Re: [Sursound] Patent application: Data structure for HOA
Unless things have changed a lot, last I checked lossy compression messes up phase relationships, and that would be an issue for things like UHJ, which as long as portable stereo players with limited battery life (and thus limited CPUs), is the only viable, because stereo compatible, distribution format. At this point in time, not only is most music listened on mobile devices, most music is even purchased on mobile devices, and that's strictly a stereo (or maybe binaural) world. Try this simple experiment. Take your favourite Nimbus UHJ CD and rip it using the most evil MP3 encoder you can find .. probably the one built into the latest Windoz Media Player. Do this at 256kB/s and also (shock! horror!) at 128kB/s. Now listen to the resultant files on a mobile device. Then you can pontificate to us on how the musicality has all escaped and no one is going to find these acceptable. You can also rip to a WAV file if your mobile device will play these and compare the MP3s with the 'original'. This is just testing Ronald's assertion about compressed UHJ on stereo mobile devices. I dunno about full UHJ surround decode cos there don't seem to be any good ones in the public domain. PS I expect you to hear ve.eery slight differenes with one MP3 and probably none with the other. I won't insist on Double Blind bla bla but you might find that educational. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
Am I missing something? - for mobile use, wouldn't B-format to binaural be better than UHJ? Dr Peter Lennox School of Technology, Faculty of Arts, Design and Technology University of Derby, UK e: p.len...@derby.ac.uk t: 01332 593155 From: sursound-boun...@music.vt.edu [sursound-boun...@music.vt.edu] On Behalf Of Richard Lee [rica...@justnet.com.au] Sent: 30 October 2012 19:51 To: 'Surround Sound discussion group' Subject: Re: [Sursound] Patent application: Data structure for HOA Unless things have changed a lot, last I checked lossy compression messes up phase relationships, and that would be an issue for things like UHJ, which as long as portable stereo players with limited battery life (and thus limited CPUs), is the only viable, because stereo compatible, distribution format. At this point in time, not only is most music listened on mobile devices, most music is even purchased on mobile devices, and that's strictly a stereo (or maybe binaural) world. Try this simple experiment. Take your favourite Nimbus UHJ CD and rip it using the most evil MP3 encoder you can find .. probably the one built into the latest Windoz Media Player. Do this at 256kB/s and also (shock! horror!) at 128kB/s. Now listen to the resultant files on a mobile device. Then you can pontificate to us on how the musicality has all escaped and no one is going to find these acceptable. You can also rip to a WAV file if your mobile device will play these and compare the MP3s with the 'original'. This is just testing Ronald's assertion about compressed UHJ on stereo mobile devices. I dunno about full UHJ surround decode cos there don't seem to be any good ones in the public domain. PS I expect you to hear ve.eery slight differenes with one MP3 and probably none with the other. I won't insist on Double Blind bla bla but you might find that educational. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound _ The University of Derby has a published policy regarding email and reserves the right to monitor email traffic. If you believe this email was sent to you in error, please notify the sender and delete this email. Please direct any concerns to info...@derby.ac.uk. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
From: Richard Lee rica...@justnet.com.au To: 'Surround Sound discussion group' sursound@music.vt.edu Sent: Tuesday, 30 October 2012, 19:51 Subject: Re: [Sursound] Patent application: Data structure for HOA Unless things have changed a lot, last I checked lossy compression messes up phase relationships, and that would be an issue for things like UHJ, which as long as portable stereo players with limited battery life (and thus limited CPUs), is the only viable, because stereo compatible, distribution format. At this point in time, not only is most music listened on mobile devices, most music is even purchased on mobile devices, and that's strictly a stereo (or maybe binaural) world. Try this simple experiment. Take your favourite Nimbus UHJ CD and rip it using the most evil MP3 encoder you can find .. probably the one built into the latest Windoz Media Player. Do this at 256kB/s and also (shock! horror!) at 128kB/s. Now listen to the resultant files on a mobile device. Then you can pontificate to us on how the musicality has all escaped and no one is going to find these acceptable. You can also rip to a WAV file if your mobile device will play these and compare the MP3s with the 'original'. This is just testing Ronald's assertion about compressed UHJ on stereo mobile devices. I dunno about full UHJ surround decode cos there don't seem to be any good ones in the public domain. PS I expect you to hear ve.eery slight differenes with one MP3 and probably none with the other. I won't insist on Double Blind bla bla but you might find that educational. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound -- next part -- An HTML attachment was scrubbed... URL: https://mail.music.vt.edu/mailman/private/sursound/attachments/20121030/2451f92c/attachment.html ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
Richard Lee wrote: Take your favourite Nimbus UHJ CD and rip it using the most evil MP3 encoder you can find Sorry, slightly off-topic, but still: Some people have done terrible data reduction to UHJ recordings already: http://www.youtube.com/watch?v=FnvRtM5WDsc http://www.youtube.com/watch?v=dDo3Hn35xEo http://www.youtube.com/watch?v=cfrsjU05S8A http://www.youtube.com/watch?v=q4NXILz29dE http://www.youtube.com/watch?v=8S4ozLfK2zU Having said that, back in the years I tested between different data reduction systems and UHJ and Dolby Surround. UHJ is much more robust against data reduction than Dolby Surround, as many data reduction algorithms wipe out all such content that has large phase differences betwen the stereo channels. Thus Dolby Surround recordings tend to turn into plain mono for example in DAB transmissions that use low bitrates. Eero ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Ambisonic 'File' Formats
There's also the possibility of a new version of Broadcast Wav (BWF). Dave Thanks Dave, URL and/or other reference ??? 'Native' CAF also has the possibility for 'W,X,Y,Z' so (again without acknowledgment) I suspect that could be counted as a *.amb variant. MPEG formats: anyone have the key URLs ? Technicolor ... started this round. Anymore ? Michael On 30 October 2012 09:40, Michael Chapman s...@mchapman.com wrote: The thing that is really annoying, really frustrating, is that the community that I am trying to serve (which includes me) keeps shooting itself in the foot, incessantly, year after year, by arguing over the same details, over and over and over. ED, Sunday. At the risk of another abortive cycle. We seem to have: Richard Dobson's *.amb Widely used, widely accepted. Problematic as order number increases. Has the underlying *.wav problems (and advantages). Etienne Deleflie's 'UA' I had thought this had been dropped in favour of the fourth of these, so have rather taken my 'eye off the ball'. Perhaps Etienne could comment (he is one of the author's of the fourth). The Graz Proposal of 2009. This fell on stoney ground ;-( It was CAF based, and we did have promises of a 'CafPak' from the 'WavPak' development team. http://mchapman.com/amb/reprints/AFF The Kenticky Proposal of 2010. The current situation is: a library available at http://iem.git.sourceforge.net/git/gitweb.cgi?p=iem/ambix;a=summary and the unofficial documentation at http://iem.at/~zmoelnig/libambix/ambix_8h.html I'll happily put up a webpage with URLs / links to them all (and any others). Even try a Wikipedia type table of 'what does/offers what' (if the authors will assist). (And genericlly these re not 'file format's so much as 'interchange formats' (files, streams, ... .) Once we've got the facts straight perhaps we could recommence on a solid basis ?? __ [ . . . ] ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Ambisonic 'File' Formats
On Tue, Oct 30, 2012 at 9:01 AM, Michael Chapman s...@mchapman.com wrote: 'Native' CAF also has the possibility for 'W,X,Y,Z' so (again without acknowledgment) I suspect that could be counted as a *.amb variant. Has anyone had any success with this? It's in the Core Audio header files, but a couple of years ago, I tried writing out some B-format files and got an 'unimplemented' error. (I forget the exact error, but could recreate it if anyone is interested). Also, I've encoded 4-channel B-format files with Ogg Vorbis and MPEG-4 AAC at rates around 160-256 kbps, and wavpack lossy (which is around 350kbps, iirc) and they decoded correctly without spatial artifacts, so at moderate bit rates these codecs preserve phase information. At very low bit rates, Ogg Vorbis switches to square polar mapping mode which will corrupt the phase relationships in stereo signals. I don't know how they handle multichannel files. See http://xiph.org/vorbis/doc/stereo.html -- Aaron (hel...@ai.sri.com) Menlo Park, CA US ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Ambisonic 'File' Formats
Whatever it is, if it's a 32-bit file format, it should be considered deprecated. CAF is at least explicitly designed to be a 64-bit format. It's also designed such that during recording a crash of the app can leave a recoverable file behind by the way the header structures are designed, much like in the old days SDII files had that property and were thus a preferred choice of many in the audio field. https://developer.apple.com/library/mac/documentation/MusicAudio/Reference/CAFSpec/CAFSpec.pdf Any compression used should prioritize fast and CPU efficient decompression over max. compression ratio, because a file is compressed once, but decompressed many times, often on mobile devices. As such, an open standard that is implemented by many hardware devices would be rather useful. Apple Lossless comes to mind, which is fully documented and available under the Apache 2.0 license: http://alac.macosforge.org/ Ronald On 30 Oct 2012, at 12:01, Michael Chapman s...@mchapman.com wrote: There's also the possibility of a new version of Broadcast Wav (BWF). Dave Thanks Dave, URL and/or other reference ??? 'Native' CAF also has the possibility for 'W,X,Y,Z' so (again without acknowledgment) I suspect that could be counted as a *.amb variant. MPEG formats: anyone have the key URLs ? Technicolor ... started this round. Anymore ? Michael On 30 October 2012 09:40, Michael Chapman s...@mchapman.com wrote: The thing that is really annoying, really frustrating, is that the community that I am trying to serve (which includes me) keeps shooting itself in the foot, incessantly, year after year, by arguing over the same details, over and over and over. ED, Sunday. At the risk of another abortive cycle. We seem to have: Richard Dobson's *.amb Widely used, widely accepted. Problematic as order number increases. Has the underlying *.wav problems (and advantages). Etienne Deleflie's 'UA' I had thought this had been dropped in favour of the fourth of these, so have rather taken my 'eye off the ball'. Perhaps Etienne could comment (he is one of the author's of the fourth). The Graz Proposal of 2009. This fell on stoney ground ;-( It was CAF based, and we did have promises of a 'CafPak' from the 'WavPak' development team. http://mchapman.com/amb/reprints/AFF The Kenticky Proposal of 2010. The current situation is: a library available at http://iem.git.sourceforge.net/git/gitweb.cgi?p=iem/ambix;a=summary and the unofficial documentation at http://iem.at/~zmoelnig/libambix/ambix_8h.html I'll happily put up a webpage with URLs / links to them all (and any others). Even try a Wikipedia type table of 'what does/offers what' (if the authors will assist). (And genericlly these re not 'file format's so much as 'interchange formats' (files, streams, ... .) Once we've got the facts straight perhaps we could recommence on a solid basis ?? __ [ . . . ] ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
On 29 Oct 2012, at 20:42, Stefan Schreiber st...@mail.telepac.pt wrote: Ronald, most if not all (classical) recordings where I am participating are done in a way that they could be issued in 5.1 (or say 5.0) surround, namely several Pentatone recordings, and even the more recent television/radio stuff. I would guess that every good orchestra recording is done in this way (which means could be issued in 2.0/stereo, 5.1 or other formats). Yes, that may be true, but the vast majority of these recordings is done in a way of the one microphone per speaker mentality, not in the let's record a sound field mentality. So while it may be surround, it's not Ambisonics, and it's the latter that I'm interested in, and not some hare-brained system like traditional 5.1 recording. The only interest I have in 5.1 is as a delivery format for G-format predecoded Ambisonics. But 5.1 outside of movies pretty much is dead in the water for music, a handful of boutique recordings aside, which really don't matter. What we need is a catalog in the 10s or 100s of thousands of recordings, and recordings with the artists the general public wants to hear, not a few boutique recordings that a few surround sound fanatics are interested in mostly due to the fact that they are surround recordings, not because they crave the music and artists who were recorded. These boutique recordings are more or less nothing but technology demonstrations, and thus are mostly only of technical interest. My hint to Dolby Surround was ironic (as many guys on this list oppose anything from Dolby), but you have to admit that there exist many (matrixed) Dolby surround mixes for film use. (And also and very obviously discrete 5.1 surround mixes, which are superior.) I don't oppose anything from Dolby, but I dislike the company because they have more than once sunk good technology because it didn't fit their specific business interests, and have thus been a major roadblock for progress. If they were to pick up the baton and would advocate the right changes, I'd be all for them. Likely that would only happen if they could hold a ton of key patents and charge everyone massive licensing fees for them; otherwise they don't seem to be interested. They rather go and invent an octagon wheel, patent it and use their influence to peddle it, than use the round wheel they can't charge royalties for. UHJ works, but it is also a matrixed format and arguably not a complete surround format, because 2 cannels are not enough. (I would say 5.1 is better, this doesn't seem to be an opinion.) Secondly, the UHJ system should have some issues even in stereo, because of the matrix. Of course, 5.1 (as G-format) would be better than UHJ, but 5.1 isn't widely used for music, while stereo is. So unless that changes, we can either ship RIGHT NOW UHJ into the stereo music channel, or we can bitch and whine that there is no surround recordings on the market, because there's no proper distribution channel for the format that would be ideal. My approach is: use what's available. If it's available, more and more people have an opportunity to discover what good surround sound that's more than an SFX button on a receiver can do, and with that demand can build up. The more demand, the bigger the catalog, the bigger the catalog, the more commercial interest to make things better, i.e. to eventually provide a better delivery format than a stereo container. That's what I mean with baby steps. Start with what's available now, instead of waiting for the glorious future that never comes, because people try to skip a few steps at the beginning. Also, UHJ opens the door for guerrilla tactics, i.e. sound engineers with a passion for surround can make UHJ mixes for people who ask for a stereo mix, because UHJ is stereo compatible. So surround mixes can slide into popular items without explicitly being asked for by the producers or artists. If they like the mix in stereo, they won't care/notice that it's actually UHJ. Write to Apple that they should publish 5.1 (and maybe .AMB files etc.), and forget about old compromises. No, because there's no interest in pushing something without perceived demand, particularly if it's something that's too complicated to explain to a non-technical audience in a sound bite. (You can continue to promote UHJ, but I am sure this won't fly because you say people ideally would have to record via soundfield mics. If you mix a UHJ recording from spot mics, you also could mix to 5.1 ...) UHJ, 5.1 are delivery formats. What matters is the recording and mixing technique. If the 5.1 mix is done with an ambisonic panner, then the resulting product is G-Format, and thus acceptable. If it's done with pan-potting, it's an abomination, or if one's friendly, just an SFX, but certainly not proper surround sound. Frankly, who cares about the 3 dozen high-end surround recordings being made? This is
Re: [Sursound] Patent application: Data structure for HOA
On 30 Oct 2012, at 06:24, Peter Lennox p.len...@derby.ac.uk wrote: Am I missing something? - for mobile use, wouldn't B-format to binaural be better than UHJ? Dr Peter Lennox Of course it would. Do you know of a mobile playback device with multi-channel audio support, multi-channel audio market place, and a binaural decoder? Lacking that, putting UHJ encoded stereo into iTunes, Amazon, CD-Baby, etc. is easy. And an audio playback app with UHJ-to-binaural is easy to place in to the Apple/Android app stores. It's not about technical superiority, but about what can be done in the main stream market place. I'm not interested in lab solutions and technology demonstrations, I'm interested in what works for millions of iOS/Android users RIGHT NOW. Ronald ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Patent application: Data structure for HOA
On 29 Oct 2012, at 20:56, Stefan Schreiber st...@mail.telepac.pt wrote: Oh yes, go to Apple and look if they listen to your ideas, and let others do their stuff instead of doing some promotion for some stylish, fahionable campany offering super slim products. You make my point: they won't listen, at least not at this point in the game. But that's exactly why it's futile to advocate lab type stuff, and stick with what's available to the mass market now. The mass market right now offers exactly two things: a) stereo audio b) custom apps though the various app stores. And that means that any mass market solution must at this point be stereo compatible and may employ a custom playback app made available through the various app store channels. (Samsung and Amazon sell also a lot of smartphones and tablets, by the way. If it i just about numbers, Samsungs sells actually more mobile phones...) I use Apple as an example, because it's the dominant company in this field, while the rest are imitators and followers; not leaders. Who sells more devices also doesn't matter, what matters which devices are used. And if you e.g. look at the web traffic statistics you'll see how clearly Apple dominates that field. Apple is also the company with the bargaining power. So if they see surround as the future, they can make that future happen. Therefore, getting surround sound into their platform by means of a Trojan horse (like e.g. putting UHJ-encoded material into the iTunes store) is a start on that path. I am really angry about these postings. Look for surround in your local Apple store, and if you find somen give us some news about. Otherwise, Apple and their fashionable products are offtopic. (I don't see any relationship to this thread, and even not to this audio list.) You're angry at reality. I'm not making these things up, nor do they constitute my ideal world. But I'm willing to face the reality and ask which small steps can we take to get from here to there by infiltrating what actual consumers use, rather than being preoccupied with lab experiments and boutique recordings that cater to a bunch of enthusiasts. Nobody who matters (i.e. average consumer) is interested in a dorky head-tracking headphone setup that makes him/her look like a Borg from Star Trek. Headphones are accessories that need to be fashionable, because people know they are going to be seen in public wearing them. That's reality. Get used to it. That's why stuff like Beats by Dr. Dre sells (cool DJs have them) and nobody would want to be caught dead wearing top-notch studio head phones. Ronald ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound