Re: [Sursound] Patent application: Data structure for HOA
I received a message back from Jan-Mark Batke, to the effect they will pass my comments on to the patent authorities. It is classified at this stage as a disclosure. The four inventors are members of Technicolor, and the new system is briefly featured here: http://community.calrec.com/?p=8268 It does seem expressly targetted at cinema applications, so it remains to be seen how relevant it may be for musicians etc. I have (at last) updated by description page for AMB**, and have indeed added a link to the UA description. Now the attention in previous posts was very much on the phrase most sophisticated format, which was guaranteed to wind people up; whereas the key word is really available. The UA format is not really available to ~composers~ to use. The description is very much one for prospective developers - acquiring wavpack, and one way or another implementing all those equations (and apparently creating a WAVE file with a large number of silent channels!). The clue is for example in the observation on that website that no player is currently available; and when someone comments positively on a piece of yours, you are obliged to suggest they decode the file themselves, but Unfortunately, getting the software to decode ambisonic stuff is kinda annoyingly painful In short, for any file format to be deemed available there ~must~ be some associated application or set of applications that can be used to create, process and render a file. This means also that there must be no political or cultural platform aversions - to be available the format must have support not merely in Linux but, arguably much more importantly, in Windows and OS X. Users really do not need, or want, to deal with mathematics or complex configuration steps drenched in jargon. Reasonable defaults must be available, so a composer can launch an app, pan a sound as intuitively as possible, and write the file. And then automatically play it back. And send it to a friend who can also automatically play it back. To me this is obvious, which is why the publication of the AMB format (1999/2000) coincided with its incorporation in the CDP Multi-Channel Toolkit, which many people have used subsequently to make and publish AMB files. So until this situation materially changes, while AMB is clearly not the most sophisticated file format ~published~ it may still be the most sophisticated one ~available~. Whatever objections people here may have to AMB (and clearly they are legion), the one thing the Toolkit programs can justifiably claim is that they are not annoyingly painful to use. The only challenge, indeed, that they represent to the user is the basic ability to use a command line. I get a nice trickle of emails from people thanking me for their availability; sadly not so many of then go the extra mile and click my Paypal button :-(. So updates and extensions will be infrequent at best. So for those new file formats to become available is is down to those who can afford the time; university departments, etc. Richard Dobson **http://dream.cs.bath.ac.uk/researchdev/wave-ex/bformat.html On 25/10/2012 01:16, etienne deleflie wrote: So is this, in fact, the ultimate file format that folk on this list have been arguing for (and over) for so long? I dont know about ultimate formats ... but one existing format is Universal Ambisonic (UA). It is documented Here: http://soundofspace.com/static/make_ua_file And there is lots of material in this format available on http://soundofspace.com This format is my attempt to *conclude* on the many discussions we had here and on other lists. I don't pretend that it is better than other formats ... nor that it satisfies everyone's needs (even though it tries pretty hard). The point is ... other ambisonic formats exist ... and UA is one of them! ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
As I've said ad nauseum, the guy who first integrates an Ambi decoder into VLC, getting around the evil Windoz mixer etc. gets to choose the data structure for next important Ambi format. This will be a lossy compressed format probably based on the public domain Vorbis. Ambisonia was the 1st major breakthrough, source material. I hope everyone is aware and thankful for Etienne's huge amount of effort work. We are lucky that York have taken over this and hope it will continue to increase and prosper. The 2nd is the ongoing decoder work by BLaH and others on this forum. At least the theory of how to design a good decoder is available. Sadly, only Fons' Happy Days decoder is flexible enough to take advantage of the new work. This isn't descrying the decoders available from D McGriffy, Richard Dobson, the York Mafia others; just pointing out that they are 'fixed'. But none of these will conveniently play MP3s ripped from CDs or youTube, surround videos etc .. no nice database for music .. so will remain niche interests. Happy Days doesn't even run on evil Windoz and probably never will if its inventor has any say in it .. 8D I finally managed to compile VLC this year but can't seem to do it again. 8( When Ambi VLC happens, I predict the re-surrection of UHJ. Simple 2 channels will remain the most important distribution format in the forseable future. But it will pave the way for HOA and other exotics. -- next part -- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 2770 bytes Desc: not available URL: https://mail.music.vt.edu/mailman/private/sursound/attachments/20121028/48dac521/attachment.bin ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
Please not! He who is happy with lossy compression is hardly a candidate to have a properly set up surround system, much less one suitable to Ambisonics. Lossless compression is OK, even desired, as an option, preferably something that's freely licensed and enjoys commercial support e.g. ALAC Sent from my mobile phone On 28 Oct 2012, at 03:11, Richard Lee rica...@justnet.com.au wrote: This will be a lossy compressed format probably based on the public domain Vorbis. -- next part -- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 5863 bytes Desc: not available URL: https://mail.music.vt.edu/mailman/private/sursound/attachments/20121027/4ff2ef7c/attachment.bin ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
Hi Richard, Now the attention in previous posts was very much on the phrase most sophisticated format, which was guaranteed to wind people up; whereas the key word is really available. The UA format is not really available to ~composers~ to use. The description is very much one for prospective developers - acquiring wavpack, and one way or another implementing all those equations (and apparently creating a WAVE file with a large number of silent channels!). I really didn't want to get pulled into a defence or argument about ambisonic formats ... but, just to clarify ... the choice to include some empty channels in UA is intentionally designed so that authoring environments don't need to change all the channel routing when working at different orders. The choice of Wavepack was determined on its ability to compress empty channels to take up no space. Wavepack also efficiently losslessly compresses all sound data. Already on these two points UA is far more practical for composers. You only need one setup to work at different orders. UA was actually designed *for* composers. I agree that there are many remaining tools required for it to be *actually* practical for *listeners*. But there's the grab ... I think the ultimate mistake is to think that ambisonics should be a consumer oriented format. (both ambisonia.com and soundOfSpace.com distribute the files as already decoded speaker feeds) That's where so many issues start to creep in. When a consumer gets an ambisonic file then: - the audio player needs to be smart, it needs to do work far beyond what audio players are used to doing - speaker agnosticism is a false benefit ... in reality, ambisonic order choice is largely determined by the targeted speaker array. Note there ... targeted speaker array is the opposite of speaker agnosticism The way I see it Ambisonics is a production format. Etienne ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
I suspect that - any file format that has any level of sophistication (read: complexity) will likely not get take-up (maybe even UA is too complex. Straight old B-format is fine). Its not what features are included that counts that's the engineer's mistake. - any file format which can't relatively-easily be output by a DAW will likely fail (Both UA and AMB require an encoding step after DAW output) - any file format that takes control away from the composer will be rejected by the composer. Abstracted formats which force the spatial composer to think in certain ways will only see takeup by those unaware... eg: thinking of spatial audio as mono-channels of sound which are then 'spatialised' in a cartesian coordinate system. In such formats, although the engineers may not realise it, they become as much a composer of the results as the composer themselves (see papers by Agostino Di Scipio on 'techne') Etienne On Sun, Oct 28, 2012 at 9:27 AM, etienne deleflie edelef...@gmail.com wrote: Hi Richard, Now the attention in previous posts was very much on the phrase most sophisticated format, which was guaranteed to wind people up; whereas the key word is really available. The UA format is not really available to ~composers~ to use. The description is very much one for prospective developers - acquiring wavpack, and one way or another implementing all those equations (and apparently creating a WAVE file with a large number of silent channels!). I really didn't want to get pulled into a defence or argument about ambisonic formats ... but, just to clarify ... the choice to include some empty channels in UA is intentionally designed so that authoring environments don't need to change all the channel routing when working at different orders. The choice of Wavepack was determined on its ability to compress empty channels to take up no space. Wavepack also efficiently losslessly compresses all sound data. Already on these two points UA is far more practical for composers. You only need one setup to work at different orders. UA was actually designed *for* composers. I agree that there are many remaining tools required for it to be *actually* practical for *listeners*. But there's the grab ... I think the ultimate mistake is to think that ambisonics should be a consumer oriented format. (both ambisonia.com and soundOfSpace.com distribute the files as already decoded speaker feeds) That's where so many issues start to creep in. When a consumer gets an ambisonic file then: - the audio player needs to be smart, it needs to do work far beyond what audio players are used to doing - speaker agnosticism is a false benefit ... in reality, ambisonic order choice is largely determined by the targeted speaker array. Note there ... targeted speaker array is the opposite of speaker agnosticism The way I see it Ambisonics is a production format. Etienne -- http://etiennedeleflie.net ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
