DSLR影天文相的ISO
DSLR影天文相的ISO
http://www.astropix.com/GADC/INTRO.HTM
Aloha介紹人睇Jerry Lodriguss 的A Guide to Astrophotography with DSLR Cameras, 內有提到,base on Canon 20D,ISO應該咁樣set。
If above 60F (15.5C) then use ISO 1600
If between 40F to 60F (4.4C to 15.5C) then use ISO 800
If below 40F (4.4C) then use ISO 400
Aloha話係Jerry實戰經驗,未有理論
Aloha介紹人睇Jerry Lodriguss 的A Guide to Astrophotography with DSLR Cameras, 內有提到,base on Canon 20D,ISO應該咁樣set。
If above 60F (15.5C) then use ISO 1600
If between 40F to 60F (4.4C to 15.5C) then use ISO 800
If below 40F (4.4C) then use ISO 400
Aloha話係Jerry實戰經驗,未有理論
I am not 高人, but it seems a physics problem. My view may be wrong… I guess increasing the ISO will result in increasing the thermal noise. However exposure time is shortened. As the noise accumulated depends on time, which means lesser noise, the unwanted spots, will be appeared on the image. In addition, double the ISO will half the exposure time. If what Jerry said is correct, the accumulated noise is a non-linear increasing function with a slower growth rate. For example SQRT or LN functions etc. As a result, the higher ISO noise can be compensated by the shorter exposure time.
我不是高人, 但Lordriguss的說法有點以毒攻毒的味道。XR250 寫:我記得Aloha講過read-out noise及dark current都會受ISO及溫度改變,上面係Jerry Lodriguss實戰經驗,未有理論。也可能係20d cmos的特性?? I don't know. 有無高人解釋?
可看一看這一篇攝影雜誌的文章, http://www.phototechmag.com/sample/sharma.pdf, 溫度高, 黑電流/黑雜訊(dark current/dark signal)會多, 而提高iso, 放大器雜訊(amplify noise)會增加, 那同時增加兩種雜訊, 就會有好照片?? 小弟愚昧, 百思不得其解。
業餘天文高人John Smith的說法也指出感光CCD一般是愈冷愈好(見http://www.hiddenloft.com/notes/CCDtemp.pdf), 而貴價的天文專用CCD會跟冷卻器而不是加熱器。
Find relevant info.:
http://www.ece.uah.edu/courses/ee420-500/603ch10.pdf
Equation (10-17) , the RMS value of noise voltage is a slowly increasing function of temp. T (i.e. square root function).
http://www.ece.uah.edu/courses/ee420-500/603ch10.pdf
Equation (10-17) , the RMS value of noise voltage is a slowly increasing function of temp. T (i.e. square root function).
Thanks! Quick response! BTW, are u focus on theoretical physics? u got good math sense.sptfung 寫:I agree with willis.
I also think that the main reason is time.
I have taken a course Observational Astronomy and there is some topics on calculating the three different types of noise.
I have to find out the notes.
Yes, my project is on theoretical astrophysics.willis 寫:Thanks! Quick response! BTW, are u focus on theoretical physics? u got good math sense.sptfung 寫:I agree with willis.
I also think that the main reason is time.
I have taken a course Observational Astronomy and there is some topics on calculating the three different types of noise.
I have to find out the notes.
I like theoretical physics too.
I guess ur supervisor is either Prof KS Cheng or Dr. HF Chau. Keep on ur research work!sptfung 寫:Yes, my project is on theoretical astrophysics.willis 寫:Thanks! Quick response! BTW, are u focus on theoretical physics? u got good math sense.sptfung 寫:I agree with willis.
I also think that the main reason is time.
I have taken a course Observational Astronomy and there is some topics on calculating the three different types of noise.
I have to find out the notes.
I like theoretical physics too.
Haha you made a good guess!willis 寫:I guess ur supervisor is either Prof KS Cheng or Dr. HF Chau. Keep on ur research work!sptfung 寫:Yes, my project is on theoretical astrophysics.willis 寫:Thanks! Quick response! BTW, are u focus on theoretical physics? u got good math sense.sptfung 寫:I agree with willis.
I also think that the main reason is time.
I have taken a course Observational Astronomy and there is some topics on calculating the three different types of noise.
I have to find out the notes.
I like theoretical physics too.
Thank you
我也不是高人.
Thermal noise 是 direct proportion to 時間同温度, 温度上升大約7C, thermal noise會double.
s/n ratio主要睇總曝光時間, 諗住用高d ISO去增加s/n其實是沒有大意義的.
早幾個星期前度過部40D的system gain同read out noise:
於ISO400/800時, gain分別是0.87e-/ADU及0.43e-/ADU, readout noise是8ADU及13ADU (14 bit A/D conversion).
要注意, ISO800時的saturation只得大約7000e-, D-range會受影響.
以拍星野或深空天體而論, 假如追蹤準確的話, 一張過長時間曝光的raw嘅s/n係會比多張短時間曝光疊加的高.
其實用咩ISO, 主要睇D-range同天體亮度的配合.
遲吓有時間再share吓其他數據.
Thermal noise 是 direct proportion to 時間同温度, 温度上升大約7C, thermal noise會double.
s/n ratio主要睇總曝光時間, 諗住用高d ISO去增加s/n其實是沒有大意義的.
早幾個星期前度過部40D的system gain同read out noise:
於ISO400/800時, gain分別是0.87e-/ADU及0.43e-/ADU, readout noise是8ADU及13ADU (14 bit A/D conversion).
要注意, ISO800時的saturation只得大約7000e-, D-range會受影響.
以拍星野或深空天體而論, 假如追蹤準確的話, 一張過長時間曝光的raw嘅s/n係會比多張短時間曝光疊加的高.
其實用咩ISO, 主要睇D-range同天體亮度的配合.
遲吓有時間再share吓其他數據.
If temp increases 7oC, the noise is double, then temp increases 14oC, the noise will be x4. The conclusion is that the noise is not proportion to temp. In addition, the unit should be measured in Kelvin. Correct me if I miss read ur meaning..Subaru 寫:我也不是高人.
Thermal noise 是 direct proportion to 時間同温度, 温度上升大約7C, thermal noise會double.
s/n ratio主要睇總曝光時間, 諗住用高d ISO去增加s/n其實是沒有大意義的.
早幾個星期前度過部40D的system gain同read out noise:
於ISO400/800時, gain分別是0.87e-/ADU及0.43e-/ADU, readout noise是8ADU及13ADU (14 bit A/D conversion).
要注意, ISO800時的saturation只得大約7000e-, D-range會受影響.
以拍星野或深空天體而論, 假如追蹤準確的話, 一張過長時間曝光的raw嘅s/n係會比多張短時間曝光疊加的高.
其實用咩ISO, 主要睇D-range同天體亮度的配合.
遲吓有時間再share吓其他數據.
誰在線上
正在瀏覽這個版面的使用者:沒有註冊會員 和 19 位訪客