I would like to analyze analog circuits by sending a stimulus through it and then recording the output, changing the circuit slightly and then running the test again to produce two *almost* identical .wav files that can then be compared sample-for-sample. Currently I can use a "Spot Frequency Resp." feature of the DSSF3 program on Windows to generate the stimulus and then the "Recording" utility of the same program to record the output. This is through an M-Audio Fast Track USB. I can then test the same unmodified circuit one test right after the other just as a "control". They should be identical minus noise because noise in an analog circuit is random from one run to the next. I then wrote a C program using libsndfile to open the two 48kHz 24bit .wav files, subtract them sample-by-sample and then write a third .wav file representing the sample-by-sample difference of the two recordings. The DSSF3 Recording utility automatically starts when the Spot Frequency Resp. starts but this does not quite work ... The problem is that the sample delay is not fixed. It can vary by several milliseconds. I need the sample delay to be fixed at exactly some number samples. I don't care if it's 10 or 1000, it just needs to be exactly the same every time. Actually it really needs to be a fixed number of clock cycles so that there's almost no time shift at all. Otherwise, even if the input sample was synchronized with the same output sample from one run to the next, if the time was off by 1/2 of the sample period, I think the resulting difference .wav would show differences at high frequencies that are just interference and do not actually represent differences in the circuit. I believe this because if I manually re-align the samples using WavePad, it still shows significant differences between the control files. Meaning the two files are out of phase by a fraction of the sample period. So ideally the ADC should start collecting samples at a certain fixed number of clock cycles after the DAC starts writing samples. Is this possible? Can anyone recommend software / hardware or some combination thereof so that I can digitally transmit a stimulus through an analog circuit and then record it's output after a precise fixed and completely reproducible number of clock cycles? Obviously I'm just doing this for fun so hopefully there is a solution that does not break the bank. The solution doesn't have to be professional - I would be perfectly ok with using some eval board for a codec and writing some dsPIC code or some such to drive it. Of course the ideal solution would be if there was some software that could just get my little Fast Track to do the job.