I’m encountering an issue where I see artifacts being drawn on a plot with a large amount of data (on the order of 10^6 of points). The artifacts are simply lines being drawn from one point to another, on top of the usual plot lines that I would expect to be drawn. These lines tend to flicker during a pan, and I think they are being caused by resampling of the data during a pan. The only resampling modes that this does not occur with are Max and None.
In general, what is the best resampling mode to avoid artifacts and obtain good performance? The data is essentially of a time-series nature (numeric value on Y axis, date on X axis), with fairly consistent spacing between points. I tried Nyquist resampling mode, and it seemed identical to MinMax.
Any insight would be appreciated!
At the moment, the resampling mode MinMax has the limitation that it expects X-data to be a.) sorted in the X-Direction and b.) evenly spaced in the time-domain. This is a known limitation and a high priority task for us, so it will be addressed soon.
For now please use Max or None as you’ve mentioned. In our internal build we have developed a robust algorithm to replace MinMax which is both fast and does not require evenly spaced data, however it is yet to be released.
In SciChart v3.1 resampling algorithms were significantly improved. Unsorted data is allowed on DataSeries as well (UnsortedDataSeries is no longer required), but not desirable because it makes a chart significantly slower.
Please login first to submit.