K-Stereo Ambience Processor

[K-Stereo Ambience Processor]

Any thoughts on this?

K-Stereo Ambience Processor

VST processor for natural and unobtrusive enhancement, expansion, and equalization of the ambience, space, imaging, depth and width in a recording without adding artificial reverberation or causing unwanted coloration or artifacts



K-Stereo is a patented psychoacoustical process (US Pat.7076071) that extracts the ambience inherent in ordinary recordings, and is capable of spreading that uncorrelated ambience around the soundstage, and enlarging the size of that soundstage, both deeper and wider.
In addition, K-Stereo enhances the depth and imaging of the instruments and vocals without adding any artificial reverberation. It does not have a sound of its own; it fills in what’s been missing. K-Stereo is also capable of making a natural mono to stereo conversion.
The plug-in is intended to be compatible with the new VST version 3.0 and is the first in the new Algorithmix Chromium Series, a set of reference mastering plug-ins.

The World’s first Ambience Recovery Processor
Welcome to a brand-new category of audio product: the world’s first Ambience Recovery Processor. You are about to experience a sound enhancement plug-in that can add the last polish to an excellent audio recording, or improve a good recording. It’s based on a psychoacoustical process developed and patented by US mastering engineer Bob Katz.

etc....


http://www.algorithmix.com/en/kstereo.htm

[Is it bollocks!]

Any thoughts on this?

Hmm.... let's see: The patent says:

A process for enhancing ambience in an audio signal output that is derived from an audio signal input in a dual channel audio ambience extraction circuit. The process includes cross-coupling of audio signals in one channel with audio signals in another channel. Each of the cross-coupled signals is attenuated and delayed by an adjustable time period that is within a haas delay time and is then applied in the feedback path to a summing input of an opposite channel. At the summing input, the signals are mixed with subsequent audio signal inputs to that channel. All of the attenuated and delayed signals are continuously applied to outputs of the extraction circuit during the cross-coupling process. The output signals comprise the original signals plus delayed and attenuated reproductions of the original signal along with continuing signals that are submitted to the extraction circuit subsequent to the initial signals.

It’s based on a psychoacoustical process developed and patented by US mastering engineer Bob Katz.

Is it bollocks! Katz never invented anything original in his life - he's an opportunist. Let's go back to an AES report from 1983. Here's the abstract, and look at the bit I bolded:

AES E-Library: The Influence of Antiphase Crosstalk on the Localization Cues in Stereo Signals by Boers, Paul M.

The Influence of Antiphase Crosstalk on the Localization Cues in Stereo Signals

Widening of the stereo base is an attractive feature for stereo sets with closely spaced loudspeakers. This effect is brought about by the introduction of crosstalk in antiphase between the two stereo channels. An analysis of this phenomenon must be based on the directional localization cues of human hearing. The most important cues are interaural differences of level, phase delay, and group delay. We show that the localization directly in front of the loudspeakers is determined by the interaural time delay differences only; to either side of this position interaural level differences play a part, too. At some positions, however, level effects and time delay effects provide conflicting localization cues, leading to a rather vague image of a virtual source. Listening experiments were carried out to verify these results. From this investigation it follows that the introduction of a small time delay in the crosstalk circuit shifts the regions with reinforcing localization cues in the direction of the listener, thus improving the localization of virtual sources in the widened stereo image.

Paper Number:   1967    AES Convention:   73 (March 1983)

Author:   Boers, Paul M.
Affiliation:   Philips Research Laboratories, Eindhoven, The Netherlands

If anybody else attempted to create one of these, I don't think that there could be any serious attempt to uphold the patent claim - it should never have been granted in the first place. And have you seen just how much that damned plugin costs???

Another one... (dodgy patent, that is)

I was still slightly intrigued, so I read through the patent application in some detail, and also looked at the sources he quoted. Boers wasn't quoted at all - but that's not surprising, because it would have immediately alerted the patent office, I think.

He appears to have done sort-of what other researchers suggested, but has extended the delay times somewhat, and also played about with Haas kicks (which is probably the real reason for the time extensions). There's no reason why you couldn't simulate this sort of effect within Audition - there's no process on any given signal that couldn't be easily achieved - but unfortunately I don't think that it would be easy to implement, because of what has to happen across channels in terms of crossfeeds, delay and an inversion.

I can think of a few things that one could try that might achieve much of the same effect - but I think that before I say what they are, I want to try a couple first!

 

"JB Omnisone" free VST springs to mind - possibly erroneously.

"JB Omnisone" free VST springs to mind - possibly erroneously.

That's not the only one, although I'm not quite sure what Omnisone does internally. The other obvious candidate is 'only' a Winamp plugin. That's Vladimir Kopjov's Acoustic Simulation Plug-in for Winamp, which uses adjustable delay crossfeed signals, mainly to improve headphone listening.

A careful look at the patent reveals that to all intents and purposes, the K-stereo Ambience processor is using exactly the same principles. The one thing that's better about it as a plugin is that there is far more control available. But the cost of this seems to be somewhat excessive, to say the least. The idea is based on sound principles, but most emphatically not ones that Katz dreamed up. This looks for all the world like a clear case of passing-off. If I was Paul Boers, et al, I'd sue.

There's no reason why you couldn't simulate this sort of effect within Audition - there's no process on any given signal that couldn't be easily achieved - but unfortunately I don't think that it would be easy to implement, because of what has to happen across channels in terms of crossfeeds, delay and an inversion.

I can think of a few things that one could try that might achieve much of the same effect - but I think that before I say what they are, I want to try a couple first!

I, for one, would be interested in what you come up with, since I've mentioned in another thread that I've discovered some things about delay, inversion, channel mixing, and multiple light reverbs of varying lengths that greatly clear up things in headphones.  They may not be so important in normal monitor or speaker listening since the environment itself comes into play, but headphones need the correct listening ambient created for optimum effect.

I've not used this myself, but a friend has downloaded the trial version and tested it out on one of our latest releases, a 1952 recording of Tchaikovsky's 1st Piano Concerto by Noel Mewton-Wood. He's used quite a gentle stereo effect, and I invite you to download and play with the first 5 minutes of both the original and the 'stereo' versions and see if you can figure out what's being done and how:

http://www.pristinestorage.com/MWmono.flac

http://www.pristinestorage.com/MWstereo.flac

Both are lossless 44.1k 16-bit two-channel audio. Any ideas?

Bringing this subject back to life, I've just obtained a trial copy of the software and tried it out on a 1952 recording of the NBC Symphony Orchestra under Toscanini, which I've also done some basic remastering on.

Personally I think that, regardless of what it is or isn't doing, the effect is pretty remarkable. Try it with cans on, switching from stereo to mono and back again, and you'll hear what I mean - the source clearly remained focused, central and mono, yet you really do get a sense of the space and ambience around it. For people listening to mono recordings on iPods this is surely a boon, but I think it's equally pleasant and perhaps as desirable on speakers.

Take a listen:

http://tinyurl.com/1952LvB9stereo

(Stereo MP3 @ 320kbps, duration 4'51", file size 11.1MB)


And to compare:

http://tinyurl.com/1952LvB9mono

Looks like Bob has found another avenue to get the K-Factor out there...  but I don't get the hype on this one.
I see nothing here that can't be created with existing tools.

The often forgotten or overlooked Pan/Expand function has allowed CEP and AA users incredibly variable image tweaking for over a decade.
Create points to find a static setting that accomplishes your goal.

Better yet, load a file and preview it while you swirl points around within the grids...   it's amazing what a simple stereo track of guitar can become.
Using "preview" I can record the output from Audition with my RME card to capture what I'm creating in real time....  simple, and very cool.

If Audition's Pan/Expand function had automation we'd have a truly fantastic and creative imaging effect.
Add THIS to a feature request list!

Here's a link to a free plugin called SHEPPI that claims to rival the $1K K-Stereo plugin; not having the latter, I can't say much about the comparision, but it does seem to give a nice spatial ambience to mono material; worth checking out and perhaps someone might what to make the comparision and share the results here

http://dallashodgson.info/articles/OpenAmbienceProject/

[*update*]

Here's a link to a free plugin called SHEPPI that claims to rival the $1K K-Stereo plugin; not having the latter, I can't say much about the comparision, but it does seem to give a nice spatial ambience to mono material; worth checking out and perhaps someone might what to make the comparision and share the results here.

I haven't tried it yet, but hey, I like that guy! If you read through what he wrote, and what I said earlier in this thread, you'll see why... (hehe!)


*update* I just listened to both sets of demo files, and I have to say that whatever difference there is between the K-St-rippof and SHEPPI, it's certainly not $1000-worth. In fact it's pretty plain that there are actually advantages to SHEPPI, for instance in terms of bus use, so that would be the one to go for anyway!

There you go, Andrew - that will save you about $1000 at the end of your trial... (I hope you didn't spring for it) 

*update 2* I've just noticed that the bottom of the SHEPPI page has a link to this thread! 

I got a trade price, Steve - but it's repaid itself many times over in sales...