I changed the light and the camera, and it looked very different. See the attachment for the new photos.
That looks normal. The numbers look contrasting and easy to read.
Don't pick on the little things. 
In the photo I see slightly different starting capacitance on all the devices. Is it calibrated or is it still to be calibrated?
Thank you for your deep question.
To be honest, the moment I saw your question, I also started to think.
But I found the answer to the question right away.
You know that at the two ends of the tweezers tip, one end is the output of the excitation signal, and the other end is the input of the excitation signal.
The tip of one of the tweezers is close to the tip of the other tweezers, and a mutual coupling capacitance is formed between the two "metal plates".
As you can see, both tweezers have 10kHz excitation signals. This useful (Interference) signal is detected, and there is a strange unstable offset and noise.
And the other unusual phenomenon is “
Insulated” desktops will also affect the measurement results
Use a video to illustrate this problem couldn’t be better.