As described in “About latency” in the Audio Basics chapter, you generally want the lowest possible latency, to get the best response when you record and play in Reason Essentials in real time. However, selecting too low a latency is likely to result in playback problems (clicks, pops, dropouts, etc.). There are several technical reasons for this, the main one being that with smaller buffers (lower latency), the average strain on the CPU will be higher. This also means that the more CPU-intensive your Reason Essentials song (i.e. the more devices and audio tracks you use), the higher the minimum latency required for avoiding playback difficulties.
If you are using an ASIO driver (Windows only) specifically written for the audio hardware, you can in most cases make settings for the hardware in its ASIO Control Panel. This panel (opened by clicking the ASIO Control Panel button in the Preferences-Audio dialog) may or may not contain parameters for adjusting the latency. Usually this is done by changing the number and/or size of the audio buffers - the fewer and smaller the audio buffers, the lower the latency. Please consult the documentation of your audio hardware and its ASIO drivers for details.
If you are running Reason Essentials under Windows and are using an ASIO driver, or if you are running Reason Essentials under Mac OS X and are using a Core Audio driver, you can adjust the input and output latencies on the Preferences – Audio tab.
Below the Input Latency and Output Latency rows in the Preferences-Audio dialog, you will find an item called Recording Latency Compensation. This value is used internally in Reason Essentials to compensate for latencies when recording audio using external monitoring.
If you're recording audio and not monitoring through Reason Essentials (e.g. monitoring directly through the audio card or via an external mixer), the audio you record will reach the program slightly late. This is because you play along with background tracks or the metronome - and you hear these delayed by the output latency. Also, the sound you record is sent to the program via the buffers in the audio hardware - it is delayed by the input latency.
Reason Essentials compensates this by moving the recorded audio earlier by the sum of the input+output latencies. This compensation is also done in "Manual Monitoring" mode, if Monitor was Off for the recorded track (when you started recording).
If Monitor is On, there is no Recording Latency Compensation. This is because the monitored sound will also be delayed, and numerous tests have shown that the performer will actually play slightly ahead to compensate for this. In other words, when monitoring through Reason Essentials, the performer is expected to do the Recording Latency Compensation himself/herself.
If you are monitoring via an external mixer, and have selected “External” in the Monitoring section on the Audio tab in Preferences, there might be situations where you experience that the recorded audio is generally played back too early - or too late - in the song. This could be because your audio card doesn’t actually have the latency values it reported to Reason Essentials.