When I was collecting D7000 Read Noise data from the "Optical Black" columns I encountered a data pattern I had never seen before.
The first 4 columns had a very consistent set of values and standard deviation. (Unknown 1 below)
The next 12 columns had a very consistent different value and standard deviation. (Unknown 2 below)
The next 56 columns had the Read Noise data that I'm accustomed to seeing. (Noise below)
I noticed that on my typical log-log plot the two unknown data sets are above and parallel the Read Noise.
So, I'm looking for theories on what this data is.
If you care to play with the numbers, here they are:
| ISO | Unknown 1 | Unknown 2 | Noise |
| 100 | 1.659 | 1.286 | 0.721 |
| 125 | 2.013 | 1.533 | 0.838 |
| 158 | 2.400 | 1.845 | 0.989 |
| 200 | 2.936 | 2.245 | 1.186 |
| 251 | 3.520 | 2.685 | 1.405 |
| 317 | 4.391 | 3.291 | 1.683 |
| 400 | 5.343 | 3.965 | 2.059 |
| 503 | 6.447 | 4.932 | 2.504 |
| 634 | 7.717 | 6.096 | 3.056 |
| 800 | 9.717 | 7.618 | 3.755 |
| 1007 | 11.134 | 8.942 | 4.402 |
| 1269 | 11.401 | 8.776 | 4.438 |
| 1600 | 11.104 | 8.746 | 4.426 |
| 2015 | 11.184 | 8.679 | 4.387 |
| 2539 | 11.252 | 8.854 | 4.410 |
| 3200 | 11.505 | 8.815 | 4.389 |
| 4031 | 11.229 | 8.871 | 4.401 |
| 5079 | 11.205 | 8.796 | 4.399 |
| 6400 | 11.294 | 8.810 | 4.410 |
| 8063 | 11.347 | 8.925 | 4.454 |
| 10159 | 11.505 | 8.820 | 4.355 |
| 12800 | 11.292 | 8.794 | 4.417 |
| 25600 | 11.286 | 9.113 | 4.369 |
Bill
