passing points

I would first check to make sure you have a good master/cal done on the robot.

Are both points off by the same amount?
Are there only those two points that are the issue?
Is there an offset used?
Why was an image restored?

If there is a difference on the points after a master/calibration it means that the program points were modifed once or several times in the past. It's not uncommon for old robots to correct their programs due to reducer plays or bad calibrations. Put the robot joints on 0 position and verify that 0 software position is same as the hardware position (witness marks). Keep in mind that after calibrating the robot to witness marks its IMPOSSIBLE to hit the exact same spot as the previous calibration. In my programs at least I always have minor differences when I have to recalibrate the robot after a motor change for example. Even 0.1 degrees out of the previous calibration in one or many joints will cause program points to shift and sometimes this shift makes a huge difference in your program.

There is a fix around that but I would advise you against it.
1) Run your program and go to the first point or any point that the difference margin is noticeable.
2) check your joint coordinates. Write them down.
3) put the robot in joint mode, and adjust each axis separately to correct the point.
4) make a new point below the original one.
5) You may need to adjust 1,2,3 or all 6 axis to correct the problem by several degrees + or -.
6) write down the JOINT coordinates of the new point.
7) Now calculate the difference in degrees between the original point coordinates and the new one. For example, if J1 was at 123.600 degrees and you positioned your J1 to 124.500 to correct the point then the difference will be +0.9 degrees.
put your robot to SOFTWARE 0 position. (not witness marks).
9) activate the position screen and switch to JOINT frame.
10) add or substract the differences you wrote down to each axis.
11) remaster your robot to those positions.

Again, I would strongly advise you against this strategy. It effectively means your robot has a "custom" calibration away from its marks.

Quote from programmer0011

put your robot to SOFTWARE 0 position. (not witness marks).
the zeros exactly at witness marks after quick master and calibrating when I restored the image .
if you mean home position the amount for axis are :
j1 15.000
j2 -25.000
j3 -5.000
j4 0.000
j5 5.000
j6 0.000
would you like me to set the differences into the zero positio at witness marks or home position ?
11) remaster your robot to those positions.
Is it necessary to do calibrating after remaster to those points ?

Display More

SOFTWARE zero position is the position that every axis reads zero degrees on the teach pendant position screen. This is only to verify that your previous calibration was done right and you didn't make a fault or forget something. Put every axis on zero degrees in the teach pendant and check if this position aligns the witness marks of each axis.

This step is optional. After, go inside your program, move the robot to a point that you know it has shifted, write down the JOINT COORDINATES and try to correct the point in joint frame using as less axis as possible. Experiment by slowly turning each axis separately. You will eventually figure out which axis has shifted and to where (+ or -). Make a new point as I described previously and write down the difference between the new and old points in JOINT Coordinates. Then put the robot axis on zero position again and add or substract the differences you wrote down. After, quickmaster the robot to those positions.