The soundcard is the timing reference
MMSSTV depends on the computer sound card for both audio and timing. SSTV does not require the sound card to run at a perfectly exact nominal sample rate, but MMSSTV must know the card's real timing accurately.
Clock error appears visually as a slanted SSTV picture.
Sampling frequency
The original documentation discusses several sampling frequencies and notes that 11025 Hz, 22050 Hz and 44100 Hz are widely supported.
Higher sampling frequencies require more CPU power and memory.
RX buffer
The RX buffer temporarily stores received audio so MMSSTV can redraw a picture after correcting slant or phase.
The original documentation states that the buffer contains approximately 270 seconds of audio data and can therefore be relatively large.
FIFO
MMSSTV uses FIFO buffers for low-level sound processing. Increasing the FIFO can help when a computer loses audio samples or synchronization.
Accurate clock calibration
For high-accuracy calibration, the original manual suggests receiving a accurately timed reference signal such as WWV, JJY or a suitable FAX transmission.
- Open Option → Setup → Misc.
- Press Adj.
- Receive the timing reference.
- Tune to a periodic timing signal.
- Allow a vertical reference line to develop.
- Click the upper and lower points of the line.
Separate TX and RX clock offsets
Some sound cards use different effective sampling clocks for recording and playback. MMSSTV can therefore calibrate TX separately from RX.
The original procedure uses an external loopback from line-out to line-in, transmits a Martin 1 image and adjusts the received loopback image before storing the result as the TX offset.