Is there any way to improve cursor resolution?

[JayZ]

Thanks, Ben.

I will stop thinking about this.

[chilly_charly]

Aahem...
Sorry Ben, but I think it's really a bug/flaw. As long as a point is clearly distinct from the others, it should be possible to select it. You are considering only the X coordinate, and forget about the Y coord.

As demonstrated in the attached modified vi, the detection of the max point in a peak works nicely, whatever the scale size and the number of points.

The problem appears to be specific to waveform graphs, and does not occurs with XY graphs.

I have made some further experiments with waveforms : When the curve is duplicated, the snap-to-point cursor property does not confuse the two curves. Means that the find-nearest-point function works properly. Surprizingly, it does not work for points belonging to the same curve.

I can't find any reason why the cursor snap-to-point should not take the Y coordinate into account in a waveform graph. I think we may call that a bug.

[Ben]

Hi CC,

You may be right!

I will let you post this to the Bug Thread.

Ben

[Ben]

Hi CC,

Thank you for posting this as a possible bug. Bugs (as you know) can be difficult to identify without a detailed understanding of how a function is supposed to perform. I am very curious in the oficial response to this issue.

Trying to learn,

Ben

[JLS]

Hello everyone,

Actually, I think you can use the mouse to make the cursor be on each peak, but I understand how the behavior is at first a bit misleading.  I am guessing by the descriptions I read that everyone is moving the cursor horizontally, and watching it "bounce" around the plot it is snapped to.  Indeed, using this technique I cannot get the cursor to jump to all the peaks.  However, what you can confirm is that for each vertical line of pixes in "dense" plots, there can be more than one data point represented.  In the example CC posted, this is indeed the case.  However, you can navigate to a peak by grabbing the cursor at the actual "point" it is currently snapped to (ie. at the intersection of the horizontal and vertical lines defining the cursor), and move it vertically.  For reference, let's call that point on the cursor the SPECIAL point.  You should then see that the cursor jumps between points that are defined for the plot, but due to compression, are represented within the same vertical line of pixels.  In fact you can drag that point all over the plot to snap to arbitrary points.  I have attached a screenshot and a trivially modified version of catch_peak 2.vi which I used to play with this.

I think what LabVIEW does, and in all fairness must do, is for a particular pixel relative to the horizontal, pick one of the plot's values which is represented in the corresponding vertical "line" of pixels.  Then, if you want to jump/snap to other points which are represented in that vertical line, you will have to click and hold carefully on the "cursor point" and move it up or down - since it is still snapped to the plot.

If I have missed something subtle in the discussion, don't be shy, and I'll explore this more and take it to R&D if necessary!  Otherwise, I think this behavior is at least reasonable 🙂

Best Regards,

JLS 

Message Edited by JLS on 03-30-2006 04:53 PM

[chilly_charly]

JLS,

thanks for popping in. However, you missed the point,,, 🙂 The catch peak 2 vi was not exposing the problem, it was showing that XY graphs behave correctly !  The problem only exists with waveform graphs...

If you want to face the problem, then look at the original catch peak vi provided above by jiankang.

And don't be afraid to comment further, we are not shy :D:D:D