Address Noise By Cultivating Signal Quality
Vložit
- čas přidán 9. 06. 2024
- Killing noise starts long before you get to denoising apps. To really address noise, cultivate good imaging fundamentals and developing processes that maximize signal quality every step of the way.
*** *** ***
Check out a portfolio of astro imagery shot from the Sky Story Observatory at Astrobin: www.astrobin.com/users/SkyStory/
We are Telescope Canada's first affiliate. It's a great place for astrophotography gear. Using the following link adds nothing to your cost but helps us produce more educational videos about this amazing universe around us. telescopescanada.ca?bg_ref=KK15sLXmWw
All videos best viewed in 4K.
Peer into the cosmos on Sky Story and along the way come to appreciate our world all the more. Here, we explore the realm of nature beyond our Earth through the fields of astronomy and astrophotography.
All our programs are committed to the standards of scientifically accurate, high quality content that aims to educate and inspire. New videos are posted almost weekly, so please subscribe.
#astronomy #astrophotography #space #nebula #stars
You had me subscribed at "Blink sucks".
Best astrophotography channel on CZcams, congratulations! Big fan here!
Many sincere thanks!
Spot on. Alot is beyond my skillset and gear right now, but still a lot is really useful for me as a beginner. As i improve i can go back and watch the more advanced bits
Thanks for another great video. I agree with everything said here. I honestly don't know how you find the time to put out these high quality videos at such a rate. Salute
Great video as always. Seeing that you take 60 second subs. How often do you dither?
This video is great. Never underestimate camera and scope compatibility. It can make or break your hard work collecting data.
Highly recommend nina flat wizard for flats
11:48 this is the big problem in microchip manufacturing. As area increases you can fit fewer and fewer chips on a single wafer, but the same size speck of dust that will kill a 1x1mm chip will kill a chip 10-20x its size. Thus the yield will decrease exponentially with increasing area.
Great video.
I wanted to ask you about stacking especially when you stack huge number of subs as you described together with taking flats for each session. So do you calibrate and stack each night separately of you calibrate by session and stack all the calibrated images at the end of the project? What's the best practice? thx
There's no way to tell how good the informaiton is going to be even if I image all night. Could be clear as a bell but bad seeing. Or clouds could roll through during the night. There could have been some kind of tech failure. So I tend to stack all the data the next morning and see how far I've gotten. I just keep adding integration time till I feel the image has reached a standard of appearance I am happy with.
@@UNDERST0RYhow do you “add integration time?” Do you take a stack from each night and then stack those sub-stacks together, or do you run night 1 on morning 1 and then nights 1 and 2 on morning 2 and knights 1, 2, and 3 on morning 3, etc…?
@@ryanwood1911 I save all the subs till the project is done. In truth, I don't usually throw culled subs away anyway. You never know when they'll come in handy. I also save the associated bias and flat frames. Each group of data is co-stacked in WBPP in PixInsight by dividing the data and calibration frames into sub folders. I'll make a video on it some time.
@@SKYST0RY A video about that would be nice. 👍
I am concerned about the Astro Tools CCD suitability calculations. They use image scale and seeing. I’m considering pairing the Redcat71 with ASI2600, which gives about 2”/pixel, which is “slightly undersampled”. But nobody seems to talk about the airy disk diagrams that manufacturers, like William Optics, provide. That scope has an airy disk radius of over 3um in the center of the frame, and it gets larger as you move out. It’s a very good performance compared with most scopes. To me, it sounds well matched to 3.76um pixels because that would be about 2pixels per airy-disk diameter (Nyquist). Isn’t the airy disk size the best resolution possible, regardless of aperture, focal length, and seeing? If you have a scope’s airy disk performance, shouldn’t you match that to the sensor pixel size? The pixel scale and seeing are also limiting factors, but I think often the airy disk might be the limiting factor. Or am I missing something.
Good question, but not one I've really thought about before.
Dss gives a score for each sub after a quick register, even if you dont stack in dss
It's been so long since I've used DSS I barely recall it. WBPP in PixInsight is such a superb stacking tool that I pretty much only use it for DSOs these days.
I shoot OSC and use basically the same PixInsight workflow you describe. One question about Drizzling. At what point (if any) do you resample down. I use Adobe Photoshop for final processing and if I don't resample I often end up using the large document file format to save intermediate steps and the files can become amazingly large not to mention Photoshop responding sluggishly since It's almost time to get a new computer with a bit more horsepower under the hood.
Affinity Photo has the same weird sluggishness with very large image files. Frequent saving addresses it. I figure it's a price worth paying for the benefits of drizzling. I never sample down. I've found in testing it's not good for the images. If you go to my videos page and look up drizzling, I recently did a whole series on the benefits of drizzling and experiments to determine best practice.
@@SKYST0RY I'll do that, thanks!
Not able to find the software you mention for culling sub-exposures. I've tried Earthin View, Earth and View, Earth in View etc. The exact name you give is not clear. Could you please provide a link. Thank you.
Sorry. The sub culling app is a tool in PixInsight called the Subframe Selector. It's very good for DSOs. I'll do a video on it some time.
I think Irfanview was mentioned. From an earlier community post: @mr.b3054 Usually I shoot 30-60s subs these days, so there are anywhere from several hundred to over a thousand subs to examine. I can breeze through a visual inspection of these subs using Irfanview in a few minutes. For visual inspections, I just discard anything with a satellite or plane track in it, or anything that is clearly flawed (perhaps by cloud). Then I let SFS do a cross check. I have a pretty good SFS routine that assesses for FWHM and several other factors that cleans up the surviving subs pretty nicely.
If shooting even shorter subs (like with lucky imaging where there will 3600 to 10,000 subs/hr, I do the whole routine on SFS. SFS can assess several thousand subs in a few minutes.
Sorry, I see what you mean. The software is Irfanview.
I have heard the Narrowband channel and some other people say that stacking too many DSO subs has diminishing effects and as such for the given integration time you would get better results with longer but fewer subs. for example 100 5min subs are better than 1k 30 sec subs. Have you observed such a thing in your experience?
I think you are referring to the law of diminishing returns. You get the most benefit early on in shooting and as you continue to accumulate information, you get less overall benefit but it shows up as finer and finer detail and better resolved color and dynamic range. The longer you go, the subtler the improvements become (though they may still matter). They may also have been referring to shot noise or read noise. Read noise isn't much of an issue with modern sensors but it can certainly accumulate if you are shooting from a region with light pollution. From a dark sky site, shot noise is virtually a non issue and the value or more integration time outweighs it by pushing up SNR faster than the shot noise. But since I don't have to deal with light pollution, I don't honestly know much about how to handle it.
@SKYST0RY Thanks so much for replying! Argument was related to the concepts you mentioned but not exactly the same as I understand. Some people argue that for the realy faint targets, the mono camera is better because when shooting lrgb separately, you get to stack 4 times less subs for each channel than with osc for the same integration time (Stacking 64 each of lrgb subs vs 256 of osc). So they get affected less by the diminishing returns than osc. Similarly, getting same integration time with fewer subs would yield better results. To me, this sounded both really right and really wrong at the same time. So I wanted to understand what's really going on here.
Narrowband channel posted video 6 month ago (mono vs osc resolution and details which is better?) talking about this from 7 min onwards (in case I butchered the argument)
I agree with taking flats with different gain is also ok, I tried myself. However, I was doing specific darks/bias with the same gain as the flats to calibrate them.
Because shooting bias frames is so easy, I don't think I've ever experimented with changing their gain. But given that gain really doesn't mean what it used to, it shouldn't make much difference with a modern sensor.
@@SKYST0RY So, you only use bias frames to calibrate offset of your flats and lights, right? I guess you don't use same temperature for the bias either... Interesting approach. I would like to run some comparison tests myself. Thanks!
@@JoseLausuch I don't really worry about temperature on bias frames or flats. They get shot at -5C on the Ares-M because that's the temperature I operate it at, but on the Uranus-C (which is a planetary camera) they get shot at ambient temperature.
An excellent video and tutorial.
Please can you provided a link to the image inspection alternative to Blink.
I agree that Blink really sucks and I would like to try out the alternative.
Thanks 🙏🏿 in advance!!
Blink would be much better if it allowed you to delete the files instead of moving them to another folder and deselect...
I should have included a reference to that in the link. The app is Irfanview. You can learn about it and download it here: www.irfanview.com/main_download_engl.htm
@@SKYST0RY y
Thanks 🙏🏿!
@@SKYST0RY not for Mac :(
There is a dither calculator online worked great for my walking noise. Just input gear details :) which im sure you know.
Yep. I did a video referencing it a few weeks back in the video about how to calculate dithering.
@@SKYST0RY yup i followed your video and works great for me on 2min subs, every 3 subs. Gonna do same for 1 min subs :)
Another good example is iairplanes or satellite s across a 10min sub ruins 10mins. Only ruins 1 min of a 1 min sub. Only down side to shorter subs seems (i could be wrong) to be more storage space used and more stacking time. So make a coffee :) these streaks may be removed via stacking but maybe not
One is that short subs use more storage space. But I save directly to a 1 TB SSD, and even 1 second subs will not come close to filling it up on a winter night when I get more than 12 hours of darkness. Memory is so cheap these days, I don't figure it's worth worrying about. The biggest downside to short subs is shot noise, but that only applies to persons who live in areas of light pollution.
@@SKYST0RY i have bortle6 so im giving 1min subs ago to dee how it goes on my stock dslr
@@SKYST0RY What about long term storage of those subs? Or do you trash them after you stack them?
@@ryanwood1911 I keep them till at least a project is done. Generally, permanently. I just accept getting a new 14 TB storage drive is necessary every couple years.