Reply to thread

Message: [QUOTE="Bombastus, post: 91937, member: 2178"] The original Fairchild 670 used a stepped input attenuator in 1 dB increments, while its front-panel threshold control was continuously variable. Those two controls can both change gain reduction, but they do not put the circuit in the same operating state. Treating Input Gain and Threshold as interchangeable “more compression” knobs misses the useful part of the design. You can arrive at similar meter movement with different combinations, yet the signal can hit the gain-varying tubes at a different level before the sidechain pulls it back. [HEADING=2]Input gain sets drive before the limiter reacts[/HEADING] Despite its name, the original Input Gain control is an attenuator rather than a little preamp waiting to be cranked. Fairchild’s normal setup procedure called for roughly 10 dB of attenuation to reach unity gain, and the hardware provided 1 dB steps. Fully opening the control simply removes more attenuation from a circuit that already has gain. Modern versions do not always preserve those numbers. Universal Audio’s current 670 model, for example, labels its input range from -20 to 0 dB and places unity around -4 dB, so a setting copied from vintage hardware is not automatically the same operating point. A clone or another plug-in may choose a different calibration again. Input level matters because the control sits before the gain-varying stage. The 6386 remote-cutoff triodes alter amplification as their grid bias changes, which is central to [B][URL='https://www.dafx.de/paper-archive/2012/papers/dafx12_submission_9.pdf']variable-mu gain stage behavior[/URL][/B]. Driving more signal into that stage can increase distortion even when you later adjust Threshold to bring the gain-reduction meter back to the same reading. A cleaner setup therefore does not necessarily mean less compression. Pull the input level back, then move Threshold toward more limiting until the meter returns to the amount you wanted. You have kept the control action similar while reducing how hard the signal hits the tube stage first. [HEADING=2]Threshold changes limiting without replacing input gain[/HEADING] The original operating instructions start with the AC Threshold control fully counterclockwise, where limiting is removed. Turning it clockwise increases limiting until the desired reduction or output level is reached. Threshold controls when and how strongly the sidechain makes the gain stage back off, not simply adding more audio level at the input. Its scale should not be read like a modern threshold value in dBFS. The number on a Fairchild-style control is meaningful only inside that particular implementation and calibration. Copying “Threshold 3” from another plug-in can be useless if its developer mapped the control differently. Hardware recreations make this more confusing because some reverse the numbering or behavior while keeping familiar Fairchild graphics. One unit may put heavier limiting toward the clockwise end, while another may use a scale where the larger printed number means less compression. Watch the gain-reduction meter and learn the direction of the specific unit instead of trusting a screenshot. The distinction also matters when you are chasing color. Raising Input Gain and backing off Threshold can preserve roughly the same reduction while feeding the audio path harder, whereas lowering Input Gain and asking Threshold for more limiting can produce a cleaner version of similar dynamic control. The meter can look nearly identical while the harmonic contribution changes. [HEADING=2]Matched gain reduction can hide a different sound[/HEADING] Take a mix bus sitting around 2 dB of reduction. One setting can reach it with more input drive and a less aggressive threshold position, while another reaches it with less drive and more threshold action. The first can sound denser because the gain-varying stage sees more level before control voltage reduces its gain. Level matching matters before deciding which setting is better. Use a downstream trim, or an emulation’s added output control, to bring processed and bypassed loudness close enough for a fair comparison. Otherwise, the louder setting can win even when the difference you wanted to judge was distortion rather than dynamics. Timing comes after the operating level is under control. A release choice can feel much more obvious once the limiter is being driven harder, so [B][URL='https://goldmidi.com/community/resources/fairchild-670-compressors-attack-and-release-times.49/']Fairchild 670 timing choices for attack and release[/URL][/B] are easier to judge after Input Gain and Threshold have stopped moving. Numbers still need context when you save or copy a preset. Record the plug-in or hardware version, the input calibration, the amount of reduction, and whether any added output trim is active. Two Fairchild-style processors showing the same knob positions can be operating at materially different internal levels. [/QUOTE]

Name