detailed nebula shots we see all to rarely ( cmon get that 11k shooting)

cheers... a low budget AL

[g__day]

I wonder why its only a USB 1 interface?

[jase]

Indeed, this is a one shot colour CCD thus making it a logical step for DSLR users. Not everyone can be bothered with using individual filters. It certainly increases the work load. I look at the high quantity output of others in the DS forum, such high volumes take a lot of work with a mono camera and filters. Where you might shoot two or three objects in one night with a one shot colour camera, you are likely to image one object over a course of a few nights with a mono camera as you take exposures through each individual filter to get the best S/N ratio possible. It is no exaggeration that the work load with a mono camera is three fold considering you also need to take flat fields through each individual filter to remove dust motes. There is no argument when it comes to quality - mono CCD chips all the way.

Filters do however offer incredible flexibility. Narrowband imaging from the light polluted suburbs or on full moon becomes reality. You can optionally take narrowband imaging up a notch by using dedicated continuum filters (aka companion filters) at each narrowband wavelength (Ha,SII,OIII) and subtract the two. The subtracted image is void of all stars and other features leaving only the gaseous densities of each wavelength. This is what I'd consider the purist form of narrowband imaging. An example of this type of work can be seen here - http://www.rc-astro.com/photo/id1170.html
You also have the flexibility of blending conventional RGB data with narrowband data to give richer colour. Another benefit of individual filters maximising object location based on filter. For example if the object is low, I typically image with the red filter, then as the object gets higher in the sky I change to blue and green. This minimises the effect of atmospheric dispersion or scattering of blue light. Possibilities are endless. Buying a CCD camera today, I'd still head down the mono path despite the additional work loads.

Quoted from SBIG regarding USB1.1 vs. 2.0;
"Many people have asked this question and while we don't have a specific answer we can give you some information. A lot of people ask this because they think that USB 2 cameras must be faster than USB 1.1 cameras. In theory that could be true but often times in practice you'll find out it's not the case. Our USB 1.1 based cameras download pixels at approximately 400,000 pixels per second. If you ask the other manufacturers how fast their cameras are you'll find many USB 2 based cameras that are slower than that. So it's not just USB 1.1 vs USB 2 that's important but the actual download rate. Also, remember that in Astronomy you tend to take long exposures, typically 20 minutes to an hour so what do you really care if after an hour you image downloads in 20 seconds? Where you like fast download rates is in focus mode where our software use binning or partial frames to speed up the image throughput.
Finally, we have implemented USB 2 on our ST-402ME camera and with our design we achieved download rates of 1,100,000 pixels per second. While we plan on upgrading our other cameras to USB 2 at some point in the future no specific time table has been set."

[rogerg]

Hmm, nice advantage over a DSLR having the built in guide chip, and binning, and I could handle colour vs mono with filter wheel... but I really couldn't handle that QE. It's not for me.

[jase]

I should point out that the Kodak KAI-4020CM chip installed the ST4000XCMI isn't something new. SBIG have had the mono and colour chips available in their STL professional series cameras for sometime. What makes this deal special is the ST body is much cheaper, but still feature rich. The STL body has a built in filter wheel which is obviously not a requirement for a one shot colour.

I fully concur Roger. The QE of this camera is low. But after all, if your goal is to take "pretty pictures" its absolutely perfect for the job. I don't know anyone that would do research such a photometry with a one shot colour camera or for that matter one which is ABG (bleeding can alter absolute star brightness measurements). With a built in guider along the same optical axis as the main imaging chip, you can guide at very long focal lengths to hunt down those distant galaxies. Saves having to worry about an off axis guider or heaven forbid a separate guidescope that has the potential to not detect guiding errors if imaging at such long focal lengths.

Actually, the QE of the colour one shot is identical to that of the STL11k.
KAI-4020CM (http://www.sbig.com/large_format/qe_4020cm.jpg)
KAI-11000CM (http://www.sbig.com/large_format/qe_11000cm.jpg)

The mono version of the KAI-4020 has a slightly better QE than the KAI-11000M!!!
KAI-4020M (http://www.sbig.com/large_format/qe_4040m.jpg)
KAI-11000M (http://www.sbig.com/large_format/qe_11000m.jpg)

Unfortunately, the ST4000 is likely to only come in colour. If you want the mono chip, you need to go with the STL body - at a considerably higher cost. There is always a trade off between chip size, QE and cost.

It will be interesting to see how the market accepts such products. The low/mid-end astronomical DSLR audience is quite large. It would have to be for someone like Canon to manufacture a DSLR specific for astronomical use. They would have done a careful risk assessment before making such an investment. Obviously high-end manufacturers want a piece of this market. Interesting times ahead. I wait with abated breath to see some first light images from users with this camera.

[JohnG]

I am actually seriously looking at this camera as the natural progression into CCD imaging, personally I am not into the scientific side of things and I can see this camera as a reasonable alternative to the high end CCD's which a lot of us just cannot afford or justify.

I just enjoy taking the pictures and this alternative seems like a reasonable way to do it. The arcsec/pixel appears to be quite good for use with my Tak FS-102 and the elimination of guidescopes, separate autoguiders and all the associated cables with a DSLR must be a plus.

Cheers

[Bassnut]

Gee Jase, Im over 2 minds on this new model. 1st impression was, cooled SBIG megapixels at a nice price and internal guide bypassing filters. Now thats a big bonus, guiding without filters, makes it a given, unlike the icky problem laden choice with mono SBIGs. But given Houghies cooled 350D (or the new 40D) at a quarter of the price, and that youd be interested in this SBIG cause of the megapixels and quick colour action, the end result margin is really closeing up even though the SBIG is a relatively good price (for an SBIG).

Unless you really dont want to get tricky, I would see any colour SBIG as a transition to getting serious, so even at this low SBIG type price, I would go a cooled or cutting edge DSLR and save the diff for an astro mono. That way, you could end up in the long term with both with less up front pain, and great pics straight up. Also of cause, by the time you went mono, new models would appear at a lower price.

[jase]

The ST4000xcmi coupled to the FS-102 would be a good combination John. 63.8 arcmin squared fov @ 1.87 arcsec pixels would frame many objects extremely well.

I don't know how well these one shot colour camera hold their value. I've seen some secondhand YankeeRobotics cameras on Astromart go extremely cheap considering they "buy new price". I guess there is a high churn for the one shot colour camera. As astro-imagers gain experience, they want to move up ("graduate") to something that offers the flexibility and performance and of a mono chip. Expected delivery of the SBIG ST4000xcmi is flagged for Oct.'07. I feel certain we'll see some seconds on the market by mid next year.

I see your point Fred. IMHO, the discussion of guiding and filters is over-rated. Its really not a big issue with adequate knowledge and acquisition planning.

The price vs. performance is always an argumentative point.
A user supplied 350D costs ~A$1,300 to get modified for cooling (based on prices obtained at http://www.centralds.net/en/price.htm).
The 350D when it originally came out to by new was ~A$1,400. You could possibly do the cooling mod yourself, but looks rather "intense" - http://www.centralds.net/en/process1.htm. (Is the cooling active and controlled i.e with a set temp point for consistency)??
Then, to make your 350D Ha sensitive by replacing the filter - it will set you back another $450 - http://khromagery.com.au/ir_conversions.html. Again you could do this yourself if keen - http://www.astrosurf.com/~buil/350d/350d.htm

So you tally ~A$3,150 for a decked out 8mp, Ha sensitive, cooled DSLR.
Regardless, the cost is certainly under the asking price for the ST4000xmci @ A$4,800. However, its the total package we need to consider. Internal guiding is a excellent advantage and a clean solution. No off axis guiders or guidescopes required. If we add guidescope+rings and dedicated guide camera to the equation, it may begin to balance $. We must also consider sensitivity/well depth, gain and read-out noise into the formula. These values are never easy to get a quantitive answer on with CMOS chips. The CCD will also provide greater dynamic range 16-bit vs. 12-bit for the DSLR. The 350D has 8mp vs 4mp for the ST4000. 350D has 6.4um pixels compared to 7.4um for the ST4000.

Tough call to make.

[Bassnut]

well, the way you put it the diff is not that great I guess, for a fully commercially kitted 350D, well researched ;-). But I understand the 40D is 14 bit (and live focus),. As I said, the margin is narrowing but DSLRs have a distance to go it seems to match the real thing.

OK perhaps the ST4000xcmi is way ahead for now, and not such a tough call. And BTW, I think internal guiding minus filters is an important feature with anything other that an ME, given my experience with a G11 and long FL ;-)..

[g__day]

I'd be very interested in know what factors would folk in the know apply about where and when to go past a stock standard DSLR - e.g. a Canon 350D or 400D?

Do you simply mod your DSLR to add Ha + peltier cooling or go direct to a dedicated CCD?
Do you value a CCD with guiding inbuilt greatly over a non guiding chip?
Is anti-blooming AGB mush have?
Is USB far preferable to parallel port connections?
Do you go colour or straight to mono with filters (what factors should most determine this choice)?

What factors should best guide your camera progression?

I'm learning shot processing now that I have pointing, tracking and focusing going reasonably well. My next purchase is likely to be a reasonable CCD - but how should I plan this move?

Obliviously I can do alot more with the software side of imaging - but just seeing Jase's shots compared to mine shows I have alot to learn. At present I say go to the Helix Nebulae - I lock on and shoot say a 1, 2, 4, 6, 8 and 10 minute series of shots on a C9.25 then start Photoshop CS2 processing (levels and curves and a bit of filtering) - real basic stuff. I'm yet to stack or really play hard with my images.

How show I plan my next steps forward? When should I buy more gear (i.e. what skills must I master before I get more high end imaging gear) and what gear should it be? All thoughts greatly appreciated!

[RB]

Hi Matthew, I'd like to see some of your images, have you posted some on here?

[g__day]

Andrew,

Please too - only posted 1 or 2. There probably should be a short tutorial on how to shrink a 10MB RAW file (28.8 MB unprocessed) into a decent 150kb jpeg file. From photoshop the size at high level of detail saves as about a 750KB - 880kb. Do I simply set the quality from High to very low on the file save as option to meet the 150kb limit.

Once I know this simple thing I'd be happy to post a dozen of my better shots!

[[1ponders]]

Matthew, try this thread "How to reduce images using Photoshop" (http://www.iceinspace.com.au/forum/showthread.php?t=7779)

[RB]

Yes have a look at the link Paul gave.
In short try this in PS:

From Raw, do any pre-adjusting and open into PS.
Adjust all your usual levels, curves etc.
If you like save it as a tiff file for safekeeping.
Then go to

Image | Image Size | then apply a reduction to 800px in either width or height (whichever is the larger) and
choose Bi-Cubic Sharper.

Then

File | Save for Web

Here you choose the size of the jpg file.
Note IIS now allows jpg files up to 200kb

Adjust the quality slider (make sure you've chosen jpg) so that the file size is around 200kb max.
Press Save and choose a location & name for the file.

:)

[jase]

Valid questions indeed. You should ask yourself what you're interested in imaging. Wide field, narrow field, conventional RGB, mapped colour (SII,Ha,OIII) narrowband etc etc. If you don't know, do some research. Find out what makes you tick. For me, I enjoy wide field vista as it puts things into perspective. But occasionally I mix it up and image something with a longer focal length. Image processing flow is different depending on what object you're imaging and what you're trying to emphasise in the image i.e spiral structure, dark nebula wisps etc.

A C9.25 is a fine scope for hunting down galaxies due to its long focal length. It has a low magnification secondary which reduces field curvature - a major benefit. A variety of cameras would suit this instrument with exception to large format chips. You can still use them but you'll have problems with vignetting and an chips size will emphasise the field curvature so you'll end up heavily cropping most images.

As I've mentioned in a previous post. Internal guiding has its advantages, but there is nothing stopping you from getting a camera with an imaging chip only. Then continue to guide with an off axis or guidescope with a web cam. Plenty of options available. You need to work out what works for you. Personally, I would recommend getting a camera that does internal guiding - it's then up to you if you want to use the guide chip or not. Of course, SBIG are the only manufacturer that provide an internal guide chip as this design is patented. So this limits your options. What is important is the imaging chip capabilities and whether it meets your imaging goals.

For someone who is just getting their feet wet, go with an ABG chip. Dealing with blooms of an NABG chip can be a pain. Sure, the great QE of an NABG chip is appealing, but if your goal is pretty pictures you can work around this by increasing the sub exposure times with an ABG chip. If you plan on imaging narrow fov's, then a NABG can work well. Stay away from them if you intended to do wide field imaging. You can guarantee that there will be a few bright stars in the FOV that will severely bloom resulting in considerable processing to recover the image. This is a real pain if they cut through nebulosity. You can get trick to dither and subtract the blooms, but when you are just starting off I don't think you'll want such a step learning curve. There is a minor advantage of using a NABG chip with narrowband filters as these filters are typically quite dark i.e. transmission ratings aren't extremely high so the sensitivity/high QE of an NABG chip can reduce your exposure time.

You'll find that most if not all CCD camera's today are USB. You may pick up a seconds parallel version on Astromart or other forums. I'd suggesting going USB. Some manufactures such as SBIG offer parallel to USB upgrades, though these upgrades are typically "return to base" so you need to ship the camera off to the US.

I've highlighted the colour vs. mono (using colour filters) in this thread. I would consider a logical progression for a seasoned DSLR would be a one shot colour CCD. Though everyone is different. Mono with filters will give the what I'd consider the most superior result, but its considerably more work.

Finally (leaving the best last) - image processing. Yes. You need to understand the concepts of image processing and there is really no right or wrong way, providing you're happy with the end result. If you can't image process, then you will not be utilising the full potential of your equipment imaging capabilities. Don't be afraid to push your data hard and experiment with different combining (medium, average, sigma reject/clip etc), stretching (levels, curves, DDP, shadow/highlights etc), layering (both single and grouped) and masks (highlights and noise control etc). There is so much to try such as Ha blending with different data channels to bring out object structure. Experiment, experiment, research and experiment more. Seriously, I spend more time in front of a computer screen processing images than I do at the telescope.

Post your recent images for comment on the DSO forum is a good way to get up to speed in processing. Plenty will offer assistance and tips. Once you're comfortable with the quality of your images, post it on the SBIG yahoo forums where the big players are. I've done it a few times and found it extremely constructive. Good advice and criticism can only bring out the best.

When it comes to image processing - we learn by doing!

[h0ughy]

Jase the filter is part of the package in the modding of the camera - see his filter specs and results on yun's website. it is very Ha sensitive!! so you dont need to get another filter job done saving you the $450 for David Burren's modification. and it is sealed and moisture proof - he has a heated element to dispell the dew formed by cooling