
OPTIMIZING PERFORMANCE
66
Introduction
Reason is a program of infinite possibilities. You can create as complex
songs as you like, using endless racks of devices. While this is one of the
most exciting properties of the program it does have a drawback – it means
that you must be careful with how you manage your computer processing
power.
Each device you add to the rack uses up a bit of computer processing power
– the more devices the faster the computer has to be. However, you can set
up your devices to require more or less processing power. For example a
sound on the Subtractor synthesizer that only uses one oscillator and one fil-
ter requires much less processing power than one using both dual oscillators
and dual filters.
Samples used in your songs also require RAM - memory - to load properly.
The use of RAM can also be managed, as described at the end of this chap-
ter.
When creating songs for other people, for example for publishing in the Rea-
son song archive (see www.propellerheads.se for more information), you
should do what you can to reduce the requirements for playing back a certain
song, both in terms of processing power and in terms of RAM requirements.
Other users may not have as powerful a computer as you do!
Checking Processing Power
On the transport you will find a meter labelled CPU. This indicates how much
processing power is used at any given moment.
The CPU meter.
The higher this meter goes, the higher the strain on your computer proces-
sor. You will note that when your processor is heavily loaded that graphics
will update slower. Finally, when there’s too little power left to create the au-
dio properly, the sound will start breaking up.
Optimization and Output
Latency
There is s a certain connection between output latency and the demands on
your computer (for more information on latency see page 194):
When your computer doesn’t have enough resources to play back a song,
audio playback will start breaking up. One way of reducing the overhead on
the computer is to raise the latency. There are two ways to do this:
➜ If you are using an ASIO driver, open its control panel (from the
Preferences – Audio dialog) and raise the buffer setting.
This is mainly effective if you are currently using very small buffers, 64 to
256 samples. If the buffers are already big (1024 or 2048 samples) you
will not notice much difference.
➜ If you are using an MME or DirectX driver, adjust the Output Latency
slider in the Preferences – Audio dialog.
✪ Please note that raising these values introduces more delay when
playing devices in real time.
For more information on various driver types, see the chapter “About Audio
on Computers”.
Comentarios a estos manuales