Click here if you are having trouble logging into the forums
Privacy Policy |
Please read our Terms
of Service | Signup and
Troubleshooting FAQ | Problems? PM a Red or a Green Gu.... uh, User
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
I seem to be having some trouble with the color balance of my flats, and I was wondering what best to do.
I have a flat field box (foam core, couple of diffusers) and a LED light, which emits a bluish light. I use a modded DSLR with a CLS filter. I use a custom white balance when imaging, which works just fine to correct for the IR filter modification and the CLS filter, but when shooting flats, I noticed my red channel is underexposed, and therefore very noisy, while my green and blue channels are just fine (i.e., 1/3 of the histogram).
If I apply these flats, I seem to introduce unwanted noise into the red channel. I tried to use other light sources to shoot flats, such as an ordinary flash light, which emits yellowish light, and that gave me a red cast, with an underexposed blue channel. When I used my laptop screen, it gave me a green hued flat field. 
I'm sure I'm not the only one here with a modded DSLR and a CLS filter, so I was wondering what you guys with similar gear do to make balanced flats. Should I shoot my flats in monochrome mode, or should I make an additional custom white balance for shooting my flats?
Thanks for your input!
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
David Rosenthal
Carpal Tunnel
Reged: 06/23/06
Posts: 1645
Loc: NJ
|
|
What software package are you using to calibrate your lights?
The flats should be desaturated before dividing them into the lights. I use Maxim DL which takes care of this for you. I have a modded 400D and a CLS filter and I have not had to do anything differently when shooting with the filter vs. without.
--------------------
David Rosenthal | Midland Park, NJ
My Image Gallery
http://www.EastCoastAstro.com
Atlas EQ-G EQMOD USBDIR [Self-Hypertuned]
8" Meade SCT | Borg 76ED | Takahashi FS-60C
Self Modded 400D | SX MX716 | DSI Pro | SPC900NC
Series 5000 26mm, Hyperion 17mm, TV 11mm Plossl
|
zAmbonii
scholastic sledgehammer
Reged: 01/19/08
Posts: 841
Loc: Ypsilanti, MI
|
|
Well...your problem may be the LED lights that you are using. The LED light probably doesn't have a broad enough spectrum for the CLS to handle. The CLS may be cutting out the red that the LED is producing leading you to low values in the red channel.
I ran into the same sort of problem with using my laptop screen for flats + my LP filter. I tried using a light box with tungsten lights, but I seem to be getting a different problem, and have reverted to going back to sky/twilight flats until I get around to figuring things out with my light box.
--------------------
Check out my Astrophotos on Flickr
C6-N 150mm f/5 Newtonian
CG-5 ASGT mount
Canon 300D self modded + IDAS LPS-P2 FF
Meade 70AZ + Meade DSI for autoguiding
|
MaxTheHitman
sage
Reged: 12/11/07
Posts: 260
|
|
hey look DHB....I wake up at 6am to take flats!!!!!!!!!!!! they rock!!!!!! no flatbox no problems!! lol
for your info my gear is similar to you with cls filter and 300D  )
all you need white t-shirt and some patience as the sun starts its ascension!
have fun!
--------------------
---------
CG5-GT
canon eos 300D modified!
88mm f/5.6 scope imager
60mm guiding winner! with QHY5 all the way
Edited by MaxTheHitman (07/19/09 02:53 PM)
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
David,
I'm using DeepSkyStacker to calibrate and stack, and from what I've read in other threads on this forum, it should be no problem for DSS to tackle colored flats. That is, as long as all color channels are sufficiently illuminated. However, as zAmbonii mentioned, the LED I'm currently using is probably not emitting enough light in the red spectrum, and this appears to be introducing noise in the red channel from the inadequate exposure. As I said before, not only does the LED give me color imbalance, but so do an ordinary flash light and my laptop screen, albeit that the dominating colors are red and green, respectively.
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
David Rosenthal
Carpal Tunnel
Reged: 06/23/06
Posts: 1645
Loc: NJ
|
|
DSS should do the job fine.
I just opened up one of my RAW flats in PSCS2 and saw B*G channels at just below 1/2 and the R channel up at almost 3/4. This was with the CLS and a modded 400D (no added glass, no UV-IR cut filter) I attached a screen shot of the RAW converter in PSCS2 with the default settings. I wonder if I am overexposing my R channel. I am using an ordinary Tap-Light as the light source.
--------------------
David Rosenthal | Midland Park, NJ
My Image Gallery
http://www.EastCoastAstro.com
Atlas EQ-G EQMOD USBDIR [Self-Hypertuned]
8" Meade SCT | Borg 76ED | Takahashi FS-60C
Self Modded 400D | SX MX716 | DSI Pro | SPC900NC
Series 5000 26mm, Hyperion 17mm, TV 11mm Plossl
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
I had a similar result when I used an ordinary flash light. I suppose it's better to have a channel that may be a little overexposed than having one that is underexposed, since your red channel is probably just as noisy as your blue and green channels? At least that means you're not adding noise to the stack when using those flats.
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
David Rosenthal
Carpal Tunnel
Reged: 06/23/06
Posts: 1645
Loc: NJ
|
|
I have a flat box that uses 3 White LEDs as the source. I'll shoot a flat with that and post the same screen shot as a comparison.
--------------------
David Rosenthal | Midland Park, NJ
My Image Gallery
http://www.EastCoastAstro.com
Atlas EQ-G EQMOD USBDIR [Self-Hypertuned]
8" Meade SCT | Borg 76ED | Takahashi FS-60C
Self Modded 400D | SX MX716 | DSI Pro | SPC900NC
Series 5000 26mm, Hyperion 17mm, TV 11mm Plossl
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
I think I've got a solution!
I tried applying a CWB to my blue LED flats, which did correct for the color imbalance. However, when I looked at the 3 individual color channels, I noticed the red channel was still significantly more noisy than G or B. So the LED is definitely out now.
I then tried to apply the same CWB correcting technique on flats made with my ordinary flashlight, which emits the familiar yellowish light. Attached is a screenshot of the color balance before CWB, which is pretty much like yours, David. The red channel is obviously overexposed and not really usable like this. However, when using this image to set a CWB, you'll get a well balanced picture (see attachment in next post), which has pretty much the same noise level in all 3 RGB channels, which is very little.
Case closed, problem solved IMHO.
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
And here a screen shot of the custom white balanced flat:
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
David Rosenthal
Carpal Tunnel
Reged: 06/23/06
Posts: 1645
Loc: NJ
|
|
Interesting....I will have to give that a shot and see what the noise levels in the Red channel look like.
--------------------
David Rosenthal | Midland Park, NJ
My Image Gallery
http://www.EastCoastAstro.com
Atlas EQ-G EQMOD USBDIR [Self-Hypertuned]
8" Meade SCT | Borg 76ED | Takahashi FS-60C
Self Modded 400D | SX MX716 | DSI Pro | SPC900NC
Series 5000 26mm, Hyperion 17mm, TV 11mm Plossl
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
Last night I made some 30 flats "new style" to see if it really would affect the end result the way I thought it would. I made 3 stacks of 2 light frames:
1) Dark subtraction only
2) Darks + my old blue flats
3) Darks + "new style" flats, as described earlier
I opened the 3 images in PS, and applied the exact same levels adjustment, so that the noise became more visible.
The difference in the level of noise in the red channel between 1) and 2) and between 2) and 3) was significant, i.e., much more noise in 2) than in either 1) or 3).
When comparing the level of noise between 1) and 3) I noticed no visible increase in either R, G or B, which means the flats don't add (any noticeable) noise to the stack. So I guess that brings us one step closer to creating the perfect image. 
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
David Rosenthal
Carpal Tunnel
Reged: 06/23/06
Posts: 1645
Loc: NJ
|
|
I need to rethink this whole discussion, at least from the point of view of Maxim DL. I just opened up my last session's master flat (IEEE Floating point FITS image converted to 16 bit TIFF) in PSCS2 and it is a greyscale image (single channel). Therefore I am no longer sure how this applies to my workflow. I think I still want to try out the CWB for the flats just to see if there is any sifference in S/N
--------------------
David Rosenthal | Midland Park, NJ
My Image Gallery
http://www.EastCoastAstro.com
Atlas EQ-G EQMOD USBDIR [Self-Hypertuned]
8" Meade SCT | Borg 76ED | Takahashi FS-60C
Self Modded 400D | SX MX716 | DSI Pro | SPC900NC
Series 5000 26mm, Hyperion 17mm, TV 11mm Plossl
|
DHB001
super member
Reged: 07/07/09
Posts: 145
Loc: The Netherlands
|
|
I know what you mean. I thought my LED flats would also be somehow transformed into greyscale in DSS, which would probably result in a usable flat, but apparently it doesn't work that way, in DSS at least.
Could very well be that Maxim works differently, and that having one under- or overexposed color channel makes no difference for the master flat.
--------------------
Regards,
Boy
C6 SCT @ AS-GT mount
SW 70/500 guide scope
Canon Rebel XS / EOS 1000D, self modded
Astronomik CLS EOS clip filter
Orion SSAG, PHD guide
Bahtinov focusing mask
|
|
4 registered and 13 anonymous users are browsing this forum.
Moderator: Charlie Hein, knuklhdastnmr
Print Thread
|
Forum Permissions
You cannot start new topics
You cannot reply to topics
HTML is disabled
UBBCode is enabled
|
Thread views: 456
|
|
|
|
|
|
|
Previous
Index
Next
Threaded
Edit
Reply
Quote
[Re: 
