I've noticed this since receiving the Boom, but since nobody else had mentioned it, I just thought I had a faulty display.

Jim: Does it look better with the latest firmware?

Dean, I still see it easily with firmware 6.

What brightness level?

Created attachment 3318 [details]
test version with high filament current.

    I've seen 2 problems:  

    1) the display gets dim at the edges and bright in the middle.

    2) The displays have a mottled look of bright-dim-bright-dim from top to
    bottom.

    The #1 should be fixable by increasing the filament current in firmware.  I was
    trying to limit the redness of the filaments by lowering the current.

    The #2, may be related to filament current.  Try this firmware.

Caleb's #2 description more accurately portrays the issue I meant to report with this bug. The firmware you just posted looks significantly better but if you look you can still see the mottled brightness. It appears most on the 3rd and 4th column of pixels. 

I think that there's something weird going on here, there's a stripey pattern.  

Sean?

Yes, I still see it.  Like Caleb describes as #2, or a stripey patter.  If I look closely, what I see are 3x3 pixel blocks that are dimmeer than the 3x3 pixel block above and below.  The effect gradually fades from left to right on the display.

Also, take a look at the volume bars displayed when using the two smaller fonts.  There's a very discernable pattern caused by lighter and darker blocks of pixels.

Caleb, is this a hardware issue?  Do we need to fix for ship?

This is because we lower the filament voltage to reduce the redness of the filament at low brightnesses.  We can PWM the filament to get intermediate filament levels.

This is a tradeoff between even brightness across the display vs. brightness of the filaments.

This can be tuned in software.

So, nothing to be done here in hardware.

So shouldn't this be made a software bug and left open until it's resolved?

Okay, so Caleb and Sean, are you aware of what changes Felix needs to make to the firmware to correct this problem?

Caleb/Sean: What should we do with this?

I don't know what could cause this - we have to ask Noritake. Since the bands run horizontally, not vertically, it is more likely an electrical issue or some physical property of this display, not something that could be affected by scan timing. The scan runs left to right, in groups of three columns at a time. 

I don't see this symptom here during the modern age.