search

goori / braintrust

Automatically exported from code.google.com/p/braintrust
0 stars 0 forks source link

EEG viewer defaults to nonintuitive waveform rendering #438

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
The default of "no filter" is doubly NONintuitive: (1) It really does filter 
with screen "antialias," and (2) it results in radically different signals than 
expected. No known EEG viewer beautifies artifact to the point that it shows no 
artifact.

For example, 1st page of 60Hz-infested Study005: In this subject difference in 
Portal between default antialiased view & lin interp (the "real" signal) is 
HUGE.

Default should really be the line plotter, the one currently with semi-misnomer 
of "linear interpolation."

Original issue reported on code.google.com by echa...@gmail.com on 19 Aug 2013 at 2:44

GoogleCodeExporter commented 9 years ago 
Well, it's perhaps more complicated than is evident.

The issue that seems to be confusing people is that we downsample (decimate) to 
exactly screen resolution, and as a result, by default we get the usual filters 
required to anti-alias at that resolution.  This is, as I understand it, the 
"correct" thing to do.

I have changed the latest version to do 4x screen resolution, to see if that's 
more in line with what everyone expects.  This will plot 4 points for every 
screen pixel, and will filter accordingly.  Note that it also means the render 
loop takes 4x as long, and the server sends 4x as much data...

Original comment by zack.i...@gmail.com on 1 Sep 2013 at 10:04

GoogleCodeExporter commented 9 years ago 
Further modifications:

* Filters have been set to now ignore the screen resolution
* There are more filter options

I believe the behavior should now match what most people expect.

Original comment by zack.i...@gmail.com on 8 Oct 2013 at 11:58

GoogleCodeExporter commented 9 years ago

Original comment by john.fro...@gmail.com on 9 Oct 2013 at 7:43

GoogleCodeExporter commented 9 years ago 
[deleted comment]
GoogleCodeExporter commented 9 years ago 
Users of XLTek Neuroworks 7.1 are happier now with the new decimation and 
screen antialiasing filters. They had report the same complaints I submitted 
here & their soln. is along the lines of Zack's newest 4x sampling, except that 
in the "crushed" pixels (where there are more waveform points than available 
pixels) they display *the maximum* of the bunch. This leads to the satisfactory 
"true" waveform illusion.

Zack's fix also looked good enough on screen with 60Hz noise...

Original comment by echa...@gmail.com on 10 Oct 2013 at 1:47

GoogleCodeExporter commented 9 years ago 
Issue 413 has been merged into this issue.

Original comment by zack.i...@gmail.com on 11 Oct 2013 at 5:02