Many engineers have noticed that USB full speed signals appear very strange when probed with an oscilloscope. The transmitted waveform seems to have a “shelf” at about ½ the signal amplitude. This shelf is normal and is expected. What you see on the USB D+ or D- signal lines during USB traffic in a full speed environment is due to the transmission lines being terminated in the characteristic impedance on only the transmitting end of the cable. This “shelf” at ½ of the signal amplitude is the time it takes for the signal edge to propagate past the point of the probe and then reflect back from the un-terminated receiver end of the USB cable. With long 5 meter USB cables, the “shelf” or stair step in the signal can be very long in time and makes the signal appear distorted. With a short USB cable ( ~ 6 inches) it may not be possible to observe the “shelf” in the waveform due to the short time delay of the cable relative to the pulse rise and fall times for a USB full speed signal.
One other aspect of the USB signal is that it is differential, and probing in a single ended fashion can also be misleading about the true nature of the signal.