The DMM6500 waveform is smaller in amplitude than the KS34465A, much smaller on larger capacitance DUT. Both appear to be a constant current drive signal into the capacitor DUT, however the KS34465A introduces a DC offset on larger capacitance DUT, the DMM6500 does not. The linear voltage rise across the DUT due to the constant current is followed by an exponential decay as with a simple resistive discharge. Both seem to have the DUT with a repeating waveform which has a frequency that varies somewhat inversely with DUT capacitance.
After observing the waveforms from both with varying capacitor DUT, the algorithm/method is entirely different between the KS and DMM. With a 10nF, the DMM/KS frequency is 455/38Hz, with 0.47uF 93/5Hz, with 22uF 11/2Hz.
To add to HKJ, these DMMs utilize an algorithm/method where the frequency and amplitude varies with DUT capacitance with a unipolar relaxation type waveform, whereas an LCR has a fixed amplitude and frequency with a bipolar sinusoidal waveform, completely different measurement methods and waveforms, and why we mentioned "The technique utilized in these meters performed well relative to the TH2830 which utilized a Synchronous Sampling of a sine wave technique, common to most lab quality LCR meters."
Best,