The S1 and S2 deliver different samples per packet for each sample rate, so to determine the interval between data packets received take this size and divide by the sample rate.
From this table a sample rate of 1,536 kHz or 6,144 kHz will deliver a 'smoother' data flow, resulting in fewer audio pops and associated stuttering as well as lowest latency.
The solution is to run a separate ELAD reader thread which collects and buffers the data; this buffer is read at a stead rate by the main threads in the program to ensure a steady data flow. This is the same approach implemented for the SDRplay and will be added here when time permits.
All ELAD sample rates are native to the ELAD and supporting API, they are not created as a result of extra processing by this software
June 11th, 2108 User-selectable buffer sizes (low, medium, high) are now supported. Normally low (lowest latency) works well, but some motherboards / computers have considerable latency variation, so increasing the buffering reduces and eliminates breaks in the data flow which you hear as breaks / stutter in the audio.
When you start the ELAD radio there is a delay of a few seconds while the API loads code into the radio - this is normal and nothing to worry about.