The Dolby 740 is so special. Any plug ins that do the same?

There's a story from someone working with Iggy Pop, he wasn't happy with the sound of anything until they figured out he was accustomed to the sound of Dolby A, and once they started tracking with that he was peachy. Note that's proper rather than 'improper' use of Dolby A.

I could be wrong, but I don't think the "general public" was aware of Dolby until much later.  Movie and music people knew, because they were using it.  I'm sure at some point a record company or movie studio, or theater, started advertising "Improved Dolby Sound" and then people started to know it existed. 

I've got a Dolby 740 Spectral Processor arriving here tomorrow. 'Very curious about what it can do in mastering. A few days ago I heard Noah Mintz on a podcast spill the beans about it. He said it's in lots of major mastering houses and "they don't like to talk about it". I believe I heard the words "secret weapon" uttered. How could I resist that? I paused the pod and immediately swooped on the only one available in the US at the moment. I paid way too much, but if it does half of what I think it will then it'll pay for itself quickly. It's been a while since I was this intrigued about a piece of gear.

I downloaded and read the manual and can clarify what it does. It's very different from Dolby A, or anything I've ever heard of. It boosts low level detail in the signal. Everything below the threshold you set gets boosted, and the EQ/filter section selects and shapes what frequencies get processed. The "EQ" section is actually crossovers and gain pots, not center frequencies with bandwidth knobs, so functionally it's more like a multiband than an EQ. And there's a noise reduction section. I can imagine so many uses for this thing in mastering and on various individual sources.  And in a M/S network?  Strong medicine!

I was concerned about the amount of circuitry involved and how that could degrade a signal, but thinking about it, the beauty of the design is that it's parallel, so the main signal passes through unaltered and all that circuitry only affects the processed signal that's added in.

You'd have to combine several plugins to try to accomplish what this box does, but even then I doubt it would be the same. If you have some free time, see below for how complex it is.


Introduction
The Dolby Model 740 Spectral Processor is a unique two-channel dynamic equalizer that allows the creative recording engineer to bring out the low-level detail without affecting louder signals or transients. At the heart of the Spectral Processor is a complex multi-band processor, though the unit itself is very easy to use. Refer to the block diagram located at the end of this Section.

Low-level Processing - a new idea
The Spectral Processor works by boosting all low-level information using a sophisticated processing stage. This boosted signal is then passed to a three-band equalizer section to allow the user to select the appropriate parts of the spectrum. The output of the equalizer is then added to the original untreated signal (as shown below) so that the output of the Spectral Processor consists of a signal where the low- level detail is enhanced but high-level signals remain unaltered.

The key to producing a signal that sounds quite natural (even though the low-level signals have been increased by up to 20 dB) is in the design of the processing stage.

SECTION 1 HOW IT WORKS
There are 3 notable aspects to this processing stage:  The gain elements - fixed and sliding bands , How these elements are combined , A multi-level approach
The Gain Elements - fixed bands and sliding bands
The processing stage actually contains eight processors that partition the incoming signal in terms of both frequency and level. These processors work in pairs of one fixed-band stage and one sliding-band stage. The sliding bands provide a constant amount of boost in a frequency band that can slide up and down the audio spectrum while the fixed bands provide a variable amount of boost in bands which are fixed in frequency. The optimum boost for some signals may be provided by a sliding-band processor, whereas for another signal a fixed-band processor may be better.
Combining Fixed and Sliding Bands
The key to the processing is how these bands work together; their outputs are combined in a unique way to gain the benefits of both approaches without the drawbacks of either. For any part of the spectrum the output of the band that provides the most boost is used, resulting in great flexibility in boosting low-level elements.

Multi-level Processing
There are four of these fixed-band/sliding-band pairs, two for high frequencies and two for low frequencies. Within each frequency range the pairs of bands are arranged so that one pair boosts low-level signals, reducing its gain to unity at medium levels where the second pair takes over, providing boost for low-level and medium-level signals gradually reducing its gain to unity at around 0 VU. The two pairs of bands combine to create a smooth transition from a gain of more than 20 dB at low levels to unity gain at high levels. This combination not only doubles the processing effect but also doubles the steepness of the filter slopes where the signal requires it, improving the discrimination between boosted and non-boosted areas of the spectrum.

Three-band EQUALIZATION Control
The processor stage in the Spectral Processor boosts all low-level signals across the entire audio spectrum. The equalization section divides the low-level spectrum into three variable bands allowing the user to select which parts of the low-level signal they wish to enhance. The selected signals then pass on to the summing stage where they are added to the original untreated signal.
The EQUALIZATION section is similar to a three-way loudspeaker crossover network. Low, Mid and High controls set the amount of boost in their respective bands while two crossover controls set the crossover frequencies between the Low, Mid, and High bands.

Source Noise Reduction
The Spectral Processor boosts low-level signals while leaving high- level signals relatively unchanged. It can be thought of as an “audio magnifying glass” which “magnifies” detail in selectable areas of the spectrum.
A side effect of this “magnification” is that some of the details that are magnified may be undesirable -- the most common unwanted detail which may be revealed is noise.
For this reason each channel contains a “source-noise reduction” or “SOURCE NR” section. This SOURCE NR section uses a sliding-band technique that can reduce noise present in the input signal by up to 12 dB. Selectable high-frequency and low-frequency side-chain filters are also provided to prevent unwanted signals from entering the low level processing stage.

Yep, that's the one.  I did haggle him down a few hundred.  Still I probably paid more than I could sell it for, but the idea is to have a new mastering 'edge' that I'll never part with.  

I’ve stopped using it for throughput, it makes everything a bit smaller. I’ve had mine in side chain mode for a few years with that mixing in parallel via an external path. Kinda wish I’d bought a spare when they were $500-700 for so many years. There were never many out there so they never showed up much. Mine was $1495 new in the ‘90’s, Full Compass, maybe ericn sold it to me! When Dolby still had support stock they would give you another for $650 rather than repair yours. It would hurt to be without it!

DOBLY

That's also how it sounded to me at first, until I calibrated it to unity gain using the Output pots, then it passed audio transparently when set flat.  I printed tracks with and without it to A/B, and it was all good.

I spent some time getting familiar with it today, and then used it when mastering a couple of singles.  The obvious first thing to notice is what it can do to the high end.  Pretty damn sweet!  Had to hold myself back from overdoing it, and had to use the built in 8k LPF on the parallel signal to keep it from being too crispy.  Waiting for client feedback, but I loved what I heard.  So far I haven't heard what I was hoping for in the mids from it, but I'll keep experimenting.  Those highs though...

Well these are all pretty old. No surprise that mine might be different. Gear ages. I think mine is a later version, as it has a more blueish faceplate than the other 'khaki" looking ones I've seen pictured, and mine has a higher serial number than some I saw online. This one's sticker says it was built in 1996. IDK how old they all are. I might pick up a TK Mini Blender so I can use it in sidechain mode in parallel. For now I'll keep testing. Good stuff either way!

Today I used only the high band with the crossover set at it's lowest setting (500Hz), and I could tell the slope was very shallow as the effect was skewed towards the highs more than expected.

Still just dipping my toes in this pond...

Now I'm noticing what I think might be what you're talking about.  Patching it inline causes some loss of high end, and that causes some narrowing of the perceived stereo image.  But once I kick in the effect it all comes back and then some.  The masters are coming out great.  Still, I'm haunted by that unwanted side-effect.  'Currently searching for a good way to run it in parallel in my mastering chain.  But until I work that out I'll be alright with the results I'm currently getting...

Oh yeah, if those are your settings in the pic, that's definitely heavy! I go that far on restorations where things are totally out of whack.

Mine's been in a parallel mix setup for a long time now.

it's in side chain mode. If you haven't seen a manual, they suggest the threshold metering should be reading in the middle for best processing, but of course you can push that too.