I have an ST7-XME that I've been using faithfully for over 10 years. I probably have at least 500,000 images taken with it. But, two days ago I decided to run the DriverChecker and updated my drivers from version 4.85 to 4.90 Build 1. Now I'm having issues with my bias frame values. They've jumped from around 125 to 2000. So I'm not sure if something happened to the camera or if the new drivers did something to the camera. I read the post "Huge Bias" and was wondering if doing the USB Offset process is what I need to do. I'm a little nervous as this has been a rock steady camera for years. Any suggestions? Vance
Thanks, Bill. I will run the USB Offset process. Could you shed a little light on what and why this happened all of a sudden? Do the new drivers interpret the EEPROM settings differently now? Does this have any bearing on the dynamic range of the chip going forward? Thanks, Vance
I agree with Bill. I'll also note that earlier generations of the Universal Driver defaulted to have an "Automatic Bias Level Correction" feature turned on. This reduced the actual camera offset (1000 to 2000 ADU) down to about ~ 100 by subtracting inside the driver. I've never thought that was a good idea, as it does cause problems in a variety of circumstances. I told our programmers to default it OFF in newer releases of the SBIGDriverChecker. That may be why it appeared to "shift" suddenly. You can turn it back on in SBIG Driver Checker, click Config Drv. But I actually recommend that you don't. Doug
Thanks for the response, Doug. I did try turning the "Automatic Bias Level Correction" option on and off in my testing and it made no difference to the bias counts. I believe my maximum pixel count before was around 63,000 or so . Does that mean it will now be shifted up to the 65,000 mark now that the bias level is being reported correctly? Thanks, Vance
Hard to know without having the camera in hand to run a test. It doesn't matter if you hit the 65,000 mark, or if the offset is a bit high. This will have no impact whatsoever on the camera performance. What matters is that you sample the full range of the sensor output with sufficient resolution to sample the noise. The KAF-0402 has a full well capacity of 100,000 electrons. But it also has a read noise in the 15 e- range. So crudely speaking the dynamic range is 6,600:1; that's about 13 bits worth. Add a couple of bits to sample the noise adequately and you have 15 bits. Throw in an extra one for good measure, and you have 16 bits. So there's lots of headroom available in these cameras.
