0.00
#... |
|---|
|
Hi astrobins, I 've this week gathered data from Hearth & Soul on three nights, total about 13hrs. Equipment used: Nikon D750 and a 180mm ED tele Question: using DSS & Photoshop, the picture from night 1 came out best (4hrs). When adding more subs from night 2 & 3 I cannot process to that same quality level anymore. I would expect the image to get more detailed, finer texture, more pronounced details you get the "picture". So why is that not the case? Love to hear from you!  Joost Hoegaarden |
3.82
#... |
|---|
|
Maybe there are cloudy, noisy, not well guided subs from the following nights. Also take in to consideration, the first 20-30 subs reducing the noise dramatically, but after a certain point it will be not so spectacular. |
|
0.00
#... |
|---|
|
True if that was the case. The subs are however excellent. Well guided, perfect round stars. The second night transparency was a bit less, but th elast night are 48 subs of equal quality. I tried adding ight 1 and night 3. Somehow I have the impression dss cannot get the additional integration time valued into quality. How do you integrate 10 hours of subs? ith Deep-Sky STacker? Cheers Joost |
11.57
#... |
|---|
|
Hi, There are many possible reasons. DSS does have its limit, but hitting this limit is not easy. It will be much easier for us to tell what's going on if you can show us some pictures. Generally speaking, 4 hr on 180/2.8 with aperture fully open is quite lot of exposure. The image should be already very deep. Whether adding more exposures can bring improvement would depend on many other factors. For example, it is possible that your image is overwhelmed by many stars, so it is not easy to see the improvement in the underlying nebulas. This is my experience on my 180/2.8 lenses. Also, when I stack many exposures from multiple nights, I usually first stack images from each night separately. Then I further stack the image in programs like Registar. This just make things much easier. Cheers, Wei-Hao |
0.00
#... |
|---|
|
After 5 years of learning and using DSS, hit the wall with it. So, now use PI just to stack my stuff. Superior noise control , & will stack every time. I have used & still do StarTools for my post processing now. Just had to give it up, DSS that is. For what it's worth. PJM |
0.00
#... |
|---|
|
Most probably I had a similiar experience/problem: 3 nights of exposure but half the integration time (6-7 hours) it turned out that adding the last night lead to a worst final image, this wasn't easy to spot, I had to apply the same operations (a minimal set of strectch and background removal just to have an image to look at) on different stacked sets to understand it Apparently the sky was "more or less" as the two previous nights, also, the Score in DSS was not changing significantly. Most probably the problem was in the sky transparency, a thing that maybe it is difficult to notice in single frames but that alters the SNR of the DSO I didn't even understand if the problem was caused by the whole set or by just a group of bad frames the third night, I don't know if there is a way to filter bad frames, e.g. calculating the SNR of the target DSO on the single frames and discarding those with lowest value |
|
11.57
#... |
|---|
|
You can ask DSS to stack only certain fractions of the frames (such as 50%, 80%, or 90%) instead of stacking all of them. If you do so, DSS will rank the images according to their "scores" and skip the bottom portions of the images in the stacking. This is one way to get rid of bad frames automatically. However, as there are no official documentation about how the scores are calculated in DSS, there is no way to judge whether this is really optimal. All you can do is to try it and see if you like the results, not very scientific. The situation is slightly better in PixInsight. It has a weighting scheme that weight every image differently during the stacking. This should give you better S/N when you stack images taken under different conditions. On the other hand, I am not sure how the weights are calculated either. You may need to go through the documentation and see if there is any info on this. Again, without knowing the details about the weights, it's hard to judge whether it is really optimal. That being said, I am not a fan of throwing everything into the pot and stacking them all at once. This is especially problematic when the conditions in the multiple nights are different, when coupled with algorithms of rejecting outliers (sigma clipping, or even median) during the stacking. The variation in image background (and gradient) can interfere with the outlier rejection and produce results less optimal. A much better way is to stack images taken under similar conditions separately, and then stack them to get the final deep image. When further stacking the stacked images, one needs to pay attention to the relative weights. Images taken under poor conditions can be still used, just with lower weights. I am not sure how to do this in PixInsight. I use Registar to further stack the images stacked by DSS, and Registar allows me to give images different weights. |
0.00
#... |
|---|
|
Dear Joost, One thing you can do is to look at the quality of your subs - it sounds as though DSS has version of this (I've never used it), but you can also use PixInsight, or CCDInspector to measure the quality of your images (FWHM, eccentricity, median I've found to be the most useful). If the seeing is worse on your third night,, for example, you will make your images worse by including these subs. With PI I use the SubFrame selector script - I choose a FWHM limit (the limit variies, but I always get rid of the worst frames) and also demand Eccentricity < 0.6, and I also have a limit on the median value to exclude images taken too close to sunrise, or the moon! Colin |
