| Bias frames or Dark flats | |
|---|---|
| Bias | |
| Dark flats | |
| Doesnt matter, it creates same result | |
| Login to vote and view results. |
1.91
#...
·
·
2
likes
|
|---|
|
I just got my 2600mm pro (coming from 1600mm pro) and on my old camera I didnt do bias frames, but dark flats instead. What should i do with this new camera? I use NINA so after the flats are done the darkflats come automaticly, so itdoesnt really matter for me. I just want to know whats best to use.. Flats (of course) Dark flats? Bias? Darks? (I know it doesnt have amp glow, but darks reduces some noise right? Also about the gain.. Gain 100 is the way to go right? but I have seen that some people uses gain 101 "just to be sure"... thoughts? also the temp, is -10 enough or should I go for -20 if i can? UPDATE: There are still much confusion about darkflats or bias frames... Poll created! I also got an answer from ZWO: "Simon Lewis (ZWO) Aug 10, 2023, 2:22 PM EDT Hi Daniel thanks for reaching out The 2600 is fine with just bias /dark frames but will do equally well with dark flats Ive tested both the results are the same. Gain 100 ... its fine at 100 yes not 101  Temp I use -10 and its very well behaved - its a v low noise camera already Hope thats helps thanks" |
0.00
#...
·
|
|---|
| Dark flat have no sense with your camera !! |
11.99
#...
·
·
3
likes
|
|---|
| The purpose of taking darks is not to reduce random thermal noise, which is impossible to eliminate, but repeatable non uniformity in the baseline value of pixels ( mean dark current pattern or signature). All sensors, whether they have amp glow or not, benefit from this. You don’t need to take bias frames for CMOS cameras unless you plan to scale your darks. You can simply calibrate your lights and flats with darks taken at the same time and temperature, as it appears you did for the 1600. |
1.20
#...
·
·
1
like
|
|---|
|
Hey Daniel, I saw a video from a geek I trust (can't find the link as of now), who stated that modern cooled astro cams do not benefit from darks flats. (as long as you dither) He showed a side by side with and without those and pixelpeeped for 10min. There was absolutely no difference. I have a 2600 and I take flats, bias and that's it, never saw anything bad happening to my stack. In 99% of the cases you use gain 100, if I remember correctly you can't even go above 100 anymore, they removed the option.  A quick peek at the charts and you'll see why, 100 acitvates HCG with the best DR to noise ratio. Unless you need the absolute last bit of DR for some reason, you stay at 100 and never touch it again. Recently my software reset the gain to 0 and I didn't notice. I was wondering why the subs looked significantly worse than usual until I found out. |
|
11.99
#...
·
·
1
like
|
|---|
|
With respect to the dynamic range - remember that the dynamic range at gain 100 is achieved by reducing noise but also by substantially reducing full well capacity. Gain 0 allows you to use the complete full well capacity of the sensor. The noise difference - 3,5 versus 1.5 - is not a major concern. Read noise is very rarely the major source of noise at these levels unless you are using short exposures. Also- it is fine to calibrate your flats with just a bias, as long as your flats are short in duration (won’t work for the 294MM but probably fine for the 2600). When people say dark flats are unnecessary- they are basically making an assumption that the mean dark current is small for short exposures. When you take dark flats, you don’t have to make that assumption. |
|
1.20
#...
·
|
|---|
Arun H: That is very interesting, that he didn't mention, do you have a link to that information? |
|
0.00
#...
·
|
|---|
Arun H: It's not true Arun, I have the 294MM and work perfectly also with 30/40s of flat (SkyFlat). Darkflat was necessary with old CMOS cameras having really bad changing of pattern related to the time used for flat .. this is something true for the 1600 sensor or old DSLR .. not for the latest CMOS sensor and especially for the 16bit. |
|
11.99
#...
·
|
|---|
Arun H: the charts supplied by ZWO. Observe the full well capacity reduction between 0 and 100. This is basically how sensors work. When you increase gain, you cut off full well capacity.  |
3.01
#...
·
|
|---|
| At a minimum, flats. I do darks, dark flats, and flats for calibration. I don't see a need for bias since with cooled cameras you can take darks for the specific temperature. For the target temperature, depends on ambient temperature. I try to keep the cooler at 80% power or less. Might push to 90% in the summer with the assumption that it will cool down overnight. You want the cooler to be active, else you'll get more variation in the sensor temperature. For gain, I run either 0 or 100. I try both on any setup to see what exposure times are like. It's possible I might go gain 0 if the subs are very short for gain 100, typical for L filter, sometimes also for RGB filters. For 100 vs 101 gain.. run the camera through SharpCap sensor analysis and see what it tells you. I had an ASI533MC I ran at gain=101 not "just in case" but because the analysis for my specific sensor recommended 101. HCG kicks in at 100. |
|
1.20
#...
·
·
2
likes
|
|---|
|
Hey Arun, I read through a discussion on cloudynights. It is actually hard to say what is better, both have tradeoffs. The essence from that discussion is, broadband go for gain 0, narrowband with a fast scope go with 0, narrowband with a slow scope = 100. The differences seem so minuscule, I assume you won't see it in the final picture anyway. But I learned something new, since I have a slow scope (F7.5) and shoot mostly Narrowband I seem to have the optimal gain by coincidence. I will try 0 for my broadband targets. |
5.70
#...
·
|
|---|
|
In my limited experience and from previous discussions a few years ago about this, I've found that Bias frames don't work well with CMOS sensors. They're more well suited for the older CCD cameras which are becoming few and far between now. In addition to taking darks and flats, I exclusively use Dark flats for calibration of the flats and I'm very happy with the results. I don't bother with Bias at all. CS, Patrick |
4.52
#...
·
·
1
like
|
|---|
|
The only cameras that require dark flats (or flat darks) are the ASI 1600 and 294 (and their equivalents from other manufacturers). The others work fine with bias. Now in practice there is no difference in results between using bias or dark flats, but taking and processing dark flats is more tedious as they require the same exposure as flats. Taking bias is much easier and faster and works for all cases. So unless you are dealing with ASI 1600 or 294, use bias. |
7.12
#...
·
|
|---|
|
I only take Flats, and calibrate with Bias frames. Seems to be doing fine. As to the temperature question, it depends on your skies. For most locations with light pollution, anything less than 0 has no significant benefit. This presentation sheds a lot of light on some of your questions and is highly recommended. https://www.youtube.com/watch?v=3RH93UvP358 |
1.81
#...
·
|
|---|
|
I switched to the 2600mm from the OSC version, I keep a gain of 100 with temp sent to -10. I feel like I have a lot of noise in my Oiii and Sii channels. I'm shooting lrgb on iris nebula with the same parameters and feel like there was a lot of noise in Lum when I zoomed in on the preview in the asiair+. I feel like I'm doing something wrong or missing a detail for my image capture. My scope is the Askar FRA500 at 5.6. Should I drop my gain down to 0 at this speed for broadband and narrowband? I use darks, flats and now bias for callibration (I stopped using dark flats recently) |
|
11.99
#...
·
|
|---|
The essence from that discussion is, broadband go for gain 0, narrowband with a fast scope go with 0, narrowband with a slow scope = 100. There is rationale for that recommendation, which is a sensible one. For narrowband, with a slow scope, there is the possibility that you simply are not getting enough sky background shot noise to overwhelm the read noise, even as low as it is. It is much less of a concern for broadband/faster scopes. The key purpose of my post was to highlight the great benefit of the IMX 571 sensor - which is that, in many cases, you can use the complete full well capacity of the sensor by operating at Gain 0, and do so with very little or even no penalty. This is not something you could do with older sensors - where you had to worry about things like banding at low gains, the only way to get rid of which was to operate at higher gains to reduce the amount of sky background noise you needed to overwhelm the banding. |
|
1.91
#...
·
·
1
like
|
|---|
|
I just got answer from ZWO regarding the question here is the response: Simon Lewis (ZWO) Aug 10, 2023, 2:22 PM EDT Hi Daniel thanks for reaching out The 2600 is fine with just bias /dark frames but will do equally well with dark flats Ive tested both the results are the same. Gain 100 ... its fine at 100 yes not 101 Temp I use -10 and its very well behaved - its a v low noise camera already Hope thats helps thanks Simon |
4.26
#...
·
·
1
like
|
|---|
|
I stick with the old protocols of darks, flats, and bias and haven’t seen a reason to change. Also, the camera gain goes to 300. I have good results at that level, too. Some dim targets benefit nicely. I try to use 100 gain most of the time, but will go to 300 for filters like the Triad Ultra, which lets so little of the spectrum through. With enough subs (and dithering), there is little difference in the quality of final images at any gain level on the 2600. we have CMOS sensors to thank for such versatility. the point is: if you aren’t a scientist and just want to make beautiful Astro-art, use all available settings any way you want to get the results you want. You will be rewarded. |
1.43
#...
·
|
|---|
| imx571 is a pretty clean and linear response sensor. If your flats are not too long (ie they don't accumulate too much thermal signal) you could calibrate them with bias. I do this with my qhy68m (same sensor) |
|
1.91
#...
·
|
|---|
| Poll created... Still much confusion, I get different answers.. |
7.83
#...
·
·
4
likes
|
|---|
You get different answers because this is one of those topics that's eternally debated. You've got staunch supporters on both sides of the argument. Coke or Pepsi? Tastes great or less filling? Dark flats or biases? These things may never be answered  |
|
1.43
#...
·
·
3
likes
|
|---|
Jonny Bravo: agreed @Daniel Renner try both and choose your best |
0.90
#...
·
·
1
like
|
|---|
|
I woult NOT go below 100 in gain, neither for flats nor for lights. The ASI294MC has a simular characteristic and operates at a different mode above a certain gain. It seemed to me like a good idea to take flats at low gain to increase exposure time. The result was severe, grainy walking- noise which was not averaged out by taking a hundred flats. The problem is not neccesarily the readout noise as such, but that the fixed noise pattern is different. |
|
11.99
#...
·
·
2
likes
|
|---|
Jonny Bravo: What Johnny said  Also, I don't understand the point of polls. Everyone will have a different opinion. Should you weight their opinions equally? Better by far to do a search on those using the 2600 MC or MM (easily done on AB), find someone whose images you like and you feel like they know what they are doing, and ask through a PM. This is what I did when I wanted to learn how to collimate my Newt. There are a dozen different ways, a poll will give you a dozen different opinions, each equally convinced they are right. |
2.11
#...
·
|
|---|
| with my 2600mm dark , flat, dark flat , not bias , i have tested it with bias , and the same |
8.52
#...
·
|
|---|
Miguel G.:Jonny Bravo: This. Ive personally had success with bias and dark flats. Pick one and if it works roll with it. |
