5.61
#...
·
·
1
like
|
|---|
|
This can be a problem, especially at a remote site (where my 2600 is located). They really need to offer a shutter but (almost) no one does. That is why my next CMOS camera will be a Moravian, not a ZWO or QHY, etc. They have a shutter. In the meantime, I do them on cloudy nights and have set up my flip flat with a magnetically attached black 1/16 inch plywood cover. I have the tech remove it for flats. I also seal it with black felt on the front of the dew cap (the scope needs to have a dew cap that will accept that). More dustproof as well as better for darks.  |
2.41
#...
·
·
1
like
|
|---|
| Since I use DSS, I enabled Dark Optimization to level out temp differences between dark frames. I shot all of my darks for all the different Gains and Exp times that I use all at once in my utility room with the cap on the camera, inside a box. It is my static dark library I use for all of my sessions. Works great for me. |
0.00
#...
·
·
1
like
|
|---|
| I take the camera and NUC to a dark corner of my garage in winter, basement in summer. |
0.00
#...
·
·
1
like
|
|---|
Stuart Taylor: Same as a few comments above, inside a small box. |
 #...
·
·
1
like
|
|---|
Antonio Jose Paris: I put it rather loosely in a bag (actually the soft camera case that came with the ZWO camera) with the front end pushed into the dark end but the back where the cooling fan is still relatively open. Then I put it into the bottom drawer of the fridge at 4 C connected up to a computer via two cables that are flat enough not to cause any problems with door closure. The main hazard is my wife getting cross if it is there too long and/or I haven't taken essentials like milk out before I started. Tim |
0.00
#...
·
·
1
like
|
|---|
| I do not, since I don't see any benefit from using them (IMX571 mono) |
0.00
#...
·
|
|---|
|
I came upon this thread after shooting my dark library on my new 2600 yesterday afternoon. I ran them with the camera attached to a filter drawer and scope with the lens cap on, but with no other precautions as far as stray light goes. Indoors with the lights on and the camera set to 0 C. Out of curiosity I re-ran the longest exposure dark frame with the same setup but in my windowless basement with the lights off. Gain 300, 20 x 180 sec, stretched on Asiair plus with histogram sliders set to zoom auto. My ingers are too fat to set the sliders identically between two shots. Upstairs: Max 65535, Min 228, Avg 507, Std 90 Upstairs image Basement: Max 65535, Min 223, Avg 506, Std 86 Basement image Visually, I can't see any difference. In both images, there is a color shift from left to right. That looks like exactly what the dark frame should be correcting. I'm doing EAA currently, not AP, so I don't think I will re-shoot the entire library. |
10.10
#...
·
·
1
like
|
|---|
Stuart Taylor:Pedro A. Sampaio: For ZWO cameras, the offset "value" is 1/10 of the ADU offset. Ie. setting the camera to an offset of 50, results in an actual offset of 500 ADUs. Why? I dunno. |
6.05
#...
·
·
1
like
|
|---|
|
I unscrew the light bulb in the refrigerator and take my darks inside it. I take 25 darks starting at 30 sec and increase 30 sec until I have taken 25 300 sec subs. I do this for 0 gain, Unity and 200 gain at 0 degrees C and -10 degrees C. I create the MasterDarks with PixInsight. CS - Tim |
22.76
#...
·
·
1
like
|
|---|
Pedro A. Sampaio: Darks capture the dark current signal. The dark current signal is what gets subtracted from the light subs. The noise that you get in your dark data is due to dark current shot noise, and read noise, which includes all noise that is not exposure dependent. Noise in the master dark data always adds in quadrature to the noise in each sub. John |
|
22.76
#...
·
·
1
like
|
|---|
|
Like Bill, I have a shutterless QHY600M on my CDK20 in Chile. I have a "dark" panel set up next to my light panel for taking darks. It's just a large, flat black panel set up just an inch off the end of the scope. I can only shoot darks at night by pointing at the dark panel. My darks are taken at both 300s and 600s which requires a pretty dark night even with the dark panel. Since I don't use my rotator (at least yet) my darks last for a very long time (like years). John |
|
10.10
#...
·
|
|---|
John Hayes:Pedro A. Sampaio: |
|
10.10
#...
·
|
|---|
John Hayes:Pedro A. Sampaio: |
|
10.10
#...
·
|
|---|
|
With my cooled camera, I create a dark library via capture sequence softward (Nina or SGPro or whatever) at my normal sensor operating temperature of -10. I will do this indoors, with a non-leaky lens cover (on cloudy nights, when my camera will be available). You will likely want >100 bias and >20 darks at a few spaced out exposures (dark noise it is roughly linear with exposure) to create a set. Create a set for the various gains and binning (and temperatures, if you want). Then integrate, without calibration, the bias and dark images to create masters at each gain / binning / temperature. Now comes the key - when you have your lights/flats calibrate by choosing the bias, and closest master dark for your light exposures. If you use software that is able to optimize the darks (such as Pixinsight), it will adjust the master dark to best fit the exhibited correlation of noise that exists in your light. This correlation generally should be equal to the ratio of exposure times, if the noise varies linearly with exposure time as theory suggests. In any effect, I no longer have to take darks with my telescope at sessions. I find my master dark files are good for a maximum of a year. Generation of the original dark frames can take between a few hours and a day for each gain/bin/temp setting. One can do similar for a non-cooled camera by changing darks at temperatures intervals, and selecting the closest temperature to your imaging session. The same optimization of the darks will take place, but note that the noise is not linearly temperature dependent. The closer your library can come to your actual temperature the better, but note that after all temperature will vary through the night too, and the optimization will do a pretty go job at fitting your dark noise to your light frames. With a non-cooled camera, one can also build up your library as you go - just save the masters and label them with the parameters. Master Darks are optimized (bias subtracted) at the time of light calibration, so you shouldn't calibrate the darks prior to integrating them creating masters. No need for flat darks, if you use a short exposure as one of you master darks. Hope this helps. Dave |
3.61
#...
·
·
1
like
|
|---|
John Hayes:Pedro A. Sampaio: I stand absolutely corrected. Cheers! |
|
10.10
#...
·
|
|---|
Pedro A. Sampaio:John Hayes:Pedro A. Sampaio: Just to clarify. Yes the purpose of the darks is to capture the electronic noise of the sensor. Some of that noise will be +ve (higher response pixels), and some is -ve (lower response pixels) from the mean. The purpose of the offset is to ensure that both are captured. If an offset of 0 was used, any -ve noise would be clipped, as -ve ADUs are not allowed and only +ve noise would be recorded. So we set a pedestal (offset). -ve noise will be recorded as an ADU value less than the pedestal and +ve noise as an ADU value greater than the pedestal. The selection of the exact pedestal is a trade-off between ensuring the capture of -ve noise and the encroachment of the pedestal on the well depth / dynamic range of the camera. Ie. you want it low enough to avoid using up the total ADU range of the camera, while ensuring that both very -ve noise (negative in analog signal) are recorded (in digital ADUs). The default offset of 50 (corresponding to ADU of 500) is likely what you are using, or it is a combination of coincidence and a very noisy camera/light leaks that would yield an average ADU of 500 in your darks. More likely, that the -ve noise and +ve noise are approximately cancelling each other out, yeilding a mean approximately equal to your offset. I am sure if you checked, individual pixels in your dark are higher and lower than 500 and even the mean that you quote won't be precisely (exactly) 500. Total noise in your darks is approximately linear with exposure time. The bias (after subtraction of the offset) is the intercept, while the darks (after subtraction of the bias) yield the slope wrt exposure time. A sign of "light leaks" in your darks would appear if the average ADU of your darks was very much greater than the offset=50/mean=500 that was employed. Hope this helps, Dave |
4.89
#...
·
·
1
like
|
|---|
Stuart Taylor:Pedro A. Sampaio: None of these numbers look concerning to me, other than that your offset is probably higher than needed. An offset of 20 or so would probably give equally good results and gets you a tiny bit of extra dynamic range. In itself these numbers don't tell the whole story, but you say that stretching these darks you see so sign of any gradient, it's fine. The maximum signal is just some 'warm' pixels. If you take the camera off its filterwheel and screw the cap on, it is hard to imagine any light leak coming in, so you can just put it in any room that is cold enough to allow the camera to get to the desired sensor temp. Optical trains (focusers, filter wheels etc) almost always have light leaks, so darks best taken in a dark environment. But camera+cap should be good enough to use anywhere. |
|
10.10
#...
·
·
1
like
|
|---|
Willem Jan Drijfhout:Stuart Taylor:Pedro A. Sampaio: Nothing looks wrong to me either. I would expect a mean significantly higher that your pedestal/offset (500/50) if you did and a higher deviation. The high maximum is a hot pixel, while the minimum is just a lower responding one. It is a skewed distribution (I would expect log normal about the pedestal). Note that you may be able to get away with a lower offset as suggested, but it is important that the same pedestal (in your case 500 ADUs) is used for your flats and lights for correct calibration. If you make a mistake in your camera settings and end up with a different offset/pedestal, you can correct the frames by addition/subtraction of the appropriate number of ADUs using pixel math (and an image container). For example, if you decided to use an offset of 20 going forward, you can just subtract 300 ADUs from you master bias/darks and still use them for lights/flats with that offset. Clear Skies! |
4.57
#...
·
·
1
like
|
|---|
Stuart Taylor: On the 2600 the offset is multiplied by 10, so an offset of 50 should result in a mean of 500, ideally. You are just about spot on 500 so all is good. BTW, you mention trouble capturing biases. I have never found a need for biases with the 2600 if you make a dark library for the exposure duartions that you shoot; the darks have all the bias data in them. |
|
10.10
#...
·
·
1
like
|
|---|
Roger Nichol:Stuart Taylor: I think this is the case for all ASI cameras (16 bit anyways). Just to confuse people, ZWO define the offset as one tenth of the ADU offset that will actually be used. I guess I didn't explain this, but I did try. |
0.90
#...
·
|
|---|
|
I'm new at this and I have just built my Rig, (Asiair Plus, AM5, Redcat 51, EAF, Asi2600MC Pro, 120mm Guide Cam and scope, 2 Dew heaters) I have been watching YouTube vids for hours and hours. I just been studying Taking Dark's and Flats. The way I saw people do it on you tube is just put the cap on the end of the scope before or same time you do your light frames so the temp and environment conditions are the same as the image session. Now I read here people are going to the extreme and putting there camera or rigs in boxes and refrigerator's ext.... I've heard you only need to do at the time of imaging 20-30 dark's. I see one guy ran his all night??? So many different ways doing the same thing kinda of confusing... |
0.00
#...
·
|
|---|
|
Hello, I put the camera in the fridge at night while I am sleeping for several nights to create a good library of dark images with varying exposure times and temperatures. Joan |
0.00
#...
·
·
1
like
|
|---|
|
I just wait for a night where the cooler will be able to hit my desired set point. Leave the camera on the scope as it would normally be, with the lens cap fitted. Wrap a couple of towels around the lens cap to provide more of a barrier, and wrap the focuser tube in a few layers of Aluminium foil to black out any potential light leaks that may be present... Cool the camera to my desired set point, set up a sequence in NINA, start it, and check the first exposure to make sure nothing looks 'unusual'. If it all looks good, Restart the sequence, turn it all off when you're done. I replace my dark library every 12 months, and I have summer darks and winter darks as I will run my camera as cold as possible, which in winter is considerably cooler than summer. I have: For -15°C and -25°c, bin 1x1 and bin 2x2 (so 4 sets) 20x 1800s, 1200s, 900s, 600s, 300s, 180s, 120s, 60s. This way, I know, no matter if I'm shooting ultra deep narrowband in winter temperatures, or if Im shooting short -15°c R/G/B binned data in summer, I have the darks to match my lights. It takes a few nights a year to capture these frames... and once they are done, I will stack each set (median combine) and keep only the resulting master dark file (no point keeping 1200 fit files on the hard drive and re stacking them) So I'll have for example, files named 'MasterDark_-15_b1_60s.tif' and 'MasterDark_-25_b2_1800s.tif' I also keep a library of flat darks, because I know the required exposure for each filter etc, so they will be named according to the filter their exposure time corresponds to, ie. 'MasterFlatDark_-25_b1_Ha.tif' and 'MasterFlatDark_-15_b2_R.tif'. I will only shoot 10 of each, as their impact on an image is much lower than that of the actual dark frames. It takes a lot of time to get it all done, but it means that for maybe 1 week a year, you're shooting/stacking calibration frames. and for the rest of the year, you take your lights, and associated flats and that's it. Well worth the effort I think. |
4.69
#...
·
·
2
likes
|
|---|
Alex Nicholas: Interesting that you also shoot a library of flat darks. I would not be able to do that as my flats exposure is calculated each night by NINA's flats wizard and can be all kinds of exposures e.g. 0.89 sec, 1.14 sec etc so my library of flat darks would have to be infinite! |
|
10.10
#...
·
·
1
like
|
|---|
Stuart Taylor:Alex Nicholas: I am not sure what processors other than Pixinsight allow for optimized darks. For flat darks, I use 5 second darks for my sky flats. Pixinsight takes out the bias from these darks, leaving only time-linear noise and scales it to the actual flat exposure times. It does the same for the darks used to calibrate light frames and there is no need to have your darks exactly match the exposure time of the lights. Hope this helps. Dave |
