N-Trig jitter is driving me nuts!

[g3g3n3]

I don't think it's defective. It's a bug for sure, but I have completed some pretty precise work on the surface book. I guess you just have to be a little quicker with your strokes. See attachment

All 4 of my students that have Surface Books don't like it. It depends on you're working style...if you tend to draw fast and undo very often it might be ok. I prefer to draw with a slow controlled line for small details. FYI my students use it for cleanup linework for manga as well.

Also, you're drawing curved lines here. Try drawing straight lines at an angle of 45 degress slowly as you would need for drawing something architectural...you WILL get a wiggly line. It is a defect that is widely known on different forums and apparently has been around for a while. Might never be fixed as it is likely a defect with the N-Trig technology.

Here's another place where people are complaining:
Surface Pro 3 Wavy diagonal lines when inking

Finally, I noticed that you are using Manga Studio which offers a brush stabilization feature(though you might not be using it and might not need it if you draw quickly); basically if delays drawing and smooths the line as you draw. What i want is 1:1 clean input from my hand to the drawing surface at any speed at any angle. N-Trig simply CANNOT do this.

 

[g3g3n3]

I don't think it's defective. It's a bug for sure, but I have completed some pretty precise work on the surface book. I guess you just have to be a little quicker with your strokes. See attachment

Yes, unfortunately it IS defective, or perhaps you could say INFERIOR to other solutions out there...in the sense that this is unlikely able to be fixed with a software upgrade. It's been out for a while now, and there's NO SOFTWARE SOLUTION IN SIGHT. Maybe in future hardware N-Trig will get better, but the currently available tech is simply not able to do it. Can't believe I recommended it to my students. Embarrassing. Luckily 2 returned them.

 

[tunepunk]

After doing some digging into how the N-Trig technology works the there might be some quite simple upgrade to the pen. The technology itself uses an active pen on a passive digitizer, meaning the pen is generating an electromagnetic field that the digitizer picks up.

Here's what I've found on another site.

"N-Trig, an Israeli company just recently bought by Microsoft is Wacom's main contender. Their technology is based on an active pen and a passive digitizer that I use in my Microsoft Surface Pro 3. Advantages are better accuracy between cursor and pen, less parallax, lower production cost and edge accuracy. Disadvantages are a pen that requires a battery, making it to big to be housed in a silo inside today's thin tablets. Unless you want to be caught in a situation where the pen suddenly won't work you'll need to carry a reserve battery. Having power in the stylus opens it up to future enhancements, but this is a field that is still lacking."

This post from an engineer i on another site confirms some things. http://surfaceproartist.com/blog/2015/4/3/microsoft-working-to-correct-n-trig-pen-jitter

..........

So, Increasing the voltage should improve the accuracy, and that's why there seems to be more jitter when the battery is running low. It seems like the battery or the output voltage itself might be the reason for the inaccuracy.

How to improve it? as the engineer stated... improve the voltage. Well there are not really any AAAA batteries with higher voltage to get, but there are some Smaller 1.5v Button/coin cells roughly the same diameter as a AAAA battery that can be stacked in serial to increase the voltage, although the compromize would probably be shorter battery life.

2 button 1,5v cells stacked should give you 3 volts. I'm gonna try out placing 2 button cells in the surface pen to see if there is any improvement in accuracy. Although i have to put some cunducting material to fill the gap, as the lenght of the stack will be far shorter than an AAAA battery.

I'm not any battery expert, but Maybe maybe a parallel of 2x1.5 volt would work to get a stack of 3x3 volts? to improve the mAmps, and get roughly the size of an AAAA battery.

Time to go shop some coin cells and try it out, Wish me luck!

 

[GreyFox7]

Meh, just rotate the canvas 45 degrees and draw your 45 degree lines horizontally or vertically.

 

[roxasodale91]

tried a surface pro 3 before and i did notice some jitters when using sketchbook pro.
I currently use surface pro 2 and had some struggles with its original pen, so i decided to try wacom bamboo feel and i must say that it is a joy using this pen rather than the original wacom pen and yes, witht he experience i had with the surface pro 3.

 

[tunepunk]

UPDATE!, I tried using button/coin cell batteries in the Surface Pen. It didn't really work as expected. The pen didn't work but the Bluetooth eraser button did. I also tried pulling the nib out and then back in but only halfway in, so it was longer. Pen was still usable, but the jitter increased as the nib got longer. Pressure sensitivity still worked even the nib was just half way in. I'm going to order some extra nibs for experimentation, and maybe an extra pen. It seems like the longer the distance from the actual pen components inside the more jitter. Next test will be to shorten the nib 1-2mm to see if I get any improvements. Maybe the build of the pen is a major factor. If you can shorten the distance from the actual components that produces the field that the digitizer picks up you should be able to reduce the jitter, as the field gets stronger.

Looking at this disassembly of a surface pro 3 pen it seems there is some kind of "antenna" creating the field that the digitizer picks up, so shortening that distance should produce less jitter. Maybe this is the cause of the hover lag as well, since the field gets weaker as the pen is further away?

Oh well, off to try some more experimentation.

 

[Cothek]

I've noticed that the closer to perpendicular to the screen the pen is, the less jitter is registered.

 

[tunepunk]

I tried to shorten the nib a bit. I cut off around 1mm from the nib so the pen itself is actually 1mm closer to the screen now. I noticed a slight decrease in jitter, but not enough to have any major effect. Ordered a 2nd Pen and some more nibs that I'm gonna do some more experimentation with. Next tests will involve some heavier modifications trying to get the pen components even closer to the digitizer.

 

[Cothek]

I tried to shorten the nib a bit. I cut off around 1mm from the nib so the pen itself is actually 1mm closer to the screen now. I noticed a slight decrease in jitter, but not enough to have any major effect. Ordered a 2nd Pen and some more nibs that I'm gonna do some more experimentation with. Next tests will involve some heavier modifications trying to get the pen components even closer to the digitizer.

It seems to me that the orientation of the antenna in the pen makes more of a difference than how close the pen is to the screen. Perhaps try adding some wire to the antenna and see what happens.

 

[tunepunk]

Seems like the latest window update removed most of my jitter =/ Almost no jitter at all now. Did notice a slight increase in the pen lag though. Update details didn't mention any fix.