The Fairchild 670 uses separate AC and DC threshold adjustments to set where its compression curve sits and how steeply it develops. On the original hardware, AC Threshold lived on the front panel while DC Threshold was an internal screwdriver adjustment.
Most modern compressors hand you a ratio knob and a knee control, then let you treat them as separate choices. The 670 does something less tidy and more useful. Its threshold controls interact, so changing DC Threshold can alter the apparent ratio, knee width, and effective threshold at the same time.
Inside the sidechain, the modeled AC and DC threshold network uses two potentiometers whose interaction changes threshold, ratio, and knee together. Moving DC Threshold toward a broader knee makes compression begin more gradually, so moderate peaks can see a relatively gentle ratio while much larger peaks still approach limiting.
This is why a Fairchild can feel soft without becoming incapable of catching a loud excursion. A broad knee spreads the transition over more input level instead of drawing a sharp corner between untouched audio and compressed audio. Push the signal far enough, and the curve can still become steep. A medium-level phrase may sit in the gentle part of the bend while the next transient lands much farther up the curve.
Universal Audio describes its current model in similar practical terms. Turning DC Threshold clockwise broadens the knee, lowers the apparent ratio, and lowers the threshold, although the eventual curve still heads toward limiting at extreme gain reduction. A displayed ratio therefore describes only part of what your ears are hearing.
Those two points define more than a trigger level. Move the early part of the curve without moving the heavier limiting point by the same amount and you have changed its shape. The technician would repeat the adjustments until the desired slope and position were reached.
This also explains why copying one DC Threshold position between emulations can be misleading. Some recreations expose an adjustment that was hidden inside the original chassis, while others rename it Ratio, DC Bias, or something similar. Control direction and calibration can differ even when the underlying idea comes from the same Fairchild circuit.
The factory setting was only one operating curve, not the only legitimate one. Original documentation describes the 670 as adjustable between compression and peak limiting, with the threshold controls determining where the machine sits between those behaviors. A plug-in that exposes DC Threshold is effectively giving you access to a calibration decision engineers once made inside the box.
This matters when you set the processor by meter alone. Two settings can both show 2 dB of gain reduction on ordinary program material while reacting very differently when a snare crack, vocal consonant, or bass transient arrives several decibels hotter. One curve may ease into compression earlier, while another stays relaxed until the peak pushes it into a much steeper region.
Timing remains a separate decision. The Fairchild 670 attack and release behavior controls how quickly gain reduction develops and recovers, while DC Threshold decides much more about how input level maps to output level. Similar timing with a different curve can feel like a different compressor.
A useful comparison is to match ordinary gain reduction first, then feed both settings a few louder passages without touching the output level. Listen for how quickly the extra peaks stop getting proportionally louder. The setting with the broader knee may start working sooner and more gently, while the steeper curve can leave moderate material freer before clamping down harder on the largest events.
Most modern compressors hand you a ratio knob and a knee control, then let you treat them as separate choices. The 670 does something less tidy and more useful. Its threshold controls interact, so changing DC Threshold can alter the apparent ratio, knee width, and effective threshold at the same time.
DC threshold changes the shape before it changes the number
Calling DC Threshold a ratio control is convenient, but it is incomplete. A modern Fairchild emulation may label the behavior in familiar terms, yet the original circuit does not hold one fixed ratio above a single clean breakpoint. The slope keeps changing as level rises.Inside the sidechain, the modeled AC and DC threshold network uses two potentiometers whose interaction changes threshold, ratio, and knee together. Moving DC Threshold toward a broader knee makes compression begin more gradually, so moderate peaks can see a relatively gentle ratio while much larger peaks still approach limiting.
This is why a Fairchild can feel soft without becoming incapable of catching a loud excursion. A broad knee spreads the transition over more input level instead of drawing a sharp corner between untouched audio and compressed audio. Push the signal far enough, and the curve can still become steep. A medium-level phrase may sit in the gentle part of the bend while the next transient lands much farther up the curve.
Universal Audio describes its current model in similar practical terms. Turning DC Threshold clockwise broadens the knee, lowers the apparent ratio, and lowers the threshold, although the eventual curve still heads toward limiting at extreme gain reduction. A displayed ratio therefore describes only part of what your ears are hearing.
The original calibration used two gain reduction points
The old setup procedure reveals more about this control than the usual compressor vocabulary does. Fairchild instructed technicians to choose desired output levels at 3 dB and 10 dB of limiting, then use different controls to establish those two points. DC Threshold set the 3 dB point, while AC Threshold set the 10 dB point.Those two points define more than a trigger level. Move the early part of the curve without moving the heavier limiting point by the same amount and you have changed its shape. The technician would repeat the adjustments until the desired slope and position were reached.
This also explains why copying one DC Threshold position between emulations can be misleading. Some recreations expose an adjustment that was hidden inside the original chassis, while others rename it Ratio, DC Bias, or something similar. Control direction and calibration can differ even when the underlying idea comes from the same Fairchild circuit.
The factory setting was only one operating curve, not the only legitimate one. Original documentation describes the 670 as adjustable between compression and peak limiting, with the threshold controls determining where the machine sits between those behaviors. A plug-in that exposes DC Threshold is effectively giving you access to a calibration decision engineers once made inside the box.
Soft-knee compression still becomes limiting at high levels
A low apparent ratio on a Fairchild-style curve does not behave like a conventional 2 to 1 compressor with a fixed slope forever. As input level climbs, the variable-mu action increases the amount of control, and the curve can steepen. Very large peaks are therefore treated more aggressively than smaller excursions above the beginning of compression.This matters when you set the processor by meter alone. Two settings can both show 2 dB of gain reduction on ordinary program material while reacting very differently when a snare crack, vocal consonant, or bass transient arrives several decibels hotter. One curve may ease into compression earlier, while another stays relaxed until the peak pushes it into a much steeper region.
Timing remains a separate decision. The Fairchild 670 attack and release behavior controls how quickly gain reduction develops and recovers, while DC Threshold decides much more about how input level maps to output level. Similar timing with a different curve can feel like a different compressor.
A useful comparison is to match ordinary gain reduction first, then feed both settings a few louder passages without touching the output level. Listen for how quickly the extra peaks stop getting proportionally louder. The setting with the broader knee may start working sooner and more gently, while the steeper curve can leave moderate material freer before clamping down harder on the largest events.
