Demonperformer Posted October 6, 2018 Share Posted October 6, 2018 Outside there is something coming out of the sky that strongly resembles the Atlantic Ocean being poured through a seive, so I thought I would give the 'sensor analysis' option in sharpcap a go on one of my cameras. This resulted in the table attached. I'm sure all the figures are great, but as to helping me decide which gain setting I should use for this camera, I am not a lot wiser. There definitely seems to be something strange (impressive?) happening to the read noise between 60 & 61, but it continues to decline after that so I'm not sure that gets me very far. So, it's time to get educated ... (1) which is my best option, and (2) why? Thanks. 224 sensor analysis.txt Link to comment Share on other sites More sharing options...
vlaiv Posted October 6, 2018 Share Posted October 6, 2018 Depends what you want to do with it Jump between 60 and 61 is indeed interesting, but documented feature - there is switch between "modes" at gain 60: If you plan on using it for planetary capture at 10bit ADC, I would say that you should go with Gain 255 (I'll explain in a minute why). If you plan to use it for longer exposures - sort of lucky DSO imaging (5-10s exposures) - go with either 135, 195, 255 or 315, depending on how important read noise vs full well is. Shorter exposures benefit from lower read noise and don't need that much full well capacity. For planetary imaging where FPS is important, you will capture in 8bit and camera will be internally working in 10bit mode, then you want e/ADU to be 0.25, and gain 255 will give you this value. This is because in conversion from 10bit internal ADC to 8bit data you want to avoid any quantization clipping that can occur. Since 10bit -> 8bit is done by discarding two least significant bits - equivalent of dividing with 4 and truncating to integer value (floor function), you want your "electron" count to be already divisible by 4, this way no truncation will take place. Gain with e/ADU of 0.25 will "multiply" any electron count with 4 (when dividing with reciprocal value - 0.25), so it will definitively be already divisible with 4. Link to comment Share on other sites More sharing options...
Demonperformer Posted October 6, 2018 Author Share Posted October 6, 2018 9 minutes ago, vlaiv said: go with either 135, 195, 255 or 315 Sorry, don't see where you get those figures from the table. Link to comment Share on other sites More sharing options...
vlaiv Posted October 6, 2018 Share Posted October 6, 2018 2 hours ago, Demonperformer said: Sorry, don't see where you get those figures from the table. It's not from the table. ZWO makes their cameras with db Gain system - unit of gain is 1/10 db, so Gain 200 for example is 20db gain over 0 gain. This means that every 60 units of gain, e/ADU multiplies by 2 since 20*logbase10(2) = ~ 6.0206 db. Now unity gain is 135, this means that gain 195 will have e/ADU of 0.5 (two times larger ADUs), while gain 75 will have e/ADU of 2 (two times lower ADUs) - steps are increments of 60. Just to check this, you can use your table: Gain 100 e/ADU: 1.20500746448319 Gain 200 e/ADU: 0.387946002185814 In e/ADU units, by going from Gain 100 to Gain 200 you "amplified" signal by e/ADU of 100 divided by e/ADU of 200 and that is: ~ 3.106 While by going with gain in db units, following applies: 20*log(X) = 10 log(X) = 0.5 X = 3.1623 (Very similar figures differing on second decimal place, but I wonder how sharpcap measures this - it actually needs a sort of flats for this measurement - comparing light levels in ADUs with shot noise to get gain factor). Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.