I dropped one of the digit anode resistors to 1.5K. Measuring the voltage across it when the digit 8 is displayed gives 1.9 volts. So the current through this tube now is 1.27 ma, which is less than what it was with the 3K resistor. The brightness of the tube didn't really change much. Anyway, this seems counterintuitive! Must be something in the programming/multiplexing that is compensating.
The thing that comes to mind is not enough available power. Have you tried a different power supply? What is the voltage at the 170v TP when it's running? Have you done a factory reset to re calibrate the HV switching timing? These are all things that could produce insufficient brightness.
Also, if you have a scope, we can take a look at the anode control lines. At full brightness the anode drive should be more or less 1/6th to each tube. If you are seeing very thin spikes here, it's likely that dimming is the problem.
Just another thought. Does the brightness increase if you only drive one tube? (e.g. take out all the other optos).
The average voltage (measured with a volt meter) across the 3K resistors is not there. While displaying the digit "8" I get 3.5 volts across the 3K resistor for an average current of 1.2 ma, hence the brightness problem.
Keep in mind that I've disabled the LDR in settings for all of my measurements.
I've pulled all of the optos except for the minutes digit. Without resetting the clock, when I apply power, the average voltage across the minutes 3k resistor starts off at about 5.0, but after 30 seconds or so reduces down to about 3.5. Brightness follows.
I'll let you know what happens at 9 minutes and 15 seconds after the hour.
I do have a scope, and we can go down that path if necessary.