I am trying to achieve the best possible accuracy we can for start and end of exposure with an Aluma 47-10 camera. The trigger-in signal doesn't look like it will offer the most accurate timing, because the manual says there may be up to 200 ms delay: "When you click Start to begin an exposure, the camera will flush the CCD one last time and then start the actual exposure, with a possible delay of up to 200 ms." Alternatively, I find that if I poll the camera for status changes, the typical polling time is a bit under 4 ms, which puts a better accuracy on the start time. Is there an even-faster alternative to polling ICameraPtr -> queryStatus() ? Finally, I see that there is also a trigger-out option, which remains active during the exposure. Is the trigger-out signal tightly coupled to the actual exposing time, or is its timing affected by things such as CCD flushes? Would it be more accurate than checking queryStatus()? thanks much for any info, -will
Hi Will, You are correct about the manual, it's not very precisely written when it comes to the Trigger operation. At least not the part that you're interested in. We should probably update it. The way Aluma cameras work right now is that while the camera is waiting for the External Trigger signal, it will keep flushing the sensor row by row. If you press Abort, it will stop and go back to the Idle state. But once the TriggerIn signal comes, it will resume normal operation, which is to do a full flush of the sensor and then start an exposure. We do that so that all frames are consistent. I think this note about a delay up to 200ms is to cover all sensors supported in the Aluma Series. I can't say exactly how long it takes for 47-10. Checking for Status changes is one way to go, but I would highly recommend using TriggerOut signal. Its operation differs depending on whether the sensor is Interline or Full Frame. 47-10 is a full frame sensor, which means that the pixels are collecting light all the time (that's why it has to be fully flushed before exposure) and the exposure to the light image starts when the shutter starts opening (from side to side). Immediately after actuating the shutter, the camera will set TriggerOut signal. TriggerOut will stay on until the exposure ends, at which point the shutter will start to close, starting from the same side, ensuring that all pixels are illuminated for the same period of time. One thing you need to keep in mind is that the full shutter operation (opening or closing) takes approx 100ms. So while all pixels will be exposed to light for the same amount of time, the pixels on the left will be exposed slightly earlier than the ones on the right. It sounds complicated but it's basically a rolling shutter that moves from side to side. To sum up, TriggerOut will be set when the first pixels on the left are starting to be exposed to the light (shutter starts opening), and it will be cleared when the same pixels are starting to be covered from light (shutter starts to close). The pixels on the other side of the sensor will still be exposed to the light until the shutter stops, which would be approx 100ms later. Hope that explanation was clear enough and solves your problem with precise timing. Kris
Hi Kris, Thanks for that detailed explanation. Our application involves synchronizing 4 separate detectors across several spectrographs, and we are using external shutters that are synchronized to about 1ms. We don't need the internal shutter from any of the cameras (and our engineers will remove the internal shutter from the Aluma) but we do seek the best-possible timing on knowing when each exposure begins and ends. There are two "science cameras" which will do an ordinary integration for the full exposure time, while the Aluma will be read out every ~1s (actual time depends on the object, of course) and these images are used for two purposes: (a) as an exposure meter, to determine when to end the exposure; and (b) to let us make a "photon-weighted" exposure midpoint estimate. Of course, these images will suffer some smearing since the shutter won't close during this time, but that's ok. Since TriggerOut is associated with the shutter, it sounds like we should prefer polling the Status values. -will
