My Patreon: www.patreon.com/cuivlazygeek Here is an EXCELLENT explainer by Dale Ghent, one of the NINA developers: daleghent.com/2020/08/understanding-camera-offset My videos on gain, etc: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-GKXkM_FbFgg.html and ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-SYQ1i4k62eI.html Amazon affiliate: amzn.to/3hTB5Ne Agena affiliate: bit.ly/3Om0hNG High Point Scientific affiliate: bit.ly/3lReu8R All-Star Telescope affiliate: bit.ly/3SCgVbV Dwarf 2 Smart Telescope: bit.ly/3SyChXu Seestar S50: tinyurl.com/3n62hpzx
Two problems with making a video on 25th December are the fact that for at least 364 days of the year, people will be watching on the wrong day for you to greet them as you did. And of course countless millions of people don't celebrate the day. Nonetheless, thank-you for your thoughts and for the video. 🙂👍
That histogram would be more useful if it could be switched to a logarithmic wavelength display. That would spread the low end for you, if the software has that facility,
Great video Cuiv. When I started this hobby back in 2021, I searched for answers to the mysterious offset, it seemed no one wanted to talk about it or if they did it was "yeah just set it at 50". I finally figured it out on my own, after much trial and error, this video will help many newcomers to the hobby.
The issue is not explaining it, as that is relatively easy. It is just most novices have far better things to worry about messing with. That and devices like the ASIAIR do not let you adjust it.
Cuiv, most logical and rational method I have seen for setting offset. Nice one mate. One other small point is that yet again NINA gives us astrophotographers all the tools we need.
What I have learned is that you should always make sure your histogram is as far away from the left side of the histogram as possible to prevent clipping since what we image is so faint. Because this is a linear RAW image that will be stretched later, don't worry about the peak being too far to the right as long as it does not exceed the 1/3 of the histogram as we can fix this with the histogram stretch in post processing. Already messed up and imaged at zero offset? Don't throw those frames away, you can add a pedestal when you integrate. Yes you lost data but you can still process your image. And yes whatever you change whether it be the gain, offset, temperature, and even the USB limit (yes that can cause issues too like banding!) take new calibration frames. Thanks for making this important video Cuiv and Happy Holidays!
As a fellow geek, I like how you deep dive and dig into these details that most others overlook without giving it a second thought. You are one of the few youtubers finding truly original tips!
This video is a good opportunity to again say thank you for your commitment to astrophotography and to your viewers. Your explanations make sometimes complicated topics understandable, and your sense of humor makes your channel a welcome place to be. Merry Christmas Cuiv.
I was absolutely setting my offset wrong. Thank you very much for explaining how it actually works! Now to retake all my master-darks in preparation for the next clear night....
I did mine last year in SharpCap and ended up with a value of 6 on my ASI585, but the offset only ranges from 0-200 on this camera. I thought I'd give it a shot with this method just now and I think it was pretty close, but I'll probably bump it up by a whole digit to 7. I found that the numbers change a lot if I loop my images in NINA though; for me at 1*e-5s it goes like: 0(x1), 40(x2) , (0(x1), (160(x1), (128(x3), (0(x1)... And at 7 everything is just a little bit higher, I still hit a few 0s, but less frequently. I found the SharpCap method to be a bit "hacky", and I was never quite sure if it was correct, but I think I like this NINA method a bit better because you get to see the real numbers, plus it's additional confirmation that I was in the right ballpark. I think the default in the ZWO driver for my camera was originally 25, which was far too high!
Thank you Cuiv, it's good to have this valuable nugget of information. It's been ages now watching rain drops UK running down my windows. clear skies are rare as rocking horse poo💩💩.. So applying tips etc has now become more interesting than taking the light frames!!! 😂😂 Happy day & new year, thank you for all you do.. Atb Tom dull wet UK....
Finally someone answers my question on Offset. I have seen it on NINA, in the sequencer, and never knew where to set it, or what to set it too. I have the OGMA AP26CC camera now, which I think is the same as the Topek ?? that you have, with the Sony 571 chip. I love the camera, but this will really help me get the maximum performance from it. Thank you very much for making this video. I really love NINA as well, but wish I knew how to use it better. Thanks for this, and many of your other videos...all very helpful. Mark
Very helpful, now at least I have some method behind my setting. I just read somewhere it should be set at 70 and I had no idea why. Going to try this method before next session. Thank you!
Merry Christmas, Cuiv, and thanks for another great video. Interesting to see how different manufacturers using the same sensors, apply their firmware/electronics differently. In your Touptec camera, apparently 1 offset equals 1 ADU. For QHY this is 16 ADU, and for ZWO this is 40 ADU. So correct offset values between cameras differ a lot. Also, the Touptec apparently can report a ‘0’ value, whereas QHY and ZWO cameras report ‘1’ as the minimum value, or ‘4’ if it is a 14-bit camera.
Merry Christmas Cuiv! Back in the long, long ago there was an article from Dr. Craig Stark (original author of PHD Guiding) that described the process for CCD cameras. When I got into CMOS cameras I read Dr. Q's article (QHYCCD) that described the same process you go over here. I basically read his article and followed the same process you highlight here. It does help to optimize your dynamic range which helps minimize frustration in imaging from Bortle 9 zones like I am in near Dallas, Texas, USA. One of these days I need to take my daughter to Tokyo so she can meet her Japanese relatives.
Another great video, thanks! I've been taking short exposures that are long enough to swamp the read noise and usually used a 500 ADU offset but lost the offset setting. I collected many nights of images at 0 offset and had clipped badly which really impacted the processed image quality. It is important to set the offset high enough that there is minimal clipping.
Merry Christmas. A real present for us. I’ve watched a lot of videos and read a lot of literature and this has not come up before. Thanks for this, looks like an important setting usually overlooked.
Very useful video. Thank you. I must admit I've always gone with the "ask someone else" method of getting the gain value for my camera, so this will allow me to correct that mistake!
I breathed a sigh of relief after checking some old bias files. With a gain of 120 and an offset of 50 on my ASI294MC pro, it appears that I've been doing it right, more or less. Thanks for the clear explaination of what offset is all about.
@@CuivTheLazyGeek I should mention that I don't take biases everytime I shoot. In fact, I just subtract off the bias value when I do my stacking in SIRL--supposedly less noise added to the result.
@@richardshagam8608 having a bias library you keep reusing is fine, I do the same :) But what do you mean by "you just subtract off the bias value when you do your stacking"?
I remember this one time when I was trying to apply my calibration frames to my light frames with different offset values. I kept on getting error warnings from Siril , Astap or any other preprocessing software. I could not understand why until I figured that I was working with different offset values. Lesson learned after many hours of frustration. I wish I had my hands on this video back a few years from now! lol
Happy Christmas Cuiv. Good explanation of this setting. I have been leaving the offset on my mono and OSC cameras at default settings as the images seem to be ok. Will try adjusting the offset the next time I get a clear night ( have only had three in the last two months )
Nice video, and well done. Yes, everything in the video is correct. I've given the same advice to people before when they asked how to adjust the camera offset. But this is the first video I've seen on the topic (at least that I've seen). Thanks Cuiv. This video should prove quite useful to a lot of people.
Merry Christmas to you and yours, and thank you for being such a wonderfully enthusiastic lazy geek. May clear skies follow you to wherever you may roam.
Dude... this video, and your channel as a whole, is SO informative and straightforward. I sincerely appreciate your demonstration of each step you take and what it all means and why you do different things. One question I have... you mentioned that you don't use dark frames any more. Can I ask why? I am new to astrophotography, but I thought dark frames were essential to getting the best images?
Thanks a lot for this explanation! For my Touptek-based Omegon VeTec 751C and based on your video I found an interesting relationship between the gain and offset. No matter which gain I use, taking the same value for the offset - 8% works great (gain 100, offset 92; gain 316, offset 291; gain 2500, offset 2300). Though, I know that i.e. the gain of the ZWO cameras is calculated differenty, so this will probably not work for any camera
Merry Christmas. An excellent video as offset seems to be quite a mysterious topic out on the net and on the forums with a whole range of settings being offered without any real explanation as to why that number was settled on. Your video gives a nice explanation of how to calculate it with meaning. Thanks again and wishing you a great 2024 (hope we see some more paragliding videos too).
Well Cuiv, if there was no way to prove you were the Lazy geek before or even convince some of us, this has done it! I did wonder why my astronomy photos didn’t have any stars.
I love the work you do and your contribution to AP. I have a RASA11 and I am worried about doing a "tilt correction". Please consider a review of this factor and how to do the procedure.
Thanks Steve! Tilt is one of those things I have never had success with - and imo unless someone is using full frame and a Tak, it may not be worth the effort... Right now I'm definitely not qualified to make such a video unfortunately!
Super informative video Quiv, very understandable explained, thank you for all those great videos you made during the past years. Wishing you and your family happy holidays
Cuiv, Interesting Video. However, I have an issue. The video seems to be a some what comparison between the ZWO-ASI2600mm and the Touptek/RisingCam 2600mono camera as they both use the following sensor: SONY IMX571 CMOS. Accordingly, they supplied information on the specifications of the camera, for the vendor to select functions that they wish to pass on to the consumer. It appears that ZWO seem to have gone out of their way to make it easier for the consumer in calculating the optimum gain and offset, known as 'Unity Gain'. Also allowing the user to have the ability to select a manual option and set values of their choosing. All good so far. So here is my issue. You have a Touptek camera and set your 'Gain' to 100 (101),'Offset' to be 151. Using the short Bias exposure to calculate values for the 'Min (0)' and 'Max (65536)' Pixel values. I don't see a problem with this, other than the values you have defined as correct for your camera. It appears to me that you are using the ZWO 'Unity Gain 100' as the starting point gain for the Touptek camera. I know the Touptek Min gain is 100, which is just a coincidence compared to the ZWO 100 optimum gain. So my point is this. Taking that ZWO has calculated the 'Unity Gain' correctly, and we can probably assume they have, and we need a comparison of the two, as Touptek, as I know, does not supply a unity gain value, then the following formula would need to be applied. The ZWO has a gain value of 0 to 450. The Touptek has a gain value of 100 to 10,000. Note: Values under 100 cannot be selected. So, if one needs to take a direct comparison between the two cameras to find the gain for the Touptek, the following calculation would need to apply. Touptek's max gain is 10,000. We would divide this with ZWO's max gain of 450. This would give a value of 22.2222. We would multiply 22.222 by ZWO's Unity Gain of 100, and would give a value of 2222. This value would be equivalent to the ZWO Unity Gain of 100. So the unity gain for the Touptek would be 2222. Formula: 10000/450=22.2222 22.2222x100=2222 Touptek max Offset is 7936. We would divide this with ZWO's max Offset 240. This would give a value of 33.06. We would multiply 33.06 by ZWO's Unity Offset of 75, and would give a value of 2479. This value would be equivalent to the ZWO Unity Offset of 75. So the unity offset for the Touptek would be 2479. Formula: 7936/240=33.06 33.06x75=2479 UNITY GAIN ZWO GAIN 0 100 450 Touptek GAIN 100 2222 10000 UNITY OFFSET ZWO Offset 0 75 240 Touptek 0 2479 7936 I have ignored LCG, HCG etc. as we are only finding a comparison unity between the ZWO and Touptek. I have always used the above gain at 120s for my Touptek/RisingCam camera as I found 100 did not work too well even with the above activated. I am not all that concerned about a few black pixels or over bright ones, as this depends on the target anyway. I am in a high bortle zone 7 and getting worse. So what is your opinion? Is it that we have been used to values of ZWO and think these values also apply to all other cameras, such as Touptek. ------------
Thanks so much, Cuiv. Great explanation. And very much appreciate all of the information you have given us - very helpful to new (and somewhat lazy!) folks like me. Happy holidays!
Good post! It would be interesting to statistically compare images of a bright source, a glowing nebula, with offsets of double or triple the 151 and see if there is any deterioration of image quality. Dynamic range is important, but I don't see how sacrificing even up to 1% of dynamic range would make any difference in the final image quality. If you are clipping the brightest stars, it shouldn't matter since most people end up reducing the number of stars and their brightness in the final image.
This was a very helpful video - thanks! There was one aspect that initially confused me, however. Clearly simply adding a constant uniformly to an entire set of data cannot increase the amount of information; it simply resets the zero point. I'm guessing then that what is really happening is that the camera performs some kind of internal mapping from the raw data to the stored data, and the offset is applied to the _raw_ data before that mapping, so that pixels with low counts don't get mapped to zero, i.e. clipped. Applying it to the stored data would not help in that way.
Great video, thanks! I'm trying to get this right on my Altair 269C TEC, the range is a lot smaller 0-462 and NINA set the default to 16. Setting it higher, the Mean and Median go far too off and the difference is more than 1000, so it might have to do that it is a dual mode camera, or just a driver thing. Maybe I'll give SharpCap calibration a test, to see what values it shows. Love your channel and I can see how Japan has formed your kindness! ありがとう ございます
Hey Cuiv, first of all Merry Christmas!!🎄 and thank you for all your great contents and advice. Perfect timing, just last night I was searching for this topic and then you came to the rescue 😀.
Thanks, Cuiv now I know how to check the offset. My ZWO ASI2600MC Duo default offset is also 50 but this gives the histogram Mean and Median values near 500. For any offset the Mean and Median histogram values near 10X the offset. UPDATE Looking at the fits header, NINA is added 50 as the offset not the 500 being added my ZWO.
Merry Christmas! Yes, a lot of people afraid to use a high offset, somehow the loss of dynamic range scares people I guess, but seeing images from people their loss of DR is more contributed to incorrect processing and I'm guilty of that too. I would say on average; without careful image editing; people ruin their DR in their image by cutting as much as many thousands of ADU steps, compared to a few hundreds that a higher offset would do. The loss of DR with the offset just irrelevant and minuscule, compared to wrong image editing habits. I just follow how ZWO set it in Asiair, (offset 75 for zwo cameras) and seeing how that's like almost halfway up on the offset scale, I imitate it to do similar with my non-zwo cameras and I am perfectly fine losing a few hundred or even a 1000 DR steps due to high offset. Otherwise Pixinsight will report me that it had to do a "huge pedestal" correction when processing and that lets me know that I did use the wrong offset.
Merry Christmas. I just bought the same camera you have and was in doubt as to how the gain and offset values should be set. Thank you for the clarification 😊🎅
Merry Christmas ‘tis like a present under the astrophotography tree! Now that I’m starting to use MeleQ in addition, this is quite relevant. Re. ASIAIR, I heard ZWO ASIAIR set an offset of 45 or so to the 2600 at 101 gain…but I need to look again on Cloudy Night! I will read through the comments here too.🎉❤
Yep it seems the ASIAir hides the offset and sets it to a suitably high value to make it easier to the end user! And their offset units are different from what I heard
Interesting. Would like to see what the master lights look like at various offsets and how one setting might produce a better end result over the others. In PixInsight, you can set a minimum pedestal value during WBPP, which I'm not sure yields the same outcome.
Finding the correct offset was always a thing for old-school QHY CCD cameras before finding the correct gain - something lost in the mists of time I guess. From my understanding set it once and never touch it again - I check offset/gain values every couple of years and they haven't moved.
I followed this method on my ASI6200MC and determined the optimal offset is 15 at gain 0. My minimum pixel value on bias frames is 9 at those settings.
Thanks for pointing this out. Was always eager to fiddle with the offset. However when counting pixels with the minimum ADU I'm thinking now about dead pixels of the sensors falsifying statistics and so even higher offsets might be feasible 🤔 but I don't know at the end the offset just makes a small difference with regard to whole dynamic range I guess
Aha… now the offset might be why my images looks pixelated when doing background extraction in Pixinsight. As some pixels turn absolutely black. 😮 Will experiment with the offset and hopefully get some clear whether to test in 2024.
I was told by experienced astro photographer that it needs to stay ( gain ) at factory level which is 48, so now you presenting the new ( it isn't new, is it ) idea about offset. I'll get it tested for sure, thanks.
I forgot to ad that my camera is Altair 269C, I will test it anyway, I love your work and getting your tips and tricks implemented into my circumstances. @@CuivTheLazyGeek
Hi there Geek. I recently bought the Altair Astro version (AA26C) and use Indi/Ekos + altair driver. My understanding is that when taking short exposure darks (BIAS frame) the histogram should not get cut off on the left. I think I saw this on a Sharpcap forum. In Indi oontrols my camera settings are Conversion Gain = High. High Full Well Mode = off. Low Noise Mode = off. Gain=100, Offset=50, exp=.001. The histogram using indi looks fine. Peak at 50, drops to zero below 25 and above 75. In NINA I get the Ultra Mode option same as you which up to now I had set to off. 12000 clipped pixels is microscopic compared to 24Million pixel on the sensor (5 parts in ten thousand). Not sure this makes that much difference for this camera, but great that you explain how to measure things. Should I be using Ultra Mode (NINA) / Low Noise Mode (indi) ? Confusing camera.
Great information. I was wondering when you were going to mention that increasing your offset was going to decrease the dynamic range. ;) I think it's probably best not to decrease your DR so you have more room in PP.
Yes you did mention it. While watching the video and mentioning that the darks would move to the right, I knew right away the DR would decrease. So, I was saying to myself that you would be mentioning a decrease in DR and you did.@@CuivTheLazyGeek
@CuivTheLazyGeek Thanks for all your insights, I appreciate them all as an Astro noob. I don't know how you do it? I mean keeping your wife😃 I lost mine a few years back in another rabbit hole called Euro Rack😂 At least now the only one I can lose in the black rabbit hole is myself😅 Merry Christmas and Happy New Year🎄🌌🌠🥂
Thanks a lot for the video and Merry Christmas! Only to say that minimum exposure time should be 0.0001s I think. I have the OEM version OMEGON veTEC571c and it's 0.0001s
in addition...I did just the test with a veTEC533c -> the Min values are not very consistent. So would recommend to repeat shooting several times until to take decision what Offset really should be taken
