Registered: 1496756664 Posts: 2
Reply with quote #1
We are currently trying to implement ML in our lab. At first we purchased a single NI PCI-6221 DAQ for use, but since it doesn't provide us with the required temporal accuracy, we purchased another one. After installing the 2nd DAQ, verifying its' recognized by ML, and enabling A-I, the lowest latency Eyejoytrack would return (using 'AcquireFix' followed by an eventmarker) is ~46ms after toggling the fixation point on, even when the eye position was already inside the fixation target when it is turned on. This is similar to the temporal resolution we measured with a single DAQ. Does anyone know of such an issue, and what can be done to solve it? Thanks in advance!
Registered: 1435685587 Posts: 627
Reply with quote #2
First of all, the temporal accuracy that you are talking about is not the kind of accuracy you can improve by adding another DAQ board.
Second, which version of ML are you testing? Try NIMH ML2. The interval you are measuring is the weakest part of the original ML. I made a lot of efforts to shorten it in NIMH ML2. Third, why are you measuring this interval? If you are doing this to switch stimuli very fast for receptive field mapping or reverse correlation, you had better program with the new runtime of NIMH ML2. Since it is not documented well yet, I can help you, if you can explain to me what you are trying to do. By the way, NIMH ML does not require two DAQ boards to achieve good temporal resolution.