We predicted this would be the year of more smart plug-ins and now ReConnect promises to fix bad cable choices made in the recording process after the event using machine learning.

The team behind ReConnect claim that they can analyse an audio file, identify where an inferior cable was used and improve the audio as if a high-quality cable were used in the first place. Even better most of the work is done in the background using machine learning; all the user has to do is decide how much they want to improve the quality using a single control.

What ReConnect Say

When the company first approached us, we were sceptical, so we quizzed them on the science behind ReConnect. This is what they told us..

Cables are critically important system components responsible for protecting the fragile voltages as they are transferred from point to point among the links of the signal chain.

The key physical factors that affect analog audio transmission over wire are shunt capacitance, series resistance, inductive reactance, and shield percentage of coverage. All of these can affect high-frequency response, phase shift, overall signal attenuation, and noise rejection.

We then have the type of materials used to make the wire in the cable, is it oxygen free copper, meaning the cable is less likely to deteriorate internally over its expected lifespan, or monster cable, because lots of strands of copper increase the surface area over which the signal can flow. Then for analog audio cables the larger/wider coax shield reduces the negative impacts of capacitance on the signal. Also we cannot forget the diode effect of dissimilar materials with the oxidation of these connections worsening that potential problem.

That’s a lot of factors to take into consideration and without our unique Artificial Intelligence modelling (AIM), the user interface would have dozens of controls all of which would interact with each other. We have spent years researching this with the best brains in the business to develop AIM that can simplify the user interface down to one knob, to enable you to determine how much improvement you want.

The Impossibly Perfect Guitar Cable

ReConnect isn’t about fixing a broken cable, it doesn’t for example mend crackles or hum. What it does do is it corrects for bad cable choice. In the case of guitar cable the AI implemented in ReConnect analyses the audio footprint of the audio to be processed and substitutes the effect of the cable used for the effect the “right” cable would have had.

Many people who are aware that cable choice affects tone don’t realise that there is more to the effect on tone than just the cable’s capacitance behaving like a low pass filter. This is the best known effect and the longer the cable the more pronounced this filtering of the high end will become. This isn’t necessarily a bad thing and many vintage tones are in part created by this effect but the interaction between a guitar’s pickups and its cable are more complex than this.

As well as the effect of capacitance there is also the effect of inductance to be accounted for. A guitar’s pickups are similar to an inductor in that both are coils of wire, a guitar cable is similar to a capacitor in that both are two electrical conductors placed in close proximity. When an inductor and a capacitor are used in combination the result is a rudimentary EQ circuit which boosts the midrange. Different cables affect where this boost occurs and the cloud based AI in ReConnect bases its cable choice on the results of the audio analysis resulting in the possibility of the impossible: A cable which changes its length during a performance!

What Others Say About ReConnect

The team developing ReConnect have a loyal beta team helping to refine the AIM code and this is the response from a couple of the beta testers…

There is no doubt that the science behind how signals are affected by the cabling and the equipment they connect is sound. What is unique about what the team at ReConnect have been able to achieve with the AIM process is make it simple to use, so we can get on with being creative.

Not All Good News

But now the bad news, a few things may mean that most people will either not be able to afford this technology or be put off by the cloud based processing.

• It’s subscription only, Reconnect tell us that it has to be able to ensure the latest AI processing is taking place, as each time it processes the audio the machine learning improves.

• It’s not cheap with a planned monthly subscription of $179.

• All processing is cloud based, which may give some users security concerns, However ReConnect has told us that all data is anonymised and no personal data is stored. Once the files have been processed, ReConnect assure us that they are deleted from their servers.

• Users must have a physical iLok to use ReConnect. There is no support for iLok Cloud or Machine authorisation.

• ReConnect has also said that because of the professional nature of the plug-in at launch it will only be available for Pro Tool users in AAX format.

What Do You Think?

What do you make of this latest iteration of artificial Intelligence and machine learning? What about the price, expensive, definitely, but with a subscription, you rent it when you need it? Would you be concerned about security with the processing in the cloud? Have you got a project that you can’t wait to use this on?

Do share your thoughts and observations about ReConnect in the comments below.

April Fools

In case you hadn’t spotted it, this article as our April Fools article for 2019. We hope you enjoyed it. For those who had the eyes to see we left clues in the quotes, the first quote was an acronym “Crown Recording Arts Pasadena” and the second was a play on words “Bill Sheet”. :)