[cerberus]

Originally Posted by bob katz

I am now the proud papa of a newly-granted U.S. Patent I'd like to take this opportunity to announce it to my fellow Gearslutz. After a long process with the U.S. government and my patent attorney, I have just received U.S. Patent number 7,076,071, which covers the algorithms behind my K-Stereo and K-Surround processes, which I have licensed to three companies, Z-Systems, Weiss, and Digital Domain (my own company).

congrats!

Just like EQ or any other technique, K-Stereo is not a process that you just "apply automatically." I use it on a small percentage of mixes that come in, and if a recording does not need it, then K-stereo can make it sound washy or ugly or if pushed too far, reduce the level of the vocalist in the mix, or bring out the booth in which the percussion was recorded. It's meant to be a polishing technique, the icing on the cake, not the steak and dumplings! But the key is that instead of adding reverb, which can muddy up an existing recording or mix, K-Stereo can enhance what good sound is there but may have been too small.

you have helped me develop my signal chain this way using off the shelf software. i've been showing results in the wump exhibitions at brad blackwood's forum... and here at gs, where i mastered a track as a demonstration; here i intentionally used the madsen effect to make the mix seem bigger and more vivid: http://gearslutz.com/board/showthread.php?t=67257

after i read the patent, i refined my process (your process!) further. in fact most mixes i recieve are too small! i am finding the effect useful on a high percentage of jobs now.

the process can be effective at low levels where it cannot actually be heard... it seems to modulate sounds that drop into the noise floor in a converse fashion to the way dither uses noise to modulate sound in this area..except that this is correlated noise..it does a lot more than dither...much more audible at similar levels. in order to replicate my amazing results, one would need to start with an appropriate amount of dynamic range.

Interestingly, the more squashed and compressed and squeezed a master or mix, the less useful K-stereo becomes because it just doesn't help. In other words, K-stereo is a natural kind of enhancement. But it works on all types of music, just doesn't help a squashed recording.

based on what i have just said, this is a fact to me and it's no mystery why. by the time i apply "ersatz k-stereo", i've usually added significant amount of dynamic range to the mix. in most cases, i am able to pull the dynamics of a mix into a shape suitable for the process. the reason not to use it on every job would be that the client can't afford to pay for enough time for me to bring the mix to that level, or it is a restoration job... in exhibitions such as wump, i use it every time lately.

the discrepancy between what you recommend and what i find seems to come from:

1. i maintain a pure floating point signal path throughout the mastering process. i pass these signals around the limiter, never through a limiter such as l3, absolutely never.

2. i am very manipulative in general in order to get "natural" or even "surreal " dynamics. my signal path is highly parallel, yesterday i made a master for the wump exhibtion using over 30 parallel signal paths with tight control over delays... the music was an "electrified" string quartet.. (think kronos quartet) very very subtle job... i think my best ever...k-stereo made the decay so divine i want to eat it. it helped make the hihats sound like real [metal] cymbals again too... and on the violin and cello cresendoes... one can -feel- overtones modulating like flames reaching into the listener's chest.

3. reverb is out of fashion... especially eighties style.. i really feel that the tc-6000 seems to have embarrassed everyone who doesn't own one into backing off. to me the entire reverb landscape looks like: convolution 'verbs, the classic [lexicon] 80s sound, and the 6000. [to answer the original question: i did use convolution reverb on a master in january... it was electronica containing a few samples, but no acoustically recorded tracks at all; so that doesn't count. almost never then.]

4. i have a very complex parallel chain, ersatz k-stereo is a helper, but there is a role in this chain for the smallest of helpers.... i work very incrementally, so i never expect a miracle from any one process... i control delays in this chain. i often add delays... so the actual part of the process that is now derived from the k-stereo patent is indeed "icing on the cake"... it would be incorrect to take a simple chain and just add k-stereo, because then i think it would "stick out"... one has to work on the microdynamics everywhere before the signal is ready to be processed with this very subtle tool....

i also often have madsen effect on bass enhancement... very important because my ersatz k-stereo array often works below the noise floor where bass really isn't. so a secondary madsen enhancement on bass, which in this case is an "artifact" not present in the mix needs to co-exist.

so i would say that a really good engineer who gets really good mixes might find themselves using k-stereo on more than half their jobs. but that is all qualified... the light sabre itself can't make one an instant jedi master.

i've invested years in learning stereo enhancement... it's so important to me.... because i got into engineering for this reason... in terms of other aspects of engineering... the beatles records already had it nailed, so stereo was the only thing i felt i could improve on...

i can say for sure that k-stereo is solid science. however i am dyslexic with math... so my ersatz k-stereo is constructed and tuned by ear, it does not really follow the patent, of which i have a very superficial understanding of the mechanics. but the concept owes directly to the patent, and to bk's direct and generous suggestion to me on internet forums that i should attempt to harness the madsen effect.

and now i am like...holy sh*t... because i add the process towards the end of a great mastering session and it always gets better...such a small incremental improvement -is- magic to me, because it always happens just when it seems that the sonics can't possibly get any better. it is like when you look with one eye, and then open the other eye.



jeff dinces