8,720
edits
Changes
m
→Understanding Modelers
== Understanding Modelers ==
Ideally, modelers are designed to be virtual equivalents of real world gear, and their user interfaces follow such. That said, whenever you are dealing with large combinations of possibilities, unexpected complexities are bound to show up. Modelers can be confusing and frustrating to those more familiar with analog gear. Modelers operate digitally, using digital signal processing Digital Signal Processing (DSP) algorithms (usually on dedicated processor chips designed for DSP). Some may have some analog elements to tweak the signal before converting it into a digital representation - usually this consists of a ''pad switch'' or similar to prevent clipping the input analog to digital (A>D) converter. On the output side, there is typically an analog amplifier (with a volume control) for the analog outputs, after the signal is converted back to analog. This allows you to gain stage it with other gear; however, it is not powerful enough to power speakers. Some may include an actual power amp capable of powering speakers, but most modelers are designed to only provide tone, not power, like analog effects units. Modern modelers that include multiple independent effects/simulations (most of them), often provide you with a virtual signal path, allowing you to select and order the available effects as you want. The number and/or type of effects allowed in a signal path is often a bit limited, which is necessary for the modeler to maintain real-time performance. If the time required to process x seconds of audio is greater than x, then the modeler would not be able to keep up, adding greater and greater lag between input and output and eventually running out of memory to buffer the audio from the input (or it could drop parts of the audio input). Even the most powerful modelers are capable of reaching a DSP limit and throwing an error if you try to add too many effects, with some effects being more resource-hungry than others. Some even offer multiple independent virtual signal paths that can process in parallel, allowing audio to be split and merged or even routed to separate outputs. This can get tricky with analog gear, and there are even more potential issues when doing so digitally. Digital algorithms often need to buffer a segment of the input audio before being able to process it, resulting in lag. This is often so short it is unnoticeable; however, when audio is split and processed separately giving the different paths different lag times then merged back together, this can create a comb filter, where some frequencies are out-of-phase, creating "notches" in the frequency spectrum. Advanced processors and DAW software recognize lag differences and resolve them when merging audio, but not all of them do. While analog gear needs to be gain staged correctly for the best SNR, many modelers need their virtual signal paths to be gain staged even more carefully to avoid clipping. Because the amplitude of the audio is represented as a digital number, exceeding the maximum value will cause overflow. Some modelers may handle this gracefully, subjecting the audio to a form of soft clipping - there is still audible distortion, but it is not raucous. Others don't fare so well, with digital clipping sounding like severe noise and crackles. == Ways to Run a Modeler == === 4 Cable Method ===
== Major Brands/Models ==