SurroundByUs.com

Sharing the method – Do-it-Yourself examples

After reading this thread and considering it for a couple of days I thought I would share some of the product of that effort. ()

The issue to me seems less about “Layouts” and more about “Setup”. What is desired by those on the learning curve is a rapid means of leveraging the knowledge of those with skill and experience to jump-start their own efforts.

Conclusion: SPEC layout to text output routine. Such a routine would produce a common “lightweight” and human readable structure for sharing setup. The current approach of sharing screen-shots is both labor intensive and prohibits search text indexing. The alternative of sharing the layout file (which encapsulates the settings) presents problems of version/add-in/on/plugin compatibility.

Proposal: Request for Feature. A future SPEC delivery package should include a tool which provides a method to export a text file which describes all the key values, settings (ingredients) of a SPEC layout.

Some detail on the issues which complicate sharing and lead to the aforementioned conclusion:

Results are source dependent. Unless the same source (recording, bitrate and format) as used by the sender is available to receiver the conversion product will be different. What is needed is a common source reference base. I don’t have a solution for this but a description of the source might suffice (e.g. Amazon link to CD or lossless .flac of album track).

Personal Preferences play a large role. Agreed but the learner has yet to develop any preferences and is dependent on example to create them. Without guidance the self-learner is apt to adopt very bad habits or lack exposure to the fullness of feature the environment has to offer.

Target playing environment varies. While important I don’t consider this relevant to the learning process. If the providers play environment differs significantly enough from the receivers then the receiver will figure it out and adjust the settings to cater for the difference. If such changes significantly affect the product then that receiver will need to select the work of another provider to take guidance from.

Legal. Neither the source material nor the product may be shared. However, if the source material is described without being shared along with the settings then the receiver may manufacture the product themselves. Thus the legal (and ethical) problem can be averted.

Comments and critic welcome.

Cheers … Phil Usher

Additional consideration and support for a "SPEC" settings description file format:

While not currently on the horizon I can foresee a future where the SPEC/ArcTan technique may become platform (product/plouge) independent. In such a world a simple descriptor file with platform specific input/output methods becomes vital.

Given the considerable work required to refine a file structure (balancing the human need to read it with the machine reading/performance needs) it would be best to begin the process early.

More, some existing structures and methods may present themselves as models to work from.

Because like you say all music is different as well as individual preferences, constructing a "right way" instruction list is not feasible.
SPEC allows for different approaches to upmixing; the best way to learn is to try all combinations. Sometimes you can set up and run a whole album with identical tweaks (rarely); other times (mostly) a song-by-song approach is optimal.
I know for example that river161 and I approach upmixing in different ways; also there is often more than one approach to accomplish the very same effect.
If you want to be very good at upmixing you must spend the time learning and experimenting. I personally do not take a mediocre approach to upmixing and how others pursue it is unimportant; I know what I know, so to speak.
If you are unsure about the function of some of the controls then experimentation is the key to understanding.
Let your ears be your guide! But if you need help with something, post your questions and we will try to help.

Exactly.

Working with SPEC, and doing all this in general for about six years now, has led to me enjoying what I hear in two, four, five, six, etc. channels more. I appreciate good (and bad) production jobs more.

A lot of folks have asked me over the years for tips as to how you convert a certain genre, for instance. It's impossible, since every artist and album is mixed differently. The key isn't just to take that information in that stereo mix and expand it into 5.1. The key is to manipulate as much of as you can so that you can expand it into 5.1 AND IT SOUNDS GOOD. That's why we're here and not just throwing a stereo source at, say, HD Audio Solo or whatnot.

There are often decisons to make based on what YOU think is right.
For example, say you have an album by one artist with 7 songs.
On every song you come up with centered vocals, and maybe the bass mostly in front right.
But, you have one song where if you center the vocals, the bass might shift slightly to either front left or rear right.
Obviously this song for whatever reason was mixed differently from the rest.
What do you do? There is no right answer, necessarily. What sounds the best? That's the decision you have to make for yourself.

While I must bow to those with greater experience I would also point out that much can be learned by what others have done.

Knowing how someone approached the problem can help one understand a great deal. For those with experience such examples highlight the differences in preference or raise useful discussion. For those without experience (me) it helps me understand the cause and effect of the functions and features.

"Rivers both" - Try to consider for a moment what it might be like if you could remove your knowledge and then look again at the 4.8 SPEC/ZAG world of choice. You both have had the advantage of working with the product and watch/help it evolve. You are experts with a clear knowledge of not only what effect a feature will have upon the sound but why the feature exists in the first place.

You know what to try if the result sounds too bright or if the rear is lacking. You listen to the Stereo recording and already have an idea of the approach you will take even before Plogue has loaded. Those starting (me) don't have that - it's hard to even find the right control let alone know what to do with it.

Encouraging experimentation is fine and justified but with limitless permutations a better primer is needed. The ArcTan features seem grouped in a way that makes sense but what the actual anticipated effect of any given adjustment is most conspicuous by its absence.

Also, while I am on my soap-box let me also take a whack at the no-right-answer opinion - I expect the people who started working on the Music Genome Project faced the same argument. Personally I expect a couple of thousand SPEC setups would do a pretty fair job when matched up against the Music Genome data.

Anyway. What I am proposing is the means to export the setup data which isn't the layout file or screen-shots of the same. Such a text output would provide easy delivery of insight and method which is otherwise unavailable.

As a race driver and instructor of race drivers I am keenly aware that there are a great number of factors associated driving a car fast and that considerable practice is required. However, I am also aware that examples are vital part of learning. Introducing good technique early is far preferred over later correction. I am also aware that some drivers are technically knowledgeable, some have feel but only the very best have both and even among those only a few develop adequate "race craft" to become consistent winners.

It isn't yet clear to me if "Upmixing" has a similar attributes "technical knowledge" and "feel" or if the there is such a thing as "mix-craft". However, assuming there maybe similarities then what I am seeking is a few more expert ride alongs (videos of an upmixs being performed) and track maps (spec settings for specific music).

Unless the barrier to entrance is lowered then only a small elitist community can exist - which would seem a pity given the effort which has gone into the tool-set.

Leaving most of the discussion aside, what would be gained over just having access to the layout itself?

Layouts are just xml by the way, so they are text based already.

My guess is by the time it was developed you wouldn't need it anymore.

Here again you are not asking specific questions other than have someone show you what to do.
None of us here are "professionals" in either software development or instruction in this very limited area of upmixing, and it's all so subjective to the individual doing the conversion. Neither are any of us professionals in music production, mixing, or any other aspect of music other than as a group of people who wanted to share a fabulous technique for converting stereo to surround.
It's not that no one wants to help you or that no one wants to tell you any "secrets" (sic), you just need to have the patience and initiative to spend the time learning the craft.
I can't tell you why I make the decisions I do when upmixing other than experience and understanding what I can do with SPEC to affect the outcome.
....and as I stated before two different people might approach the same music quite differently, and no one wants to get into a "no, I would tell him to do it like this" sort of thing.
All of us have busy lives and make a living doing other things so in some respects it's a wonder this site ever came to be at all!
Having said all that, given time I'll review the guides and if I think it's warranted I'll try to put together a basic primer and if I think it's worthwhile I'll put it out.
In the meantime if you have specific questions about the controls, ask.

Ditto, Please ask your questions here.

That will help me improve the documentation and SPEC in general.