Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
I'm fairly new to all this stuff so I have a long ways to go, but I'm going to toss out a couple of images that I took last evening. I'm not happy with these. They are not sharp, the color is off, the background seems to have a red hint to it, and I think they just look very noisy.
I shot both of these at ISO 800 with a stock Canon XS at 55mm f/5.6. With the original frame I did some adjustments and then I cropped the area I wanted with an 800 by whatever to result in these images at 100%.
This M42 was 10 at 3 minutes. I only had one dark frame to use because I had left the auto rotation on.
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
This next one is M31, and it is a total of 90 minutes using the same techniques I used above. To me it just looks noisy.
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
justabob
Carpal Tunnel
Reged: 05/05/07
Posts: 1892
|
|
Focus is a bit soft. Using the live view with a wide lens can be tough unless you have very dark skies. More darks will help, I would also to iso 1600 to gather more signal.
--------------------
http://www.pbase.com/rkn/astro&page=all
Vixen Sphinx SXW
Meade sn6
Canon EF 200mm f/2.8L II USM Lens
Hutech 1000d
Self modded 350d
SBIG ST8300c
DSI PRO II
Bob
|
Eric Gage
sage
Reged: 12/13/05
Posts: 298
|
|
Although, if I understand correctly, changing the ISO merely amplifies the signal gain, and does not actually increase the total amount of signal.
--------------------
Eric
---------------------
WO 80mm 10th Anniversary
EZ Touch
PST
TV-85
Vixen Sphinx
Canon 500D (T1-i)
15" Obsession with ServoCat
|
waassaabee
Post Laureate
Reged: 11/26/07
Posts: 3055
Loc: Central California Coast
|
|
Auto rotation? I thought that was just for the camera display, nothing to do with the frame itself..?
But you definitely need darks, and flats never hurt! You may have over stretched causing the blown out core and stars, but it's all workable.
You never mentioned what mount, and @ 3 minutes I'm guessing you are guiding..?
--------------------
Gary
34N 120W
-My kingdom for blue squares!-
WO Megrez 90FD/TV 0.8x FR/FF
AT8RC
mini Borg 50/Q-Guide/PHD
CGEM & CG5ASGT
Canon 350D Hap Griffin Baader mod
My Friend Flickr
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
Yes Gary, I piggybacked the camera on my SN6 on the LXD75. Tracking, no guiding, but the tracking was fine. What is strange to me is that M42 had one 3 minute dark frame and 30 minutes integration and M31 had 10 darks at 3 mins and 90 minutes total integration, yet the M31 sky background looks horrible.
Something to keep in mind is that these are at the full scale of the frame. When you reduce the size and keep the scale the same the background contrast improves a lot as does the density of the image itself. Maybe I just expect too much.
Still, feel free to play with these images and let me see what can be done with them. I am curious.
James
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
Tonk
Post Laureate
Reged: 08/19/04
Posts: 4498
Loc: Leeds, UK, 54N
|
|
You have not said how you are calibrating and stacking the images - what software?
Are you using RAW or JPEG images from the camera? You should be shooting RAW
autorotation only effects the camera display - all your darks should be usable.
Chenging ISO does not increase/decrease the amout of photons captured. It only affects the mapping of the analogue signal range to the digital storage format range. There is an ideal ISO that best maps the data. That is model dependent so you'll have to look that up (google)
--------------------
Televue 85, GM-8/Gemini, Canon 40D (unmodded), Canon 450D (modded w/Astronomiks clip-ins - UV/IR, OWB)
Coronado SM60/BF10, Baader Herschel Wedge
Leeds Sky Clock Ripon Sky Clock
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
Quote:
You have not said how you are calibrating and stacking the images - what software?
Are you using RAW or JPEG images from the camera? You should be shooting RAW
autorotation only effects the camera display - all your darks should be usable.
Chenging ISO does not increase/decrease the amout of photons captured. It only affects the mapping of the analogue signal range to the digital storage format range. There is an ideal ISO that best maps the data. That is model dependent so you'll have to look that up (google)
I'm sorry for not mentioning this. I am using Deep Sky Stacker to stack the images and calibrate them. And I am using RAW images (actually RAW + jpeg because I like to preview them).
My process is as follows.
First I aim the scope with camera on back at Jupiter. I turn on the live view, magnify 10 times on Jupiter and get as precise a focus as I can. I turn the live view off. I don't touch this at all after this.
Next I move to my target. Then I shoot my light frames, right now doing 10 frames at 3 mins apiece, ISO 800. I actually collect about an hours worth of data but wind up discarding some because of trees, planes, etc.
Finally, at the end of the evening I cover the lens on the camera, put the scope in the parked position, and shoot my 10 darks at 3 minutes while I am tearing down my gear.
I download the images using the Canon software, and batch process the darks and lights from .CR2 into TIFF 16 bit files. Those are the files I use in DSS. After stacking, I load the resulting stack into Digital Photo Professional (comes with the camera) and adjust my curves and levels there. This is something I know nothing about, I just do it until it looks good. I don't even know what a good histogram should look like, or how my curve should sit on it. Then I convert it using that software into a full frame JPEG, load it into Jasc Paint Shop Photo Album (OK it was FREE) and do my cropping and sizing, rotation, etc.
Hope that helps.
James
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
waassaabee
Post Laureate
Reged: 11/26/07
Posts: 3055
Loc: Central California Coast
|
|
Have you tried letting DSS do the dark calibration?
--------------------
Gary
34N 120W
-My kingdom for blue squares!-
WO Megrez 90FD/TV 0.8x FR/FF
AT8RC
mini Borg 50/Q-Guide/PHD
CGEM & CG5ASGT
Canon 350D Hap Griffin Baader mod
My Friend Flickr
|
m2434
professor emeritus
Reged: 01/01/09
Posts: 592
Loc: on the road
|
|
Quote:
I turn on the live view, magnify 10 times on Jupiter and get as precise a focus as I can. I turn the live view off. I don't touch this at all after this.
I have never done this, so I can't say for sure. However, while Jupiter is very far away, I am not sure that it the resulting focus will truly be at infinity. I would try getting a Bahtinov mask and focus on a bright star.
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
I reprocessed M42 and applied some cosmetics during the stacking, and cropped a larger area before sizing it for posting here. Take a look at this one and tell me what I'm doing wrong.
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
m2434
professor emeritus
Reged: 01/01/09
Posts: 592
Loc: on the road
|
|
Color seems fine, but I see noise, gradient and focus issues.
Noise of course means you need a stronger S/N ratio. It's tough to determine the reason definitively, by looking at a processed image. I would guess, 10x3 min isn't working, in part, because of light pollution or poor seeing, but tough to say. I would start with more darks, more subs and maybe if you aren't doing so already try a CLS or IDAS filter.
The gradient could be from camera noise (i.e, need more darks), but to me, looks like flats are needed. And again possibly a light pollution filter.
The focusing, as I said, your optical system, probably is not focused at infinity with your method. Also, are you autoguiding, some of the odd shaped stars could be resulting from pec, as it's very difficult to tell from a stacked and processed image, but to me anyway, it looks more like focus.
|
m2434
professor emeritus
Reged: 01/01/09
Posts: 592
Loc: on the road
|
|
After reading through it again, I had two more thoughts. it sounds like you are not guiding. So, while tracking can be fine, there can still be cycles of error due to flaws in the gears, known as PE or periodic error. Also, as for the difference between M42 and M31 backgrounds, seeing conditions and even thin, almost unnoticeable(at night) cloud layers sometimes can add noise and do that.
|
quantumac
sage
Reged: 12/17/07
Posts: 499
|
|
Some of the noise might be bias (i.e. read-noise). The rest is thermal.
Below are the steps of a typical DSLR processing run. This discription is necessarily terse (there are books and web-sites that elaborate and expand on this process). How you achieve the steps depends upon the software you are using, but the basic idea stays the same. These steps assume you are capturing data as RAW (i.e. no debayering).
Bias Frames: set the camera to the fastest shutter speed it can take at the same ISO you will use for your light captures. Cover the lens so no light comes into the scope. Take ~20 shots and average them together. This is your master bias.
Flats: Take ~20 shots of the daylight sky at a short exposure, same ISO as the lights, shutter speed adjusted so that the image is gray (i.e. not too over-exposed or under-exposed). If you can't get short enough exposures, put a white T-shirt in front of the telescope to cut down on incoming light. Average these frames together. Subtract the master bias from this average. This is your master flat.
Darks: Take ~20 exposures, same ISO as lights, same exposure as lights, same ambient temperature, with the lens cap on telescope so no light comes in. Average these together. Subtract the master bias from this average. This is your master dark.
Preprocess: For each light exposure, subtract the master bias, subtract the master dark, divide by the master flat. These become your preprocessed lights.
Debayer: For each of your preprocessed lights, debayer the images to produce a color image. These become your color preprocessed lights.
Align and Stack: For each of your color preprocessed lights, align and stack to produce a stacked color master. Clip the stacked color master so that only the area which appears in all subs is visible. Adjust color curve histogram so that the peak of red, green and blue are in alignment.
Curves: Perform curve stretching to taste.
--------------------
Scope: Meade 10" LX200R
Guide/Planetary Camera: Imaging Source DBK41AF02.AS
DSO Camera: QHY8
Guide Hardware: Celestron OAG, Shoestring Astronomy GPUSB
Software: Mac OS X, Starry Night Pro, Nebulosity, PHD Guiding, PixInsight, Astro IIDC. No Windows anything.
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
Quote:
Some of the noise might be bias (i.e. read-noise). The rest is thermal.
Below are the steps of a typical DSLR processing run. This discription is necessarily terse (there are books and web-sites that elaborate and expand on this process). How you achieve the steps depends upon the software you are using, but the basic idea stays the same. These steps assume you are capturing data as RAW (i.e. no debayering).
Bias Frames: set the camera to the fastest shutter speed it can take at the same ISO you will use for your light captures. Cover the lens so no light comes into the scope. Take ~20 shots and average them together. This is your master bias.
Flats: Take ~20 shots of the daylight sky at a short exposure, same ISO as the lights, shutter speed adjusted so that the image is gray (i.e. not too over-exposed or under-exposed). If you can't get short enough exposures, put a white T-shirt in front of the telescope to cut down on incoming light. Average these frames together. Subtract the master bias from this average. This is your master flat.
Darks: Take ~20 exposures, same ISO as lights, same exposure as lights, same ambient temperature, with the lens cap on telescope so no light comes in. Average these together. Subtract the master bias from this average. This is your master dark.
Preprocess: For each light exposure, subtract the master bias, subtract the master dark, divide by the master flat. These become your preprocessed lights.
Debayer: For each of your preprocessed lights, debayer the images to produce a color image. These become your color preprocessed lights.
Align and Stack: For each of your color preprocessed lights, align and stack to produce a stacked color master. Clip the stacked color master so that only the area which appears in all subs is visible. Adjust color curve histogram so that the peak of red, green and blue are in alignment.
Curves: Perform curve stretching to taste.
Thanks for this information, it is helpful. I realize that there are many books out there on this, but an overview of the process is more helpful than all the technical stuff out there, especially when you don't understand the bigger picture. Right now I am just using darks, no bias and no flats.
I think some of my problems are thermal because it was quite warm when I started and it was a lot cooler when I shot my darks, and this may have contributed to the processing issues. I even noticed some "banding" in one of the processed images and that lead me think that this was less mechanical and more electrical/thermal related.
As for guiding and focusing errors, at 55mm, if the object stays in the center of the eyepiece of the main scope for the entire time then the tracking is good enough. I realize that when I step up to imaging at 750mm of focal length guiding is important and I will be prepared to use an autoguider then. And as for focusing issues, I do think Jupiter is more than enough for focusing at infinity, even the moon would suffice, but even on a pixel level the stars look decent enough, albeit a bit larger than I would like. Many here have pinpoint star images and that is due to larger aperture scopes, longer focal lengths which results in larger image scales so they have the option to reduce their images, the luxury of which I don't have.
Overall my question has less to do with mechanical issues and more to do with achieving better results through post-processing. I need to find a good tutorial on using DSS and probably get a copy of PhotoShop. I wish I could take on online class on post-processing as it would be helpful.
James
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
Nils_Lars
Post Laureate
Reged: 01/04/08
Posts: 4273
Loc: Santa Cruz Mountains , CA
|
|
I would not convert the CR2 files to Tiff , DSS will work with the CR2 files.
You can use Jupiter to auto focus on if you pull up the focus points in your viewfilder and put the center one on Jupiter , it will work on that and a few other very bright stars and Venus.
The Bahtinov mask is always a good idea , there are some good tutorials on post processing but I think most are Photoshop based.
--------------------
Erik
Orion Atlas Self Hypertuned (EQMOD)
Orion ED 80
Williams Optics VII reducer
Celestron 8" SCT
Orion Starshoot Autoguider
PHD guide
Canon 500D Gary Honis Mod w/Baader filter
Astronomik CLS , 12nm Ha , and OIII clip ins
Bahtinov mask
Canon EF 50mm f/1.8 II and Nikkor 180mm f/2.8 ED lenses
NexImage webcam
http://www.flickr.com/photos/31986095@N05/
|
Falcon-
professor emeritus
Reged: 09/11/09
Posts: 702
Loc: Gambier Island, BC, Canada
|
|
I was thinking the same as Erik - Your converting of the CR2 files to TIFF files does the debayer step *before* all the other steps listed in quantumac's proceedure.
Perhaps try processing again by just feeding the CR2 files to DSS rather then converted TIFFs.
--------------------
Intes Micro ALTER-M606: 6" Maksutov-Cassegrain Astrograph, 912mm f/6 
Tasco 11TE-5 'Lunagrosso': 4.5" Newtonian, 900mm f/7.9
Meade DS-2114S: 4.5" Newtonian, 1000mm f/8.8 (For sale!)
Galileoscope: 50mm Achromatic Refractor, 500mm f/10
Tasco EQ-2-like mount w/ clock drive
Celestron CG-5 ASGT mount
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
Thanks Sean, I looked at the re-process briefly and will email you back when I get more time. It proves how little I know about this stuff, but I am going to re-process everything using just the raw files instead of TIFF's and see what it does for my results. I don't know why I thought I had to convert everything to TIFF's first.
Also I am going to dig into some reading on the net and keep a notebook of things I learn from it all, most of what I've learned has come from this forum, and I am in constant awe of the results many here are producing. Give me a few days to play with what I have and I'll be back with more questions. 
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
Tonk
Post Laureate
Reged: 08/19/04
Posts: 4498
Loc: Leeds, UK, 54N
|
|
Quote:
and batch process the darks and lights from .CR2 into TIFF 16 bit files.
Nooooooooooooooo ! - don't do this - feed the RAWS straight into DSS. The RAW to TIFF has done three things that now work against you.
They have be "debayered" so dark calibration is compromised (debaying is the interpolation of new colour values from the RGB bayer mask pattern in the image). An individual hot pixel is usually a single pixel under a red or green or blue mask. After debayering this hot pixel data is spread into its neigbouring pixels wrecking the subtraction step. Typically a debayered green hot pixel is not completely removed but leaves a "X" pattern of pixels in you image. Red and blue ones do something a bit different - but what ever its wreckage
The RAW to TIFF conversion is non linear - again this wrecks calibration
A white balance correction has been applied. RAWs are not WB corrected but contain the WB setting for later use by software. Again this wrecks calibration
Yes this explains why your results are poor
DSS and other calibrating/aligning/stacking software ensure calibration is done before debayering, it makes sure that the conversion is linear and white balance is under your control (you can do it different to the camera settings). "daylight" WB is the norm you should use (with a DSLR camera) as it matches a white G2V reference star (our own sun is G2V! so "daylight" WB is near perfect)
--------------------
Televue 85, GM-8/Gemini, Canon 40D (unmodded), Canon 450D (modded w/Astronomiks clip-ins - UV/IR, OWB)
Coronado SM60/BF10, Baader Herschel Wedge
Leeds Sky Clock Ripon Sky Clock
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1223
Loc: Kentucky
|
|
Quote:
Quote:
and batch process the darks and lights from .CR2 into TIFF 16 bit files.
Nooooooooooooooo ! - don't do this - feed the RAWS straight into DSS. The RAW to TIFF has done three things that now work against you.
They have be "debayered" so dark calibration is compromised (debaying is the interpolation of new colour values from the RGB bayer mask pattern in the image). An individual hot pixel is usually a single pixel under a red or green or blue mask. After debayering this hot pixel data is spread into its neigbouring pixels wrecking the subtraction step. Typically a debayered green hot pixel is not completely removed but leaves a "X" pattern of pixels in you image. Red and blue ones do something a bit different - but what ever its wreckage
The RAW to TIFF conversion is non linear - again this wrecks calibration
A white balance correction has been applied. RAWs are not WB corrected but contain the WB setting for later use by software. Again this wrecks calibration
Yes this explains why your results are poor
DSS and other calibrating/aligning/stacking software ensure calibration is done before debayering, it makes sure that the conversion is linear and white balance is under your control (you can do it different to the camera settings). "daylight" WB is the norm you should use (with a DSLR camera) as it matches a white G2V reference star (our own sun is G2V! so "daylight" WB is near perfect)
Thanks for the detailed explanation, very helpful. I will re-process and see how it works out.
James
--------------------
http://astronomyguy63.blogspot.com/
Meade LXD75 SN6 with UHTC
Orion ST80
Cave Astrola 10" f/5
Garrett 15x70 binoculars
FarSight binocular mount with reflex finder
Orion Paragon tripod
Canon XS (1000D) and a Meade LPI
|
Previous
Index
Next
Threaded
Edit
Reply
Quote
[Re: 