Like the thread name says will someone please explain what happened.
i have included previous images shot with the 300d cannon camera i used to use , note these have NOT been edited other than to convert to tif and then resized as jpegs so to fit here. they are as follows
NGC 2467 a 371 sec exposure at ISO 800
NGC 3324 a 386 sec exposure at ISO 800
RCW 57 a 370 sec exposure at ISO 800
Run/chook 361 sec exposure at ISO 800
all showing detail colour and promise thats what im used to , so i try out my new you beaut cooled CCD camera and expose for 600 sec, should be better yeah?
Well heres the results ,shot using supplied recomended program ignore poor tracking as thats not my concern, again these are all UNEDITED other than conversion to another format. This is the cats paw in Scorpius a VERY BRIGHT H alpha target.
An export out of WIN4AIP as a tif... note i couldnt find a debayer function so its in greyscale.
An export out of Images Plus note had to convert as hxsv9 to debayer or something like that as this was only one that looked half right for colour, the others were horrid.(the 25 version which is the same chip supposedly was awful)
Photoshop using NASA fits Liberator Plug in, again color doesnt work so its a greyscale.
All then converted to JPG in photoshop, all show the same "light"or strength of signal if you will so to speak, i have included the FITS header.... note the date is wrong maybee its my computer idont know but shot between 12.20am and 1.20 am Sunday 4th May 2008 the rest is right,
Headers for HDU 1
SIMPLE = T
BITPIX = 16
NAXIS = 2
NAXIS1 = 3040
NAXIS2 = 2016
BZERO = 32768
BSCALE = 1.0
CAMERA = 6.0 MPix Camera
EXPTIME = 600.0000
GAIN = 50%
OFFSET = 118
DATE = 11.01.2000
TIME = 01:32:02
PIXXSIZE= 7.80
PIXYSIZE= 7.80
TEMP = N/A
So if this is a better camera WHAT HAPPENNED WHY IS IT SO DARK FOR AN EXPOSURE NEARLY TWICE AS LONG
i just dont get it, what have i done wrong, this is not what i expected. Note the image looked on the supplied program just like the ones seen here..... heck i will supply the FITS files on a disk to someone really experienced if they can make sense of it. At this point i am assuming i have done something wrong.....what is it?
i will return sunday evening at 6 ish hopefully someone will know.
Tired and Gutted
Clive
Last edited by Alchemy; 04-05-2008 at 05:10 AM.
Reason: tell the picture took
Clive, the output looks fine to me. The data is there (not clipped), it just needs to be stretched. Can't stretch a jpg really well due to the limited bit depth/compression. Simply RGB convert your subs (unless a mono chip), register and combine them, followed by a stretch. You're now dealing with a different beast - the dedicated astro camera. Don't go by your DSLR intuition.
How can you make this comparison accurate Clive - considering the dynamic range difference - DSLR (12-bit) - 4096 intensity levels vs. QHY8 (16-bit) - 65,000 intensity levels? Perhaps if you performed a screen stretch to equalise the two, then evaluate the pixel values in each image it may give you an insight.
i hope youre right jase, i have limited experience in this part of the equation, hence the questions, maybee im comparing apples with oranges but the 40d has 14 bit and still is just as bright ...... why should another 2 be so different.
i appreciate your input and hold your abilities in high regard ..... i can send you the FITS for an evaluation im sure others will have questions. I can send them to anyone who has a good understanding such as yourself.
my only experience has been the dslr, and i have seen good images from qhy8 cameras.... i was hoping the new drivers were dodgy, or it didnt run quite right with xp, i dunno its all too much at the moment.....
im going to leave it alone for the day and wait to see what transpires.
Clive, I did a little bit of jiggery pokery on your image of the cat's paw to see what I could pull out of it. Performed a couple of curve functions on each channel and some brightness/contrast stretches. The only problems with saving it as a jpeg was it had already lost pixels so compressing it again lost some more. It looks a little "fake" because of this, plus the image has a bit too much red in it. But it's as good as I could get it....jase would do infinitely better!!!
Hi Clive,
Yep it is a step up in processing for a CCD camera you have to work for your image, but I think they are better -you have more control.
I guess its like a manual Vs automatic camera or your own dark room to a photo lab.
I agree with Jase & renormalised stretching seems to be the issue.
JPEG doesn't help but there is more data there.
As for exporting, don't know.
Keep at it and you will be happy with the results in the end.
i hope youre right jase, i have limited experience in this part of the equation, hence the questions, maybee im comparing apples with oranges but the 40d has 14 bit and still is just as bright ...... why should another 2 be so different.
It can make all the difference, Clive. That extra 2bits means there's 4 times the amount of information in that CCD shot than in the camera shot from your 40d. 4 times the dynamic pixel range and 4 times as many colours. What looks like a pale yellow in one might be bright lemon in another. Green could look black and whites go from creamy to snow white.
When i load my FITS subs into maxim for callibration they look quite bright. This is because maxim does a stretch on them. I then callibrate, do a max pixel stretch and save as tiff.
When i open them in ps2 to process the tiff all i can see is only the brightest stars, the data is still there though and is easily got back using levels and curves.
I am sure you have not purchased a dud, it's just a different camera and once you get the hang of working with it will get some good results.
When i open them in ps2 to process the tiff all i can see is only the brightest stars, the data is still there though and is easily got back using levels and curves.
Ditto. It's that 16 bit Tiff Clive. Once you get used to it you'll never go back.
i can send you the FITS for an evaluation im sure others will have questions. I can send them to anyone who has a good understanding such as yourself.
More than happy to process some of your data, but based on your previous work, I think you're more than qualified to handle the task. Look forward to seeing what you can do with a cooled CCD.
Well, looks like another horse has bolted out the starting gate without a saddle !.
Clive, they are all correct. You have not processed the picture. Every image from a good 16bit Astro CCD camera will be exactly the same.
This is why Imagers are so well respected for the amount of work they put in to their work.
Its a step up to start using processing techniques.
Doug (Hagar) had other starting steps which look to have been gained, as he is producing better results. You just need to understand CCD fundementals and techniques, which will come with experience now.
There is a link here http://www.southern-astro.com.au/php...ng-ngc2070.php which shows how you can process a FIT file, by Brad Moore. They are all in divx video format, so you can download it and see the same issue where you need to strech the image to start getting results.
If you dont stretch the image before converting it to TIFF or whatever, the image will be nearly black, or no detail will be evidant.
Just play with the images now, and see how processing works.
thanks for all your suggestions and encouragement, plus to the private messages i got
ive been pondering this all day and the answer i think is this.....
if we assume the camera is sensitive ie 60% + QE then it is capturing the photons , what is happening is the 'PixelWell if i can call it that is deep and holds far more than a dslr pixel well, say 45000 instead of 4500. this means the data can be stretched out..... if you get my drift.
So capturing 1800 photons for a given amount of time in a dslr gives you 40% full so a mid grey and about 1800/4500 x4096 (1638 levels of play)........ wheras 1800 photons in the Q8 gives a really dark, dark grey BUT (1800/45000 x 65000 gives you 2600 levels of play) does that make sense?
i will perservere for a while longer, i would like a confirmation from someone that my camera is behaving just like theirs as far as the comparison shot went, just to ease my mind a bit.
i will have a go withthe data i collected, it will be messy as the chip got quite dusty while iwas playing around with the condensation idea and i didnt check or see it until i did the daylight comparison shot so its got donuts and .... well a whole feed on it.
i would end up by saying this perhaps as criticism , but i downloaded a trial version of nebulosity ....worth about 50 dollars when you pay for it, and also got an 85 page manual with it...... i have bought a 2600 dollar camera and got zippo in the way of instruction manuals or documentation that couldnt be written on one page, now i will give Gama his due he did email me and say if i had a problem i could call him, but the documentation really needs some attention, both in how the camera works and its uses what to expect etc, most of the people who buy this are coming from dslr which it appears is very different.
if we assume the camera is sensitive ie 60% + QE then it is capturing the photons , what is happening is the 'PixelWell if i can call it that is deep and holds far more than a dslr pixel well, say 45000 instead of 4500. this means the data can be stretched out..... if you get my drift.
So capturing 1800 photons for a given amount of time in a dslr gives you 40% full so a mid grey and about 1800/4500 x4096 (1638 levels of play)........ wheras 1800 photons in the Q8 gives a really dark, dark grey BUT (1800/45000 x 65000 gives you 2600 levels of play) does that make sense?
clive.
This is exactly correct. The light buckets (pixels) are just much deeper. The quantum efficiency should be good and the cooling reduces the noise much more than in an uncooled DSLR so a much better S/N ratio.
Clive, read this link about processing http://www.lefevre.darkhorizons.org/...orialchap4.htm , at about 1/4 way down, he shows you the image of the Horsehead nebula RAW (Looks all black) and then when he stretches the image, all detail starts to come out..
Hi Clive, Don't get to frustrated yet, I'm one to speek of that. The camera is OK. I have had mine for about a month now and as the others have said, the data is there. It's just a matter of manipulating it better. I have certainly found it to be a huge step up from the DSLR. I have been on the verge of giving up on several occasions, but with a little help I think you and I might just make it and produce some reasonable images.
Take a look at this link, Theo sent it to me about processing CCD images and try to work through some of the process. Don't rush it. I am just starting to get images which are respectable, and they are in fact images that I really thought were rubish.
i would end up by saying this perhaps as criticism , but i downloaded a trial version of nebulosity ....worth about 50 dollars when you pay for it, and also got an 85 page manual with it...... i have bought a 2600 dollar camera and got zippo in the way of instruction manuals or documentation that couldnt be written on one page, now i will give Gama his due he did email me and say if i had a problem i could call him, but the documentation really needs some attention, both in how the camera works and its uses what to expect etc, most of the people who buy this are coming from dslr which it appears is very different.
i hope i kept it reasonably polite.
clive.
As i said in some other post Clive, the documentation for processing is not with the camera. Its with the Software package your using to process the images. The camera only takes the image, its just a dumb metal box that is cooled and sends the data down as a file, and the software you decide to use to process it will have the instructions on how to use the software and its function..Software packages are many, and all have their own method and style in executing and producing different filters and processes. What works for one package, will not work for another.
This is why there are drivers for AstroArt and Maxim, as do other cameras on the market. But they expect you to either have Maxim or Astro art as a minimum to start imaging. Thus Maxim or Astro art has the full help and tutorial manual in its box.
I hope you understand the differences. Because using the same logic, Nikon or Canon should have supplied the help to process the images in Iris or deepsky stacker or AstroArt etc.. But because DSLR has a zillion functions, there is a book for those functions. How many functions can a standard Astro camera do ?..
i just think there are a number of issues which are not understood with these sorts of cameras, as you say another horse has bolted.... not because we want to be mean or get ugly but simply because we dont understand the processes, i havent used sbig or other, some documentation would help even covering "look heres what you expect to get" or any of the basic questions asked recently.
i appreciate the fact you have been available on this forum to answer the questions , and yes some of them will be testy, i guess thats what forums are, and a question asked in person can be differentiated by tone or a look which is not evident via text.
its up to you whether you offer that level of support. i think once you have the information, even the issues gathered here from us ignorant types, it would be good to supply on a disk.
i have given the image a quick process , perhaps i should have done this first, but i was so suprised by the difference in what i usually get. having said that i will openly say i am impressed by the results and will post another thread in this column withthe processed image.
![[1ponders]'s Avatar](../vbiis/customavatars/avatar45_9.gif){kind=avatar}
Look forward to seeing what you can do with a cooled CCD.
.
