Reply to thread

Message: [QUOTE="Queen, post: 90771, member: 27"] Art Davis built the first recognizable graphic equalizer at Cinema Engineering, with six bands offering eight decibels of boost or cut in one-decibel steps. That was it. Six sliders parked at 80, 200, 500, 1250, 3200 and 8000 Hz, a narrow range, and a layout that felt radical because everything before it buried its controls behind rotary switches and guesswork. Davis kept going. He drew the Langevin EQ-252-A with seven bands, then the Altec 9062A, also seven, which stayed in catalogs into the 1970s. Nobody in that era shipped thirty-one faders, and the usual explanation blames heat and expense. That explanation is half right, and it misses the more interesting half. The real obstacle was never the number of valves. It was where a valve had to sit in the circuit, and what the filters on either side of it were doing to your signal on the way through. [HEADING=2]Valves sat at the ends of the chain, never inside it[/HEADING] Look at how the celebrated tube equalizers actually work. The Pultec EQP-1A, introduced in 1961 as a replacement for the 1956 EQP-1, is a passive filter network with a tube amplifier stage that gives back the gain the filters throw away. A 12AX7 handles the input, a 12AU7 drives the output transformer in a push-pull arrangement, and neither one is doing anything band by band. The Ling Electronics EG-10-B makes the point harder. Built in the 1950s for military work, it spreads twelve bands from 4 Hz all the way up to 16 kHz across inductors and interstage transformers, and it is fully passive apart from the tube input and output stages. Twelve bands. Two tube stages. Passive filters lose level. Stack more of them, and you lose more, so the tube's job is to hand that level back at the end rather than fight the filters along the way. [B]Passive filtering with valve gain bolted on either end[/B] was not a compromise anyone regretted, and it was the only workable arrangement once a design carried more than a handful of bands. A thirty-one band valve equalizer would therefore not have needed thirty-one tubes. It would have needed a passive filter bank deep enough that insertion loss became the dominant problem, plus a makeup stage working hard enough to color everything whether you asked it to or not. [HEADING=2]Third-octave filtering arrived after tubes had already lost[/HEADING] Narrow-band work did not show up until 1967, when Art Davis, Jim Noble and Don Davis developed the Acousta-Voice system for Altec-Lansing. It was the first analog third-octave variable filter set, and Don Davis presented the work to the Audio Engineering Society, which is where most accounts date the modern era of sound system equalization. The timing matters more than the engineering. By 1967 the industry had already moved to solid-state for anything carrying this much circuitry, so the narrow-band idea and the valve era never overlapped. There was no window in which somebody could have built the thing and decided against it. Acousta-Voice was passive as well, a variable notch filter set aimed at ringing out rooms and holding feedback down rather than shaping tone. Constant-Q third-octave designs came later. Bob Thurmond built the first one for private use in 1973, by which point nobody was reaching for a valve. [HEADING=2]Thirty-one bands was never a number everyone agreed on[/HEADING] Even the band count is less settled than it looks. Klark Teknik's DN360, a touring standard for decades, runs thirty bands per channel. The dbx 231S runs thirty-one. Both sit on ISO center frequencies, and both are right, because the ISO series extends wider than any single chassis chooses to expose. So when [B][URL='https://goldmidi.com/community/threads/m-media-audio-dropped-reference-player-third-rail-tube-eq.76097/']a valve-voiced 31-band graphic running as software[/URL][/B] turns up in 2026, it is not restoring anything. There is no vintage rack unit to measure, no surviving schematic to argue over, and no front panel anyone remembers from a session in 1972. That changes what the tube stage in such a plugin can honestly be. It cannot model a specific circuit, because the circuit was never built, so whatever saturation you hear is a designer's judgment about what a valve might have done to thirty-one bands of summed filtering. You are free to like that judgment or not. What you cannot do is check it against hardware, and anyone telling you a plugin like this sounds like the original is describing a unit that never existed. [/QUOTE]

Name