tag:blogger.com,1999:blog-69161684703769374252024-03-13T12:57:26.568-07:00Earning My TurnsSki, music, books, science, technologyFernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.comBlogger422125tag:blogger.com,1999:blog-6916168470376937425.post-27269254598304434702023-06-03T15:47:00.004-07:002023-06-03T15:47:46.423-07:00Books...<p> It's quite possible that blogging on an old-style blogging site is totally obsolete, but for a reason I got to look at some of my 2009 posts here, and I didn't feel too embarrassed by what I read. Moving to another more fashionable platform would be too much work. Concurrently, I've been asked a few times recently for book recommendations. So...</p><p></p><ol style="text-align: left;"><li><i><a href="https://nick-lane.net/books/transformer-the-deep-chemistry-of-life-and-death/">Transformer</a></i> by Nick Lane. No, not <i>that</i> kind of transformer. Three books have reset my understanding of biology and overcame my boredom with the subject in high-school. In chronological order: Jacques Monod's <i><a href="https://fr.wikipedia.org/wiki/Le_Hasard_et_la_N%C3%A9cessit%C3%A9">Le Hasard et la Nécessité</a>; <a href="https://www.degruyter.com/document/doi/10.12987/9780300128673/html?lang=en">The Plausibility of Life</a></i> by Marc Kirschner and John Gerhart; and now Nick Lane's <i>Transformer</i>. Lane's focus on self-sustaining biochemical cycles as drivers of evolution, with genes as combinatorially generated and selected catalysts, rather than the canonical view of genes in the driver's seat, is refreshing and illuminating. Not an easy read, especially given the sadly decayed state of complex (biochemical pathway) schematics in trade books, but so much worth it. There's a <a href="https://www.preposterousuniverse.com/podcast/2022/05/23/198-nick-lane-on-powering-biology/">good preview</a> in Sean Carroll's <i>Mindscape</i> podcast.</li><li><i><a href="https://mitpress.mit.edu/9780262047005/the-evolution-of-agency/">The Evolution of Agency</a></i> by Michael Tomasello. Another <i>Mindscape</i> podcast <a href="https://www.preposterousuniverse.com/podcast/2023/01/30/225-michael-tomasello-on-the-social-origins-of-cognition-and-agency/">recommendation</a> (can you detect a trend?) A bit thin and abstract, and maybe too flowchart-y in its explanation of different levels of agency from lizards to primates, but it re-centers self-monitoring and social cognition when (AI) fashion favors monolithic feed-forward prediction. It encourages, actually it forces, (re)thinking assumptions. It forms a worthy provocative triple with <i><a href="https://www.hup.harvard.edu/catalog.php?isbn=9780674237827">The Enigma of Reason</a></i> and <a href="https://global.oup.com/academic/product/the-instruction-of-imagination-9780190256623?cc=us&lang=en&" style="font-style: italic;">The Instruction of Imagination</a> by suggesting a plausible evolutionary path to our symbolic technology culture. </li><li><i><a href="https://www.antofagica.com.br/produto/memorias-postumas-de-bras-cubas/">Memórias Póstumas de Brás Cubas</a></i> by Machado de Assis. Walking in old Lisboa recently, I found a handsome new illustrated edition of this Portuguese language classic in a bookstore window. Savoring this modernist well-before-modernism masterpiece, its sharp humor and meta-narrative dislocations scratching out a bleeding picture of the racial oppression and bourgeois aristocratic pretensions that imperial Brasil inherited from colonial Portugal. The edition's <a href="https://en.wikipedia.org/wiki/Candido_Portinari">Portinari</a> etchings don't pull any punches. There are English translations, of course, but I don't know which one to recommend. </li></ol><p></p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com2tag:blogger.com,1999:blog-6916168470376937425.post-67995713518139281932022-11-27T21:20:00.003-08:002022-11-27T21:23:10.616-08:002022 (so far) creative music releases to (re)listen to<div style="text-align: left;">I can't face doing a top 10 list, I'm already skipping a lot of great music. Ordered by release date:</div><p></p><ol style="text-align: left;"><li><a href="https://577records.bandcamp.com/album/pacifica-koral-reef" target="_blank">Pacifica Koral Reef</a></li><li><a href="https://emileparisien.bandcamp.com/album/louise" target="_blank">Louise</a></li><li><a href="https://sluchaj.bandcamp.com/album/flow-of-everything" target="_blank">Flow of Everything</a><br /></li><li><a href="https://gordongrdina.bandcamp.com/album/nights-quietest-hour" target="_blank">Night's Quietest Hour</a><br /></li><li><a href="https://martasanchez-whirlwind.bandcamp.com/album/saam-spanish-american-art-museum" target="_blank">SAAM (Spanish American Art Museum)</a><br /></li><li><a href="https://jazzandpeople.bandcamp.com/album/romain-pilon-falling-grace" target="_blank">Falling Grace</a><br /></li><li><a href="https://editionrecords.com/releases/mario-laginha-jangada/" target="_blank">Jangada</a></li><li><a href="https://michaelleonhart.bandcamp.com/album/the-normyn-suites" target="_blank">The Normyn Suites</a></li><li><a href="https://www.ecmrecords.com/shop/1637149136/return-from-the-stars-mark-turner-jason-palmer-joe-martin-jonathan-pinson" target="_blank">Return From the Stars</a></li><li><a href="https://fergusmccreadie.bandcamp.com/album/forest-floor" target="_blank">Forest Floor</a></li><li><a href="https://tumrecords.com/059-2-blues-for-cecil" target="_blank">2 Blues for Cecil</a></li><li><a href="https://milesokazaki.bandcamp.com/album/thisness" target="_blank">Thisness</a></li><li><a href="https://florianarbenz.bandcamp.com/album/conversation-5-elemental" target="_blank">Conversation #5 — elemental</a></li><li><a href="https://maryhalvorson.bandcamp.com/album/amaryllis" target="_blank">Amaryllis</a></li><li><a href="https://lucianban.bandcamp.com/track/ways-of-disappearing" target="_blank">Ways of Disappearing</a></li><li><a href="https://davidvirelles.bandcamp.com/album/nuna" target="_blank">Nuna</a></li><li><a href="https://store.bluenote.com/products/nduduzo-makhathini-in-the-spirit-of-ntu" target="_blank">In the Spirit of Ntu</a></li><li><a href="https://tyshawn-sorey.bandcamp.com/album/mesmerism" target="_blank">Mesmerism</a></li><li><a href="https://taoforms.bandcamp.com/album/gravity-without-airs" target="_blank">Gravity Without Airs</a></li><li><a href="https://sunnysiderecords.bandcamp.com/album/nearness" target="_blank">Nearness</a></li><li><a href="https://billydrummond.bandcamp.com/album/valse-sinistre" target="_blank">Valse Sinistre</a></li><li><a href="https://miguelzenon.bandcamp.com/album/m-sica-de-las-am-ricas" target="_blank">Música de las Américas</a></li><li><a href="https://lecoqrecords.bandcamp.com/album/the-jon-cowherd-trio-pride-joy" target="_blank">Pride & Joy</a><br /></li><li><a href="https://jazztrail.net/blog/jd-allen-americana-vol-2-album-review" target="_blank">Americana, Vol. 2</a></li><li><a href="https://jamesbrandonlewis.bandcamp.com/album/msm-molecular-systematic-music-live-24bit-hi-res-48khz" target="_blank">MSM Molecular Systematic Music</a></li><li><a href="https://michaelwollny.bandcamp.com/album/ghosts" target="_blank">Ghosts</a></li><li><a href="https://taoforms.bandcamp.com/album/root-perspectives" target="_blank">Root Perspectives</a></li><li><a href="https://eremiterecords.bandcamp.com/album/mondays-at-the-enfield-tennis-academy" target="_blank">Mondays at the Enfield Tennis Academy</a></li><li><a href="https://jakobbro.com/web/album/2575-2/" target="_blank">Once Around the Room</a></li><li><a href="https://danweisstrio.bandcamp.com/album/dedication" target="_blank">Dedication</a></li><li><a href="https://store.bluenote.com/products/bill-frisell-four" target="_blank">Four</a></li><li><a href="https://patriciabrennanpyroclastic.bandcamp.com/album/more-touch" target="_blank">More Touch</a></li></ol><p></p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-35286277749549638892020-05-09T22:23:00.000-07:002020-05-10T19:26:54.815-07:00In memoriam Pedro CarvalhoPedro was my wife Ana's kid brother. He was 57 years old when he died in Lisbon, early Saturday May 9th 2020 from COVID-19. He had been in the ICU at Santa Maria Hospital since early March after being infected in late February on a business trip to the North of Portugal. He succumbed to a catastrophic lung hemorrhage that his outstanding medical team was powerless to stop after so much effort trying everything that had been learned from other COVID-19 patients all over the world.<br />
<br />
In a painful turn of fate, Pedro was the most prepared of us for disaster. He was a volunteer with the Santarém fire brigade from his early teens. He was also fascinated by radio from early on, and became a well-known radio amateur (call sign CT1DBS). He continued for the rest of his life his dual passions for helping others in danger, and for telecommunications: as a firefighter, as a scout, as a Portuguese Navy officer, as commander of Santarém's fire brigade, as the director of civil protection and firefighting for the Portuguese autonomous Açores region, and as an entrepreneur developing advanced emergency communication systems for Portugal. <br />
<br />
Pedro never took the easy path. He repeatedly fought for the right actions against bureaucratic foot dragging, lack of accountability, and misplaced priorities. He pushed politicians to support better training for firefighters and other rescue personnel, for creating warning systems for populations in danger from Portugal's terrible forest fires, and for teaching children how to protect themselves from earthquakes in the very seismic Açores. For his leadership and rectitude, he lost some of the jobs he loved, but he earned the profound admiration and gratitude of the marines and firefighters he led.<br />
<br />
It is maybe a consolation that Portugal's response to COVID-19 has in the main followed Pedro's principles of leadership, technical excellence, and attention to everyone and everything in the toughest situations. It was not enough to help him, who was one of the first two critical COVID-19 patients in Lisbon, but we feel his spirit of service and competence in how many others have been protected.<br />
<br />
Pedro leaves behind Clara, his beloved wife of 25 years, his brother João and his sister Ana, several nieces and nephews, and a big hole in all the work he would still be doing in advancing emergency technology in Portugal. We miss him terribly.<br />
<br />
COVID-19 is real, it can easily take the most vital, helpful, and productive among us.<br />
<br />
<a href="https://www.rederegional.com/sociedades/29370-antigos-colegas-na-ultima-homenagem-ao-ex-comandante-pedro-carvalho">https://www.rederegional.com/sociedades/29370-antigos-colegas-na-ultima-homenagem-ao-ex-comandante-pedro-carvalho</a>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com3tag:blogger.com,1999:blog-6916168470376937425.post-30807128072345321492019-11-02T20:28:00.000-07:002019-11-03T12:53:13.581-08:00Digital audio without tearsIn earlier posts I described systems I was using for listening to lossless (FLAC) digital audio. A lot has changed in my various audio setups that would take too long to explain, but I wanted to update you on a really nice lowish-cost transport for DACs and integrated audio devices with S/PDIF or AES digital inputs that I strongly recommend. The ingredients (most used in a previous setup):<br />
<br />
<ol>
<li>Pi 3 SBC</li>
<li><a href="https://www.raspberrypi.org/products/raspberry-pi-touch-display/">PiTouch 7" touchscreen</a></li>
<li><a href="https://smarticase.com/products/smartipi-touch">SmartiPi stand</a></li>
<li><a href="http://www.pi2design.com/pi2aes.html">Pi2AES audio shield</a></li>
<li><a href="http://www.pi2design.com/store/p21/PI2AES_-_EXTERNAL_40W%2C_24V_POWER_SUPPLY.html">Power supply for Pi2AES</a></li>
<li>Pi 2 Design supplies (5) with a straight barrel adapter that is too long for the stand (3), but <a href="https://smile.amazon.com/CGTime-5-5...2.1mm+female+2.5mm+male&qid=1572444556&sr=8-7">this elbow adapter</a> fits (thanks to Michael Kelly @ Pi 2 Design).</li>
<li>Whatever miniUSB 5V power supply you have around to power the PiTouch (I'm using an otherwise idle Chromecast power supply)</li>
<li><a href="https://moodeaudio.org/">moOde Audio</a> Linux-based digital music software</li>
</ol>
I set this up on my home network, which allowed me to configure moOde through a Web browser on my laptop, but I run it at work disconnected from the network, with all the music on an SSD. If your system is always networked, I'd skip (2), (3), and (6) and buy the <a href="http://www.pi2design.com/store/p20/PI2AES_-_ACRYLIC_CASE.html">nice Pi2AES case</a> from Pi 2 Design instead. The same hardware can be used with other Linux-based audio distributions. I used to favor <a href="https://www.picoreplayer.org/">piCorePlayer</a>, and I still prefer its lean, fast UI, but it has become unusable for non-networked systems, unfortunately.<br />
<br />
For those of you who have USB DACs or integrated amps, I've heard good things about the <a href="https://allo.com/sparky/usbridge-signature-pcb.html">Allo USBridge Signature</a>, which can also be bought as a complete box with a Pi and pre-installed software.<br />
<br />
I've used quite a few more expensive transports in different configurations, with different DACs and downstream audio chains. Pi2AES is definitely competitive with many times more expensive transports, if you are willing to do the hardware and software legwork, and I suspect Allo USBridge Signature will be too.<br />
<br />
Pictures of the Pi2AES-based transport feeding a nice <a href="http://www.soekris.dk/dac1541.html">Soekris dac1541</a> sign-magnitude R-2R DAC from Denmark followed by the now discontinued HP-1 headphone amp from <a href="https://neurochrome.com/">Neurochrome</a> in Canada.<br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjQZIG5ADsReTooF5BjuySiuXwEViWUmuJ1s_n6TYY2JewMvZzELZhFRgziwN-hrp49toGA8FMnuxfc_5l8xz0NdprccfMbE3mzLn60dX9bvpn9aHCSN8NDLXuJOZbtbn3_dIH9AiXrug/s1600/IMG_20191101_161007.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="1600" data-original-width="1200" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjQZIG5ADsReTooF5BjuySiuXwEViWUmuJ1s_n6TYY2JewMvZzELZhFRgziwN-hrp49toGA8FMnuxfc_5l8xz0NdprccfMbE3mzLn60dX9bvpn9aHCSN8NDLXuJOZbtbn3_dIH9AiXrug/s320/IMG_20191101_161007.jpg" width="240" /></a></div>
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi2F1mRlAXv44EoxP0r-KxO-2VqAfczfXm8ReeU1Xixt-vcqbSCKxiwUZMzsYuUjFuJUcondJ4oM5R0aSZNivC1PO4N89CHCvZvBkies74-UM-BFdDWik2JgT0-bmg78bSE7gDWh7K6hg/s1600/IMG_20191101_161346.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="1600" data-original-width="1200" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi2F1mRlAXv44EoxP0r-KxO-2VqAfczfXm8ReeU1Xixt-vcqbSCKxiwUZMzsYuUjFuJUcondJ4oM5R0aSZNivC1PO4N89CHCvZvBkies74-UM-BFdDWik2JgT0-bmg78bSE7gDWh7K6hg/s320/IMG_20191101_161346.jpg" width="240" /></a></div>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com4tag:blogger.com,1999:blog-6916168470376937425.post-63697080099913082262019-05-28T21:35:00.000-07:002019-11-02T20:32:39.416-07:00Digital audio findingsIt's a long story that I won't give in detail, but 6 years ago I started to explore digital lossless music, initially CD rips but increasingly digital downloads. My music interests are mainly contemporary jazz, modern classical, early music, Baroque, West Africa, Mediterranean, Middle East. I went through a lot of gear and setups since. I learned a lot, too much to summarize, but here are a few findings:<br />
<br />
<ul>
<li>First question is why lossless rather than MP3 or AAC? Simply, it can sound better for some material (plucked strings especially) with the right downstream gear.</li>
<li>How do you get music into that form? If you have CDs, you re-rip them. I use <a href="https://tmkk.undo.jp/xld/index_e.html">XLD</a> on my Mac. If you want to stream, you can use lossless Tidal or <a href="https://www.qobuz.com/us-en/discover">Qobuz</a> (which is what I use). Finally, there are several lossless download vendors: <a href="https://www.prostudiomasters.com/">ProStudioMasters</a>, <a href="https://www.hdtracks.com/">HDtracks</a>, <a href="https://us.7digital.com/">7digital</a>, <a href="https://www.eclassical.com/">eClassical</a>, <a href="https://www.prestomusic.com/classical">Presto Classical</a>, and best of all because more money goes to the artists, <a href="https://bandcamp.com/">Bandcamp</a> for independent artists and labels. Most of my recent music is from Bandcamp.</li>
<li>How do you play it? You need a means to move bits from local storage or streaming sources to whatever turns bits into analog waveforms in your system. Many options there, from simple Apple or Sonos setups to integrated receivers from companies like NAD, Naim, or Hegel (owned gear from all of them), to dedicated <i>digital transports</i> connected to separate DACs, preamps, and amps.</li>
<li>You can use your phone or laptop as a transport, but it is also possible to use a variety of single-purpose devices for that. <a href="https://www.earningmyturns.org/2018/06/building-small-standalone-digital-music.html">Here</a>'s what I assembled to use at work. Again, one can go crazy on digital transports, to stratospheric prices that make no sense to me. I do own a relatively pricey <a href="https://metrumacoustics.com/roon/">Metrum Ambre</a> transport for my fancy headphone system(s) at home, but one can do well in the just-digital-audio transport front for much less, such as the <a href="https://www.allo.com/sparky/usbridge.html">Allo USBridge</a> I have for my speaker system.</li>
<li>For quite a while, I was having a very good time with a single-purpose <a href="https://www.solid-run.com/nxp-family/cubox-i/">CuBox</a> running <a href="https://volumio.org/">Volumio</a>, reading tracks from a home NAS (Synology) playing to a <a href="https://www.schiit.com/products/bifrost">Schiit Bifrost Multibit</a> DAC and a <a href="https://www.schiit.com/products/jotunheim">Schiit Jotunheim</a> headphone amp for my home headphone and then work systems until I gave it to my daughter. People I trust claim that you can spend a lot less, either from Schiit or from vendors like <a href="https://www.audioquest.com/dacs/dragonfly/dragonfly-black">AudioQuest</a> for a very reasonable sound quality. The AudioQuest devices are nicely portable and work easily off your laptop or phone.</li>
<li><a href="https://www.schiit.com/">Schiit Audio</a> has by far the best quality/price ratio that I know of for DACs and headphone amps. I've never tried their preamps, and speaker amps, but I suspect they win there too. They are also responsive, don't take themselves too seriously, and respond well to nice customer inquiries.</li>
<li>If you share living space with someone, or if you have close neighbors, headphone listening has got really good with the right headphones. I have only listened closely to a small number of different headphones, mostly rather expensive (>$1K), so I have no recommendations for anything below $500. At work, I use <a href="https://www.mrspeakers.com/headphones-22/aeon-closed-back-headphone.html">MrSpeakers Æon Closed</a>, which work well for the sources and relatively laidback music I like for work.</li>
<li>It is hard for me to recommend speaker systems because I've heard way too few different ones in the last decade. I can say that for a small(ish) room, a pair of <a href="https://us.kef.com/speakers/flagship-hi-fi-speakers/ls50.html">KEF LS50</a>s with a sufficiently brawny amp (the LS50s are rather inefficient) give outstanding clarity and dynamics (LS50s are now mostly sold in a wireless, built-in amp version that I've not heard so I can't comment on). In my current larger living room, my earlier LS50s were replaced by <a href="https://us.kef.com/reference-1-ultimate-bookshelf-speaker-pair.html">KEF Reference 1</a>s on stands, with a really brawny <a href="https://www.hegel.com/products/integrated/h360">Hegel integrated amp</a>, but those choices involved a bunch of other factors as well as a lot of listening at a couple of SF dealers. The Hegel integrated can take bits directly into its built-in DAC, which is fine but a bit too polite for my taste. Instead, I use a <a href="https://www.schiit.com/products/yggdrasil">Schiit Yggdrasil</a> DAC with some experimental hardware (don't ask) that allows it to get outstanding quality from a cheap <a href="https://www.allo.com/sparky/usbridge.html">Allo USBridge</a> transport.</li>
<li>How do I get bits to my home systems? I've landed on <a href="https://roonlabs.com/">Roon</a>, which does the best job among alternatives to manage my eclectic local and streamed music collection and send the bits reliably to the two systems I listen on at home (speakers in the living room, headphones in the study). I run the Roon server on a fanless Intel NUC with Ubuntu Server 18.04, mounting its music from a Synology NAS.</li>
<li>You don't really want to go there: <div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEisN5e-rCwfcacZ_lL06LTNxNqh9nZ2Ho8bDuIfasJVFwXKrS3Zh3IXquiRCsIELQeb8w6tRalTIGKSDrgOknUN1NEKgEOjaYwM01PMrxDRBrIcXxx1gbu8aDk2wjUEg4fcQuuNdtADrQ/s1600/audio.001.png" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="768" data-original-width="1024" height="300" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEisN5e-rCwfcacZ_lL06LTNxNqh9nZ2Ho8bDuIfasJVFwXKrS3Zh3IXquiRCsIELQeb8w6tRalTIGKSDrgOknUN1NEKgEOjaYwM01PMrxDRBrIcXxx1gbu8aDk2wjUEg4fcQuuNdtADrQ/s400/audio.001.png" width="400" /></a></div>
</li>
</ul>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-83671104543960371432019-05-27T18:18:00.000-07:002019-05-27T18:21:10.257-07:00Dennis Bray's "Wetware"Just finished reading <i><a href="https://yalebooks.yale.edu/book/9780300167849/wetware">Wetware</a></i>, which I had owned for a while and started reading a couple times while traveling but misplaced both times, so I had to go back to the beginning. I don't agree with everything Bray says, and in the 8 years since he published the book, a lot happened both in computing and in biology. But I recommend <i>Wetware</i> to anyone who is puzzled about the similarities and differences between living things and computers. Three quotes that connected:<br />
<blockquote class="tr_bq">
<span style="background-color: white; color: #202124; font-family: "roboto" , "arial" , sans-serif; font-size: 16px; letter-spacing: 0.1px; white-space: pre-wrap;"><i>Neural networks can cope with large amounts of information, supplied as written text, facial features, industrial processes, and so on. They can recognize restricted sets of patterns better than we can. But what they do is the tapping of a tin drum compared with the symphony orchestra of natural environments.</i></span></blockquote>
<blockquote class="tr_bq">
<span style="background-color: white; color: #202124; font-family: "roboto" , "arial" , sans-serif; font-size: 16px; letter-spacing: 0.1px; white-space: pre-wrap;"><i>Such coupling of environmental parameters through internal circuits allows a cell to predict future events. These circuits contain, implicitly, the probability of certain life-changing events.</i></span></blockquote>
<blockquote class="tr_bq">
<i><span style="background-color: white; color: #202124; font-family: "roboto" , "arial" , sans-serif; font-size: 16px; letter-spacing: 0.1px; white-space: pre-wrap;">If the detailed chemistry of the cell is simply the outcome of a historical ragbag of ad hoc interactions, then it will be no more predictable than the weather.</span> </i></blockquote>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-38454334505407991362018-09-18T20:41:00.000-07:002018-09-18T20:46:08.786-07:00August-September music<table border="0" cellspacing="0">
<colgroup span="2" width="400"></colgroup>
<tbody>
<tr>
<td align="left" height="17"><b>Artist</b></td>
<td align="left"><b>Title</b></td>
</tr>
<tr>
<td align="left" height="17">Alisa Weilerstein</td>
<td align="left">Transfigured Night: Haydn and Schoenberg</td>
</tr>
<tr>
<td align="left" height="17">Andrew Staples/Pauline Cheviller/Esa-Pekka Salonen</td>
<td align="left">Stravinsky: Perséphone</td>
</tr>
<tr>
<td align="left" height="17">Blaze/Sakurada/Sampson/ Wörner</td>
<td align="left">J.S. Bach – Secular Cantatas, Vol.10 (BWV 30a, 204)</td>
</tr>
<tr>
<td align="left" height="17">Cécile McLorin Salvant</td>
<td align="left">The Window</td>
</tr>
<tr>
<td align="left" height="17">Chris Lightcap</td>
<td align="left">Superette</td>
</tr>
<tr>
<td align="left" height="17">Christian Tetzlaff/Hannu Lintu</td>
<td align="left">Bartók: Violin Concertos Nos. 1, 2</td>
</tr>
<tr>
<td align="left" height="17">David Binney</td>
<td align="left">Out of Airplanes</td>
</tr>
<tr>
<td align="left" height="17">Edo de Waart</td>
<td align="left">Adams: Harmonium</td>
</tr>
<tr>
<td align="left" height="17">Ensemble Dialoghi</td>
<td align="left">Mozart, Beethoven: Quintets for Piano and Winds</td>
</tr>
<tr>
<td align="left" height="17">Ethan Iverson/Mark Turner</td>
<td align="left">Temporary Kings</td>
</tr>
<tr>
<td align="left" height="17">Gilad Hekselman</td>
<td align="left">Ask for Chaos</td>
</tr>
<tr>
<td align="left" height="17">Grdina/Houle/Loewen</td>
<td align="left">Live at the China Cloud</td>
</tr>
<tr>
<td align="left" height="17">Henry Kaiser/Wadada Leo Smith/Yo Miles!</td>
<td align="left">Yo Miles: Sky Garden</td>
</tr>
<tr>
<td align="left" height="17">Henry Kaiser/Wadada Leo Smith/Yo Miles!</td>
<td align="left">Yo Miles: Upriver</td>
</tr>
<tr>
<td align="left" height="17">Jamie Saft Quartet</td>
<td align="left">Blue Dream</td>
</tr>
<tr>
<td align="left" height="17">Jordi Savall/La Capella Reial de Catalunya/Tembembe Ensamble Continuo</td>
<td align="left">Bailar Cantando: Fiesta Mestiza en el Peru</td>
</tr>
<tr>
<td align="left" height="17">Lee Konitz/Dan Tepfer</td>
<td align="left">Decade</td>
</tr>
<tr>
<td align="left" height="17">Louis Thiry</td>
<td align="left">Messiaen: Organ Music</td>
</tr>
<tr>
<td align="left" height="17">Marcin Wasilewski Trio</td>
<td align="left">Live</td>
</tr>
<tr>
<td align="left" height="17">Mary Halvorson quartet</td>
<td align="left">Paimon: Book of Angels 32</td>
</tr>
<tr>
<td align="left" height="17">Mary Halvorson/Bill Frisell</td>
<td align="left">The Maid with the Flaxen Hair—A Tribute to Johnny Smith</td>
</tr>
<tr>
<td align="left" height="17">Matt Lavelle/Reggie Sylvester</td>
<td align="left">Retrograde</td>
</tr>
<tr>
<td align="left" height="17">Medeski Martin Wood/Alarm Will Sound</td>
<td align="left">Omnisphere</td>
</tr>
<tr>
<td align="left" height="17">Mikkel Ploug/Mark Turner</td>
<td align="left">Faroe</td>
</tr>
<tr>
<td align="left" height="17">Miles Okazaki</td>
<td align="left">Work (Complete, Volumes 1-6)</td>
</tr>
<tr>
<td align="left" height="17">Phronesis</td>
<td align="left">We Are All</td>
</tr>
<tr>
<td align="left" height="17">Quatuor Debussy/Jean-Philippe Collard-Neven/Vincent Peirani/Franck Tortillier</td>
<td align="left">Claude Debussy: ...et le jazz - Preludes for a Quartet</td>
</tr>
<tr>
<td align="left" height="17">Robert Kaddouch/Gary Peacock</td>
<td align="left">53rd Street</td>
</tr>
<tr>
<td align="left" height="17">Robert Kaddouch/Gary Peacock</td>
<td align="left">High Line</td>
</tr>
<tr>
<td align="left" height="17">Rosa Brunello/Los Fermentos</td>
<td align="left">Volverse: Live in Trieste</td>
</tr>
<tr>
<td align="left" height="17">Stile Antico</td>
<td align="left">Song of Songs</td>
</tr>
<tr>
<td align="left" height="17">Tom Barford</td>
<td align="left">Bloomer</td>
</tr>
<tr>
<td align="left" height="17">Tord Gustavsen Trio</td>
<td align="left">The Other Side</td>
</tr>
<tr>
<td align="left" height="17">Víkingur Ólafsson</td>
<td align="left">Johann Sebastian Bach</td>
</tr>
</tbody></table>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-12867847588915149292018-08-12T17:26:00.001-07:002018-08-12T17:26:42.743-07:00Latest month in music<table border="0" cellspacing="0">
<colgroup width="392"></colgroup>
<colgroup width="485"></colgroup>
<tbody>
<tr>
<td align="left" height="17"><b>Artist</b></td>
<td align="left"><b>Album</b></td>
</tr>
<tr>
<td align="left" height="17">Alisa Weilerstein</td>
<td align="left">Solo</td>
</tr>
<tr>
<td align="left" height="17">Anthony Braxton</td>
<td align="left">Willisau (Quartet) 1991, Studio</td>
</tr>
<tr>
<td align="left" height="17">Bill Frisell</td>
<td align="left">Rambler</td>
</tr>
<tr>
<td align="left" height="17">Charles Lloyd</td>
<td align="left">Lift Every Voice</td>
</tr>
<tr>
<td align="left" height="17">Daniel Barenboim / Michael Barenboim / Yulia Deyneka / Kian Soltani</td>
<td align="left">Mozart: Piano Quartets</td>
</tr>
<tr>
<td align="left" height="17">David Murray Octet</td>
<td align="left">Home</td>
</tr>
<tr>
<td align="left" height="17">Don Pullen</td>
<td align="left">Capricorn Rising: Featuring Sam Rivers</td>
</tr>
<tr>
<td align="left" height="17">Enrico Pieranunzi/Thomas Fonnesbaek</td>
<td align="left">Blue Waltz</td>
</tr>
<tr>
<td align="left" height="17">Fred Hersch</td>
<td align="left">Heartsongs</td>
</tr>
<tr>
<td align="left" height="17">JD Allen</td>
<td align="left">Love Stone</td>
</tr>
<tr>
<td align="left" height="17">Julian Argüelles' Tetra</td>
<td align="left">Tonadas</td>
</tr>
<tr>
<td align="left" height="17">Martyn Brabbins/BBC National Orchestra of Wales</td>
<td align="left">John Pickard: Sixteen Sunrises; Symphony No. 5; Concertante Variations</td>
</tr>
<tr>
<td align="left" height="17">Old and New Dreams</td>
<td align="left">Old and New Dreams [1977]</td>
</tr>
<tr>
<td align="left" height="17">Paul Bley</td>
<td align="left">Memoirs</td>
</tr>
<tr>
<td align="left" height="17">Seattle Symphony / Roomful of Teeth / Ludovic Morlot</td>
<td align="left">Berio: Sinfonia; Boulez: Notations I-IV; Ravel: La Valse</td>
</tr>
<tr>
<td align="left" height="17">Steve Coleman</td>
<td align="left">Live at the Village Vanguard, Vol. 1 (The Embedded Sets)</td>
</tr>
<tr>
<td align="left" height="17">Sylvie Courvoisier/Mary Halvorson</td>
<td align="left">Crop Circles</td>
</tr>
<tr>
<td align="left" height="17">Telemann; Richard Boothby</td>
<td align="left">Solo Fantasias</td>
</tr>
<tr>
<td align="left" height="17">Thomas Demenga</td>
<td align="left">J.S. Bach: Suiten für Violoncello</td>
</tr>
<tr>
<td align="left" height="17">Tim Berne/Matt Mitchell duo</td>
<td align="left">Angel Dusk</td>
</tr>
<tr>
<td align="left" height="17">Various Artists</td>
<td align="left">Journey – music for Indian violin & tuba</td>
</tr>
<tr>
<td align="left" height="17">Wes Montgomery</td>
<td align="left">In Paris: The Definitive ORTF Recording</td>
</tr>
<tr>
<td align="left" height="17">William Tatge</td>
<td align="left">General Cargo</td>
</tr>
</tbody></table>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-50038953804720156532018-07-01T17:03:00.000-07:002018-07-01T17:03:05.626-07:00Jazz wonders<div class="separator" style="clear: both; text-align: center;">
<br /></div>
<br />
<ul>
<li>Heard Christian McBride's <i>New Jawn Quartet</i> at Stanford last night. I knew it would be good, but it was way above good, it was revelatory. All new music, no pretension, no concession to easy listening, superbly tight. I had heard them at SFJAZZ in 2016, they were good but they did not make this intense impression. Yesterday, Waits was like a fast current in a deep channel, many overlaid rhythms, no derivative splash. Waits and McBride set an intense pace, which Evans and Strickland rode creatively without ever going lax or derivative. Still recovering, like a scary steep ski descent. They'll be recording, album expected Sept-Oct. But best, they are touring widely. Go hear them!</li>
</ul>
<br />
<div class="separator" style="clear: both; text-align: center;">
<iframe width="320" height="266" class="YOUTUBE-iframe-video" data-thumbnail-src="https://i.ytimg.com/vi/5ELNHxpTIQM/0.jpg" src="https://www.youtube.com/embed/5ELNHxpTIQM?feature=player_embedded" frameborder="0" allowfullscreen></iframe></div>
<div class="" style="clear: both; text-align: center;">
</div>
<ul>
<li style="text-align: left;">Great <a href="https://amp.pastemagazine.com/articles/2018/06/jazz-icon-charles-lloyd-leads-lucinda-williams-int.html">detailed review</a> of Charles Lloyd and the Marvels + Lucinda Williams's <i>Vanished Gardens</i>.</li>
<li style="text-align: left;">Wonderful, long <a href="https://ethaniverson.com/interview-with-bill-frisell/">historical interview</a> with Bill Frisell. Includes a link to a bootleg video of Paul Motian Trio (Motian, Frisell, Joe Lovano) at Jazz em Agosto, Lisboa 1986. Frisell and Lovano look so young!</li>
</ul>
<br />
<div class="separator" style="clear: both; text-align: center;">
<iframe width="320" height="266" class="YOUTUBE-iframe-video" data-thumbnail-src="https://i.ytimg.com/vi/JaCCxBIoEZc/0.jpg" src="https://www.youtube.com/embed/JaCCxBIoEZc?feature=player_embedded" frameborder="0" allowfullscreen></iframe></div>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-61957015025009039292018-06-25T22:00:00.000-07:002018-06-25T22:00:23.582-07:00Photos of my work office small music player<div class="separator" style="clear: both; text-align: left;">
Music SSD in the foreground, player in the in the middle, DAC (green/black) and amp (blue) in the background. SSD is connected to one of the Pi 2 USB ports. Digital audio goes out through an BNC connector from the 502DAC bolted to the back of the Pi touchscreen and plugged into the Pi hat connector. BNC coax cable connects to the DAC. Left picture shows the main piCorePlayer menu; right picture shows the Now Playing screen, in this case for the first track of Marc Ribot's excellent <i>Silent Movies</i>.</div>
<div class="separator" style="clear: both; text-align: center;">
<br /></div>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjAcRDvj6s9frD6ArKzf0i2lfD5iVmg5jP2M3o8rKWKg329TwYWL8yneveB5iSyEAelS6hqXN-zrAmZzSUTrWo0RwAu9GL75w5H7w9pW9TSpnfg3qDyAPc0fe6Zl32Pq_7Bk4BfiuD4Rw/s1600/IMG_20180625_093735.jpg" imageanchor="1" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" data-original-height="1600" data-original-width="1200" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjAcRDvj6s9frD6ArKzf0i2lfD5iVmg5jP2M3o8rKWKg329TwYWL8yneveB5iSyEAelS6hqXN-zrAmZzSUTrWo0RwAu9GL75w5H7w9pW9TSpnfg3qDyAPc0fe6Zl32Pq_7Bk4BfiuD4Rw/s320/IMG_20180625_093735.jpg" width="240" /></a><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhe4R7_8QQ7hA5q_YjJoNdpv_zAJs8vSfVAyTjATbJMg_hQ5kqVk-TTRssIfD7BoLQ9MOIMQYCCUp0Mr0UYxLugXvA4YCZjuqBMxMOwejqra3TjUcdUCOhj6ciEWY5eBRQUdAzKmJ6usw/s1600/IMG_20180625_093532.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="1600" data-original-width="1200" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhe4R7_8QQ7hA5q_YjJoNdpv_zAJs8vSfVAyTjATbJMg_hQ5kqVk-TTRssIfD7BoLQ9MOIMQYCCUp0Mr0UYxLugXvA4YCZjuqBMxMOwejqra3TjUcdUCOhj6ciEWY5eBRQUdAzKmJ6usw/s320/IMG_20180625_093532.jpg" width="240" /></a></div>
<br />
<br />Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-91771117915843845162018-06-23T17:21:00.001-07:002018-06-24T09:21:53.228-07:00Building a small standalone digital music playerFor the last several years, all my recorded music listening has been from lossless PCM FLAC stored on a Synology NAS. I went through several iterations of Ethernet-based streaming from the NAS, but now I've settled on a <a href="https://roonlabs.com/">Roon</a> server on a Ubuntu Intel NUC that streams to a couple of different Roon endpoints: an <a href="https://us.auralic.com/products/aries">Auralic Aries Femto</a> for the living room speaker system, and an <a href="https://www.allo.com/sparky/usbridge.html">Allo USBridge</a> for the study headphone system. Those endpoints connect to DAC>amp>transducer chains.<br />
<br />
For work, however, I want my music system there to be standalone for convenience and security. After a bit of exploration, I settled on the following hardware:<br />
<ul>
<li>Pi 2 single-board computer.</li>
<li>AdaFruid 7'' touchscreen.</li>
<li>SmartiPi touchscreen stand and Pi case.</li>
<li><a href="http://www.pi2design.com/502dac.html">Pi2 Design 502DAC</a>, to be used as a high-quality S/PDIF source, not as a DAC. A possible alternative would be the <a href="https://www.allo.com/sparky/digione.html">Allo DigiOne</a> board, which I did not consider at the time for whatever reason.</li>
<li>Samsung 1TB USB-C SSD for music storage.</li>
<li>sBooster ECO 5-6v LPS to power the above. Probably overkill, but you definitely need 3A@5V, which is more than the typical USB wall wart.</li>
</ul>
The player connects with a S/PDIF coax cable to my <a href="http://www.soekris.dk/dac1541.html">Soekris dac1541</a> DAC/amp, which on its own it is a very nice source for my <a href="https://www.mrspeakers.com/headphones-18/aeon.html">MrSpeakers Æon Flow Closed</a> headphones. However, I currently have an extra <a href="https://www.neurochrome.com/product/hp-1-ultra-high-end-headphone-amp-2/">Neurochrome HP-1</a> headphone amp, and that sounds even better between the DAC and the headphones than the dac1541's built-in amp.<br />
<br />
What remained was to find standalone music player software that would run well on that low-power computer. I started with <a href="http://www.runeaudio.com/">RuneAudio</a>, which worked but was was sluggish, often missed command touches, and sometimes got stuck doing harder chores. Scrolling through my 1TB music library was really annoying. I also tried <a href="https://volumio.org/">Volumio</a>, but I could not get it to run stably on my hardware. Eventually, I heard about <a href="https://sites.google.com/site/picoreplayer/home">piCorePlayer</a> on an audio forum as being really lightweight -- it runs from RAM disk -- and decided to give it a try. All of those three Linux-based players are mainly designed to stream music from a separate server or streaming service, but RuneAudio and Volumio can in principle work standalone from a local USB disk.<br />
<br />
My adventure was to try to get piCorePlayer to work standalone, even though it is mainly a replacement for the formerly proprietary networked Squeezebox player. I managed to get this to work thanks to advice from kind users of an audio forum as well as a lot of searches for documentation and other forum info. In brief, I eventually succeeded, and the resulting player works really well, scrolling quickly through my big library and allowing me to select the right album for what I'm working on those precious times I'm not in meetings. Here's what I did:<br />
<ol>
<li>Downloaded <a class="externalLink" href="https://sourceforge.net/projects/picoreplayer/files/insitu/piCorePlayer3.5.0/piCorePlayer3.5.0.zip/download" rel="nofollow" style="border-radius: 5px; font-family: inherit; margin: 0px -3px; padding: 0px 3px;" target="_blank">piCorePlayer 3.5.0</a><span style="background-color: white; font-family: inherit;">.</span></li>
<li>After downloading and unzipping, used Etcher on my Macbook Pro and a SD-2-microSD carrier to flash the piCorePlayer image onto a 64GB SanDisk microSD.</li>
<li>Inserted the card into the Pi 2 microSD carrier, reassembled the unit, connected it to my home LAN, DAC, USB SSD, and power.</li>
<li>On boot-up, boot messages on touchscreen are upside-down. Don't worry, it will be solved later.</li>
<li>Connected to the piCorePlayer software running on the Pi with the Web browser (Chrome) on my Macbook Pro. For convenience (it will come especially handy later), I assigned a fixed IP on my LAN to the Pi, which is really easy to do with the Web interface to the Ubiquiti EdgeRouter that manages my home LAN.</li>
<li>Enabled Beta software on the main piCorePlayer control Web page. This will come in handy later.</li>
<li>Using the piCorePlayer Web interface, installed LMS. This requires resizing the boot partition, which involves several rebooting dialogs, and then actually installing LMS.</li>
<li>Set up your preferred name for the piCorePlayer, and also tell LMS about it on the LMS configuration page. Mine is "<a href="https://en.wikipedia.org/wiki/Ebnefluh">Ebnefluh</a>," after a memorable ski tour I did in April (all my music-related machines are named after peaks I visited on skis).</li>
<li>Make sure that the LMS flag to bypass mysqueezebox is set and saved.</li>
<li>Install jivelite package to manage the touchscreen.</li>
<li>Once jivelite is installed, use its configuration page on your browser to adjust screen rotation. In my setup, I had to select the option to flip upside-down.</li>
<li>Use the LMS Web interface to tell LMS where your music library is in the USB drive, and to get it to index it. There's a page that shows indexing progress. If you did 9 above, you won't be prompted to get a mysqueezebox account.</li>
<li>Wait until your music is indexed. In my case, this stopped somehow, and I had to poke it on the LMS Web interface. But it got eventually done.</li>
<li>Now you can test that you can control play to your DAC from the touchscreen. Enjoy testing with some known tunes!</li>
<li>Just to be sure everything so far is remembered, use the "Backup" option on the piCorePlayer control page to save your current configuration to the microSD card.</li>
<li>The final step is to make your device work standalone. On the beta options on the piCorePlayer control page, click to set a fixed IP address. That gets you to a network configuration Web page. Set DHCP to off, enter your fixed IP address, netmask, default gateway, and default DNS. You should set this to what you have on the LAN you are configuring the device on, so that it talks correctly to it when you bring the device back to it for software updates etc.</li>
<li>Backup your whole configuration again to microSD. This is critical!</li>
<li>Shut down the device, unplug it from power, and let it rest for a while so that its RAM resets. Also unplug it from your network.</li>
<li>Power up the device again. Once it is up, you should be able to access your music and control play from the touchscreen.</li>
<li><b>Troubleshooting</b>: at step 19, if you see a boot-up message that the device is waiting for network and that stays for a while, outputting periods on the screen, that means that you did on 16 did not stick, probably because you forgot to backup the configuration to microSD before power down. If that is done correctly, the network should come up right away, and LMS will also boot-up quickly. If not, after the long failed wait for network, LMS will not spin up and Squeezelite won't find your music.</li>
</ol>
<br />Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com1tag:blogger.com,1999:blog-6916168470376937425.post-1731352671645919832018-06-02T11:36:00.003-07:002018-06-02T11:36:28.040-07:00Recent musical explorations extended<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjtwp6SL8qqF2iiLrSvEGYCVTTWo56ovOLahZ_qxnlbLnHob_l1CGcUPjrr4mAjusqnW1lRECz2WgrnlMcbKfcsTVr4Jn8cpavRFWsgoHBgOkms7UbQyzg8DpdF9otrG5USfLCdcl7fWQ/s1600/Eclectic.png" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="659" data-original-width="1600" height="260" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjtwp6SL8qqF2iiLrSvEGYCVTTWo56ovOLahZ_qxnlbLnHob_l1CGcUPjrr4mAjusqnW1lRECz2WgrnlMcbKfcsTVr4Jn8cpavRFWsgoHBgOkms7UbQyzg8DpdF9otrG5USfLCdcl7fWQ/s640/Eclectic.png" width="640" /></a></div>
<br />Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-2228210036254933602018-05-25T21:42:00.000-07:002018-05-25T21:43:02.437-07:00Seven days of recorded music explorations<style type="text/css">
body,div,table,thead,tbody,tfoot,tr,th,td,p { font-family:"Liberation Sans"; font-size:x-small }
a.comment-indicator:hover + comment { background:#ffd; position:absolute; display:block; border:1px solid black; padding:0.5em; }
a.comment-indicator { background:red; display:inline-block; border:1px solid black; width:0.5em; height:0.5em; }
comment { display:none; }
</style>
<br />
<table border="0" cellspacing="0">
<colgroup span="2" width="392"></colgroup>
<tbody>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Alessio Bax / Southbank Sinfonia / Simon Over</span></td>
<td align="left"><span style="font-family: inherit;">Beethoven: Piano Concerto No. 5; Works for Solo Piano</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Angelika Niescier Trio</span></td>
<td align="left"><span style="font-family: inherit;">The Berlin Concert</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Brad Mehldau Trio</span></td>
<td align="left"><span style="font-family: inherit;">Seymour Reads the Constitution!</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Cyro Baptista</span></td>
<td align="left"><span style="font-family: inherit;">Vira Loucos: Cyro Baptista Plays the Music of Villa Lobos</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Gallicantus, Gabriel Crouch</span></td>
<td align="left"><span style="font-family: inherit;">Sibylla</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Jeff Parker</span></td>
<td align="left"><span style="font-family: inherit;">Slight Freedom</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Jeff Parker</span></td>
<td align="left"><span style="font-family: inherit;">The New Breed</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Jerusalem Quartet</span></td>
<td align="left"><span style="font-family: inherit;">Debussy: Quatuor, Op. 10; Ravel: Quatuor</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Jory Vinikour</span></td>
<td align="left"><span style="font-family: inherit;">Bach: Goldberg Variations</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Joshua Redman</span></td>
<td align="left"><span style="font-family: inherit;">Still Dreaming</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Leila Josefowicz / David Robertson / Saint Louis Symphony Orchestra</span></td>
<td align="left"><span style="font-family: inherit;">John Adams: Violin Concerto</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Marc Sinan/Oguz Buyukberber</span></td>
<td align="left"><span style="font-family: inherit;">White</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Stephen Schultz / Jory Vinikour</span></td>
<td align="left"><span style="font-family: inherit;">J.S. Bach: Sonatas for Flute and Harpsichord</span></td>
</tr>
<tr>
<td align="left" height="17"><span style="font-family: inherit;">Wil Blades</span></td>
<td align="left"><span style="font-family: inherit;">Field Notes</span></td>
</tr>
</tbody></table>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-17470983538181952632017-06-10T22:33:00.000-07:002017-06-16T20:37:37.071-07:00A (computational) linguistic farce in three acts<h3 dir="ltr" style="line-height: 1.38; margin-bottom: 4pt; margin-top: 14pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 13pt; font-style: normal; font-variant: normal; font-weight: 700; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Prologue</span></h3>
<h3>
<b id="docs-internal-guid-5d4c3b81-956a-8cea-9373-725a5ac6ff16" style="font-weight: normal;"><div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">I had not blogged for 3 years. Many plausible excuses, but the big reason is that it is easier to dash a tweet or a short incidental social media post than to structure a complex argument, which uses mental resources that I need full time at work. But the argument about deep learning, natural language, publication styles and venues that Yoav Goldberg </span><a href="https://medium.com/@yoav.goldberg/an-adversarial-review-of-adversarial-generation-of-natural-language-409ac3378bd7" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">posted on Medium</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> reminded me of something that one day (not today) I would like to get to, the complicated, sometimes hazy, often contentious history of the science and engineering of language as a computational process (I know, I know, even putting it that way could trigger many social scientists and philosophers, but this is just a blog post, not a treatise). </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">I call this a farce not in a derogatory way, but for its many misunderstandings and pratfalls, in the best traditions of comedic theatre, opera, and silent movie. Who has not </span><a href="https://youtu.be/2WZLJpMOxS4" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">stepped on a rhetorical rake</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> in heated academic discussion may cast the first water balloon. After all, these debates issued from the very serious work of intellectual giants of the 50s and early 60s: Kleene, Shannon, Harris, McCulloch, McCarthy, Minsky, Chomsky, Miller, ... One day I'd love to see a careful, thoughtful intellectual history of the origins of AI in general and of the computational turn in language in particular, but we don't have one, so I'm free to make up my own comedic version.</span></div>
</b></h3>
<h3 dir="ltr" style="line-height: 1.38; margin-bottom: 4pt; margin-top: 14pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 13pt; font-style: normal; font-variant: normal; font-weight: 700; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Act One: The (Weak) Empire of Reason </span></h3>
<h3>
<b style="font-weight: normal;"><div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Much of the work on computational models of language and language processing until the 80s was based on an implicit or explicit hope that relatively simple algorithms would capture much of what mattered. Researchers (including me) created models and algorithms that claimed to capture the “essential” phenomena in a modular, compositional way. Once that was done, practical applications would follow easily, since the nice combinatorics of compositionality would cover the infinitely many ways people express meaning. </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">That was nice, but there was the nagging problem that none of those models or systems could parse, let alone usefully interpret most of the language occurring in the wild. Even back then, artificial neural network fans argued that those crisp formal models of language failed because they did not have enough “flex” at their joints. That led to some epic food fights, but the reality is that NN models, algorithms, but mostly the puny computers and datasets we played with back then could not even match those carefully handcrafted rule systems.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">One (temporary) escape from this mismatch between models and actual language was to turn ourselves into formal linguists (I did that too) and argue that we were using computational tools to investigate the core of language, leaving that wild mess of actual language for later decades when we'd have finally dug up the keys to the treasure house. This was a nice detour for both symbolic and neural-network researchers, and it had a not totally unreasonable methodological defense in that, say, physicists also investigated simplified systems (oh, that physics envy!) Of course, this sidestepped the uncomfortable feeling that language, as an evolved biological and social phenomenon, might not have a simple description at all. Incidentally, there's a bit of a parallel here with how biology and biomedical research went on a “simplicity” trip after the discovery of the genetic code (and even later after the sequencing of the human genome), only to keep being foiled by the daunting mess that evolution has left us. Nevertheless, I'd still argue that some of the descriptive models of language developed then still capture the range of certain actual combinatorial possibilities in language at a level of detail that has not been bested. That's mainly a story for another time, except that the lure of simplified settings and models comes back in Act Three.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The field was very small back then. Everyone knew everyone, even those who might despise each other's work and say it loudly in ACL question periods. As a result, a few powerful arbiters of research taste set the tone for each sub-community. Combined with the limited means of research circulation then, that led to small, cohesive cliques. When such a group captured control of research resources (funding, plum academic or industrial roles), as did happen, alternative ideas did not have much room to grow.</span></div>
</b></h3>
<h3 dir="ltr" style="line-height: 1.38; margin-bottom: 4pt; margin-top: 14pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 13pt; font-style: normal; font-variant: normal; font-weight: 700; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Act Two: The Empiricist Invasion or, Who Pays the Piper Calls the Tune</span></h3>
<h3>
<b style="font-weight: normal;"><div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The not insignificant research funding that computational research on language had received from the late 70s to the late 80s, combined with changes in research funding climate (a whole interesting story in itself, but too long and twisted to go into here) created an opening for bold invaders to convince funders that the Emperor of Reason had been committing research in the altogether.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The empiricist invaders were in their way heirs to Shannon, Turing, Kullback, I.J. Good who had been plying an effective if secretive trade at IDA and later at IBM and Bell Labs looking at speech recognition and translation as cryptanalysis problems (The history of the road from Bletchley Park to HMMs to IBM Model 2 is still buried in the murk of not fully declassified materials, but it would be awesome to write — I just found </span><a href="https://books.google.com/books?id=EMMsCgAAQBAJ&dq=kullback+cryptanalysis" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">this</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> about the early steps that could be a lot of fun). They convinced funders, especially at DARPA, that the rationalist empire was hollow and that statistical metrics on (supposedly) realistic tasks were needed to drive computational language work to practical success, as had been happening in speech recognition (although by the light of today, that speech recognition progress was less impressive than it seemed then). It did not hurt the campaign that many of the invaders were closer to the DoD in their backgrounds, careers, and outlooks than egghead computational linguists (another story that could be expanded, but might make some uncomfortable). Anyway, I was there in meetings where the empiricist invaders allied with funders increasingly laid down the new rules of the game. Like in the Norman invasion of England, a whole new vocabulary took over quickly with the new aristocracy.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">In hindsight, the campaign of 1987-89 and the resulting new order were quite entertaining (even if they did not seem so to the invaded at the time) and brought new cultural devices that were objectively more effective in defining measurable progress, if quite stressful for funding recipients. Personally, I had already started </span><a href="https://pdfs.semanticscholar.org/e597/cb9d0a9bd767b39dccf94696ed4e3557461b.pdf" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">my own journey</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> from somewhat skeptical rationalism to somewhat skeptical empiricism, and would leave the Government-funded research world for the next 12 years, so the conflict was a great opportunity to develop a more distanced view of both the old and the new culture.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The empiricist ascendancy was fortunate (or prescient) in riding the growth of computing resources and text data that also enabled the Web explosion and the flooding of all this work with new resources for funding research, software development, and corpus creation. The metrics religion helped funders sell progress to the holders of the purse strings, and there were real (if not as extensive as sometimes claimed) practical benefits, especially in speech recognition and machine translation. As a result, the research community grew a lot (my top-of-the-head estimate is around 5x from 1990 to 2010).</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">One curious byproduct of the empiricist ascendancy that is relevant to the present conflict is that measurement became a virtue in itself, sometimes quite independently of what was really being measured. Many empiricist true believers just want the numbers, regardless of whether they correspond to anything relevant to actual language structure and use. Although Penn Treebank metrics are most often brought up for this critique, there are much worse offenders that I will omit in the interests of both not offending by naming without proper arguments, and of not spending my whole weekend on this. In summary, a certain metric fetishism arose that still prevails today for instance in conference reviewing, with the result that interesting models and observations are dismissed unless they improve one of the blessed metrics. Metrics became publishing gatekeepers, easy to apply without thinking, and promoting a kind of p-hacking culture that demeaned explanation and error analysis. Worst, for a practitioner, was that all the metrics are averages, when large deviations is what really matters if you are responsible for a product that should have very low chance of doing something really bad.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Which brings us to the final act.</span></div>
</b></h3>
<h3 dir="ltr" style="line-height: 1.38; margin-bottom: 4pt; margin-top: 14pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 13pt; font-style: normal; font-variant: normal; font-weight: 700; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Act Three: The Invaders get Invaded or, The Revenge of the Spherical Cows</span></h3>
<h3>
<b style="font-weight: normal;"><div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Empiricist practice has a deep flaw that is rarely discussed. At its origins, cryptanalysts were working on sources very restricted in content, and vocabulary. After all, it was not so likely that Enigma traffic would have much outside the military order of the day, weather, and the like, or go full Jabberwocky. When you get enough token counts for traffic like that, you pretty much know everything you can know. It's what Harris profoundly noted in the differences between technical languages and general language. Popular tasks of the empiricist era, from ATIS to PTB, were similarly restricted (travel, business news, ...). What this means is that typical count-based empiricist methods do much better on their own benchmarks than in real life. Just try to parse the Web (let alone social media or chat) with a PTB parser to see what I mean. </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Where a lot of training data can be collected in the wild — most notably, parallel translation corpora — empiricist practice with enough counts limps along the long tail, although it's touch-and-go when the counts get small, as they always do. </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Another way to see this is that empiricist dogma was protected from its own demise by a rather convenient choice of evaluations. Those of us who have worked hard to apply these methods to real data know well the struggle with small token counts, and the dismaying realization that fancier statistical methods (such as latent-variable models) are most often a waste of effort because in practical situations, a flat count-based model (or linear models) can do as well as can be hoped with the data at hand. Those of us who thought a bit more about this started to realize that token counting and its variants could not generalize effectively across “similar” tokens. We tried many different recipes to alleviate the problem (eg. class-based language models), but they were all ineffective or computationally infeasible (I know, I co-authored quite a few papers in that mode, including at least a couple of best papers — goes to show the limited horizons of program committees).</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">It's here that deep learners carrying warming GPUs descended from the Northern wastes to lay siege to the empiricist (bean) counters and their sacred metrics. First language modeling, then machine translation fell, in no little measure thanks to their ability to learn usage and meaning generalizations much better than counting could. The modularity of NN models made it easier to explore model design space. Recurrent gated models (thank you </span><a href="http://www.mitpressjournals.org/doi/abs/10.1162/neco.1997.9.8.1735" style="text-decoration: none;"><span style="background-color: white; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">Hochreiter and Schmidhuber</span></a><span style="background-color: white; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">!) managed history in a much more flexible and adaptable way than any of the history-counting tricks of the previous two decades. </span><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">It was a rout.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The excitement of the advance was irresistible. Researchers involved, experiments, and papers grew fast, in my estimate 4x from 2010 to 2017. Publication venues overflowed, and researchers burning ever more fossil fuel with their GPUs (morality tale warning!) turned to arXiv to plant ever more flags on their marches through newly conquered lands (not the most culturally apt of behaviors, it must be admitted).</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">But was the invasion so glorious? Very few of the standard tasks have the very large training sets of language modeling or translation that large-scale SGD depends on. Some tasks with carefully created training sets, like parsing, showed significant but not as striking gains with deep learning. There are exciting results in transfer learning (such as the </span><a href="https://research.googleblog.com/2016/11/zero-shot-translation-with-googles.html" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">zero shot translation</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> results), but they rely in starting from models trained on a whole lot of data. However, when we get to tasks for which we have only evaluation data, where count-based models can still do decent work (clustering, generative models), deep learning does not have yet a superior answer.</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">For continuous outputs, GANs have made a lot of progress. At least, the pictures are stunning. But as I discovered when I worked on </span><a href="https://scholar.google.com/citations?view_op=view_citation&hl=en&user=qWDmIgIAAAAJ&citation_for_view=qWDmIgIAAAAJ:d1gkVwhDpl0C" style="text-decoration: none;"><span style="background-color: transparent; color: #1155cc; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: underline; vertical-align: baseline; white-space: pre-wrap;">distributional clustering</span></a><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">, stunning is very much in the eye of the beholder. Proxies like the word association tasks so popular in evaluating word embeddings are almost embarrassingly low-discrimination compared with the size of the models being evaluated. Sensible ways of evaluating GANs for text are even scarcer. </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">In defense of the Northern hordes, the empiricist burn-it-to-the-ground campaign left little standing that could promote a new way of life. Once the famous empiricist redoubts are conquered or at least laid siege to, how does the campaign continue?</span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Idea! Let's go back to toy problems where we can create the test conditions easily, like the rationalists did back then (even if we don't realize we are imitating them). After all, Atari is not real life, but it still demonstrates remarkable RL progress. Let's make the Ataris of natural language! </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">But now the rationalists converted to empiricism (with the extra enthusiasm of the convert) complain bitterly. Not fair, Atari is not real life! </span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<br /></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Of course it is not. But neither is PTB, nor any of the standard empiricist tasks, which try strenuously to imitate wild language (their funding depends on it!) but really fail, as Harris predicted back in the 1950s. Or even the best of descriptive linguistics, which leaves in the murk all those messy deviations from the nice combinatorics of the descriptive model. </span></div>
</b></h3>
<h3 dir="ltr" style="line-height: 1.38; margin-bottom: 4pt; margin-top: 14pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 13pt; font-style: normal; font-variant: normal; font-weight: 700; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Epilogue</span></h3>
<h3>
<b style="font-weight: normal;"><div dir="ltr" style="line-height: 1.38; margin-bottom: 3pt; margin-top: 14pt;">
<span id="docs-internal-guid-5d4c3b81-95a3-2360-5810-b45014ae65e7"></span></div>
<div dir="ltr" style="line-height: 1.38; margin-bottom: 0pt; margin-top: 0pt;">
<span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The mysticism of Mozart's </span><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: italic; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">The Magic Flute</span><span style="background-color: transparent; color: black; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;"> makes me queasy, and honestly the opera is longer than it should be (at least on an uncomfortable concert hall seat). But the music, and the ultimate message! The main protagonists struggle for and eventually reach enlightenment along their different paths. We are very far from </span><span style="background-color: white; color: #222222; font-family: "arial"; font-size: 11pt; font-style: italic; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">Dann ist die Erd' ein Himmelreich, und Sterbliche den Göttern gleich </span><span style="background-color: white; color: #222222; font-family: "arial"; font-size: 11pt; font-style: normal; font-variant: normal; font-weight: 400; text-decoration: none; vertical-align: baseline; white-space: pre-wrap;">(thank you neural MT for checking my quotation from the original), but we have been struggling long enough in our own ways to recognize the need for coming together with better ways of plotting our progress. </span></div>
</b></h3>
Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com12tag:blogger.com,1999:blog-6916168470376937425.post-71947463940365439502014-05-16T23:03:00.000-07:002014-05-16T23:16:10.496-07:00Innocence<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgHHosg-JN56RVZczXIsuTY_HwKLVLsgMBvaD1pbENXzY-f_9r_HC8Q8A8xzhLWh3f4S662xhyw3wtz4c60Mcph1fUqnTPCMUs1cvYlFJTVGlOAl4wJ2z7EaYN2CRdts1qKGoFGS0_Stw/s1600/IMG_1064.JPG" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgHHosg-JN56RVZczXIsuTY_HwKLVLsgMBvaD1pbENXzY-f_9r_HC8Q8A8xzhLWh3f4S662xhyw3wtz4c60Mcph1fUqnTPCMUs1cvYlFJTVGlOAl4wJ2z7EaYN2CRdts1qKGoFGS0_Stw/s1600/IMG_1064.JPG" height="240" width="320" /></a></div>
Slowly, I've been trying to get into the video 21st century. I've finally got a Blu-ray video player (Oppo BDP-103), just wired it to our TV and audio system. The only video I could find at home to test it was a DVD of the backcountry ski movie <a href="http://www.wildsnow.com/33/sanctified-movie-review/"><i>Sanctified</i></a>. Its opening with words by <a href="http://www.oilfreefun.com/2013/12/bob-athey-wizard-of-wasatch.html">Bob Athey</a> struck me with surprising force for its old-school feel, even though it's only nine years old. The late Doug Coombs makes an appearance, and much of the words are for keeping the slopes wild, interspersed with the inevitable perfect backcountry powder runs. Maybe it's my frustratingly slow recovery from the return of my twisted lower back after unprecedented (for me) skiing 29 days out of 80 in four countries, three continents, mostly with no or just partial lift help, maybe it is the sadness of missing the usually glorious spring skiing of the Sierra and the Cascades because of my bad back and of the exceptionally dry winter, or maybe it's just nostalgia for the innocent silliness of the defunct TelemarkTips bboard, but this movie makes it all seem simpler, less ambiguous, and makes me long for those days when I had become just got good enough on skis to become totally besotted with the mountains and the sport, and managed to enjoy places and conditions that I'd not even bother to consider today.<br />
<br />
I want to go skiing, but I have no idea when my back will let me.<br />
<br />Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-66167284234284800692011-07-30T21:00:00.000-07:002011-07-30T21:00:45.932-07:00Short-blogging at Google+It's just easier: <a href="https://plus.google.com/102078454522244583150/posts">Fernando Pereira</a>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-13344478038778221912011-07-04T22:55:00.001-07:002011-07-04T23:09:18.403-07:00Social set theory<a href="http://www.tbray.org/ongoing/When/201x/2011/07/04/Google-plus">Tim Bray</a>: <blockquote><cite><strong>Math math math</strong>: I’d rather not join the parade of people shouting for one new feature or another, because it seems to me that G+ as it stands hits a decent 80/20 point. Having said that, Circles are, mathematically speaking, sets, and I think set arithmetic would come in real handy: “Post this to the intersection of my Photogeeks circle and my Vancouver circle”. I can think of lots of other amusing permutations. The reason I bring this up is because I smile, envisioning a future in which math teachers use social-network constructs to explain Set Theory.</cite></blockquote>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-14291471912710370342011-07-02T16:26:00.000-07:002011-07-02T20:50:41.375-07:00Why we publish<blockquote><cite>These Thoughts, my dear Friend, are many of them crude and hasty, and if I were merely ambitious of acquiring some Reputation in Philosophy, I ought to keep them by me, ’till corrected and improved by Time and farther Experience. But since even short Hints, and imperfect Experiments in any new Branch of Science, being communicated, have oftentimes a good Effect, in exciting the attention of the Ingenious to the Subject, and so becoming the Occasion of more exact disquisitions (as I before observed) and more compleat Discoveries, you are at Liberty to communicate this Paper to whom you please; it being of more Importance that Knowledge should increase, than that your Friend should be thought an accurate Philosopher.</cite></blockquote><p>Letter from Benjamin Franklin to Peter Collinson, September 1753, quoted by Lewis Hyde in <a href="https://creativecommons.org/weblog/entry/23204"><cite>Common as Air</cite></a>.</p><p>Irony note: I wasn't allowed to copy-and-paste this quotation in Google eBooks from either Hyde's book or from the out-of-copyright <cite>The Works of Benjamin Franklin</cite>. No better support for Hyde's theses. Ended up finding copiable text on <a href="http://www.enterstageright.com/archive/articles/0407/0407selfinterest.htm">Eric Raymond's site</a>.</p><p><strong>Update:</strong> Thanks to Lewis Hyde's gracious hint below, I've now replaced the "modernized" text I had found on the Web by the text from the Franklin papers.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com1tag:blogger.com,1999:blog-6916168470376937425.post-6076416501118964162011-06-05T12:46:00.000-07:002011-06-05T12:46:16.599-07:00Slow blogging: NIPS<p>As NIPS program co-chairs, Peter Bartlett and I have just been deluged with 1394 paper submissions to steer through reviewing with 52 area chairs and over 600 first-line reviewers. I doubt that I'll have much time for much blogging in the next 3 months. A few quick links:</p><ul><li>Peter Norvig wrote a very interesting <a href="http://norvig.com/chomsky.html">essay on Chomsky vs statistical learning</a> that has triggered a <a href="http://languagelog.ldc.upenn.edu/nll/?p=3172">lot</a> of <a href="http://languagelog.ldc.upenn.edu/nll/?p=3180">commentary</a>. I have thought about and written some on these issues, which Peter graciously notes, but I feel that there's something more to say, about contingency in language, which I hope to be able to at least blog about some time.</li><li>Stu Shieber keeps writing <a href="http://blogs.law.harvard.edu/pamphlet/2011/06/04/the-benefits-of-copyediting/">excellent stuff</a> on open-access publication. Matt Blaze also wrote a <a href="http://www.crypto.com/blog/copywrongs/">great post</a> that had some reverberations. And Wired Magazine had an <a href="http://www.wired.com/wiredscience/2011/05/free-science-one-paper-at-a-time-2/all/1">interesting article</a> on the topic. As NIPS PC co-chair, I'm more aware than ever that our pre-digital scientific communication systems are unsustainable. It's like watching a wet snow avalanche, which seems slow compared with the more photographed slab avalanches, but moves unstoppably and churns everything in its path. </li></ul>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-89999147165965768292011-06-05T12:09:00.000-07:002011-06-05T19:23:20.922-07:00May skiing<a href="https://picasaweb.google.com/lh/photo/p-r8HUVJfMCYtJgak4lbJA?feat=embedwebsite"><img src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg_NEAY8U77Wb8gaHXXRh1dSDEyWVfPCbfGb3rYenigurKtkUmHKrYTkA8gr16NRZ_0wka_4dsWKm1IBm6vnNE4J2AetRXzvjmPtIohLOHK99vxOT03nAbllAxokf8CqsdC8CfqPL2B4g/s400/IMG_1045.JPG" height="225" width="300" align="left" /></a><p>Between work and the actual skiing, I've not had time to post updates. In early May, I participated in the inaugural <a href="http://alpineskills.com/spring_roadtrip.html">ASI/The Backcountry Spring Sierra road trip</a>. <a href="http://www.thebackcountry.net/">The Backcountry</a>'s Mike Schwartz drove and provided great advice and help on and off the mountain, ASI guide Logan Talbott got us safely to beautiful summits and delicious corn skiing. Coincidentally, I had skied with two of the other clients, Jennifer and Justin, on a trip to Sol Mountain a few years ago. The whole group got along well, and made for a remarkably friction-free trip in sometimes cramped and improvised conditions. Here are some pictures <a href="http://www.flickr.com/photos/logantalbott/sets/72157626639439564/">by Logan</a> and <a href="https://picasaweb.google.com/fernando.pereira/EasternSierraRoadtrip">by me</a>. I'll do it again!</p><p>Two weeks ago, Johannes, Jim stayed Friday night at a Kirkwood condo (the place was closed, spooky like a set for a scary movie), and went for a Saturday tour around Carson Pass. We had misplaced our Snowpark permits, so we parked at the unregulated <a href="https://picasaweb.google.com/fernando.pereira/RedLakeToRoundTop">Red Lake trailhead</a> and skinned from there to Round Top. It's a longer route than the Carson Pass one, but it allowed us to discover some nice terrain behind Elephant Back. The day started blustery, overcast, with sprinkles and snow flurries, but cleared toward the afternoon. The timing worked fairly well, we got back without having to deal with the wet glop that is the main return hazard in spring days.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-12996273962959732062011-05-22T19:42:00.000-07:002011-05-22T20:29:29.460-07:00Animation and the philosophy of mind<p>The best passage I read in a while:</p><blockquote>Everything that you see, when you view a Pixar movie, is what an empiricist philosopher of the eighteenth century would call an impression. It was born and cradled in the mind of a computer, and there it lived and grew.</blockquote><p>Anthony Lane, <a href="http://www.newyorker.com/reporting/2011/05/16/110516fa_fact_lane"><cite>The Fun Factory: Life at Pixar</cite></a>, <cite>New Yorker</cite>, May 16, 2011.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com1tag:blogger.com,1999:blog-6916168470376937425.post-38985732541854846502011-04-23T18:36:00.000-07:002011-04-23T18:36:04.965-07:00Technology Has Social Consequences<a href="http://cacm.acm.org/magazines/2011/5/107695-technology-has-social-consequences/fulltext">Technology Has Social Consequences | May 2011 | Communications of the ACM</a>: <cite>The loss in quality of conference reviewing is just one result of the move to virtual PC meetings. Another outcome is the loss of socialization that took place in PC meetings. It is this lost socialization that contributed to a senior researcher being ignorant of one of the most basic rules of scholarly reviewing.</cite><p>I quoted the paragraph in Moshe's article that best summarizes his main point. The article is worth reading in full. However, I disagree with both its alleged empirical claims and its main argument. Moshe claims that reviewing quality is down in CS conferences, and reviewers are less aware of good scholarly conduct. However, he provides no empirical evidence for these claims except for one anecdote and some vague impressions, which I could easily counter with old tales of reviewing incompetence and malice. He then proceeds to argue that those failings are the result of less face-to-face interaction among reviewers. Again, he provides no empirical evidence for the claim, and he does not consider alternative explanations. In particular, he does not consider the fact that many areas of CS have grown rapidly. For example, I just did a simple calculation to arrive at an estimate that computational linguistics as a field has grown at an average 6%/year over the last 28 years, making the field <strong>five</strong> times as big now as when I presented my first ACL paper. This growth forced conferences to adopt more complex structures, with areas, multiple tiers, electronic submission and review discussion, simply to scale up to the much larger population. We can argue about the specifics of reviewing mechanisms, but the old unitary program committee was already collapsing under the strain around 15 years ago for the first-tier conferences I have been involved in.</p><p>But there's an even bigger potential problem that Moshe does not discuss. As the field has matured, it takes longer and it is harder for someone to become a good reviewer because there's just more to know. Meanwhile, the number of people coming into the field continues to grow and the number of papers submitted grows in proportion. The result is then that the ratio of submissions to qualified reviewers increases. Less qualified reviewers are enrolled, or qualified reviewers are overloaded. Either way, review quality goes down. For all we know, it is this, and not Web-based program committees, that is the root cause of all the complaining about bad reviewing in the last few years.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com1tag:blogger.com,1999:blog-6916168470376937425.post-73467682725251603252011-04-17T11:24:00.000-07:002011-04-17T11:24:39.343-07:00Sugarbowl to Squaw<a href="https://picasaweb.google.com/lh/photo/JQWEdl6H6WI_o3tsfd0pBQ?feat=embedwebsite"><img src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjWRPvU0trH5lHuPWagNOOLSBWKUHC4DUyNpHd1VNzd9i2Ekf_fjbm7LKgG_bEWRc6gZRX6VYsINrhomWMGRWmXN6ewKtL1SsEWH6xiwFhJJh16DtAwiJwyIiCt2rhlRC-tgLB2QkQr3A/s144/IMG_20110416_130923-1.jpg" height="131" width="144" align="left"/></a><p>Drove up to Sugarbowl yesterday (Saturday) to join three other ASI clients for a one-day ski traverse from Sugarbowl to Squaw guided by <a href="http://www.alpineskills.com/asi_certs.html">Logan Talbott</a>. It's just under 8 miles as the crow flies but more like 12 miles given the terrain (once again forgot to set MyTracks on my phone to record the actual track, so this is based on what Logan said). We left the top of Sugarbowl's Lincoln lift at 9:06am and arrived at the Squaw parking lot at 3:59pm, a pretty good pace that included a beautiful lunch break at the top of Tinker Knob (photo). The forecast had been for cloudy skies with 30% chance of showers, but it was dry and fairly warm the whole day. Ski conditions ranged from good corn to very sticky mank, with the good surprise being the drop into Squaw, which had much better snow than we expected for mid-afternoon because of the cooling from the breeze and high clouds. Cold lager at the <a href="http://www.squawchamois.com/">Cham</a> was a perfect way to finish the day.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-67518585752015307612011-03-31T07:19:00.000-07:002011-03-31T07:19:02.996-07:00Fleeting<p><a href="http://thetravelingskier.blogspot.com/2011/03/fleeting.html">Fleeting</a>: <cite>When I moved here I was told by one of those wise local types that I would never ski first tracks at Chamonix. Seemed like a reasonable prophecy. I don't live in Chamonix, I live an hour and a half away. Skiing in Chamonix is expensive and you pretty much have to invest in a lift ticket to access anything of interest. Add to that French toll roads, European gas prices, family responsibilities, and a general aversion to competitive crowds and the wise local saw no argument from me. He still wouldn't. That doesn't mean I wouldn't try, either.</cite></p><p>It's the same for me with most backcountry lines around Tahoe. But it doesn't mean that one can't find surprisingly empty terrain. Read <a href="http://thetravelingskier.blogspot.com/2011/03/fleeting.html">the whole thing</a>.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com0tag:blogger.com,1999:blog-6916168470376937425.post-84161802643140873242011-03-14T22:44:00.000-07:002011-03-14T22:44:18.009-07:00Fairy Tales<a href="http://xkcd.com/872/">Fairy Tales</a>: <br />
<center><img alt="Goldilocks' discovery of Newton's method for approximation required surprisingly few changes." src="http://imgs.xkcd.com/comics/fairy_tales.png" width="400" title="Goldilocks' discovery of Newton's method for approximation required surprisingly few changes." /></center><p>How does he get so well that <a href="http://en.wikipedia.org/wiki/Haruki_Murakami">Haruki Murakami</a>-like unsettling border between overactive, exhausted mathematical wondering and dream incoherence?</p><p>Why does Murakami come to mind so readily? The awful news from his homeland showing that he's been right all along about the deep channels between nightmare and reality? Fairy tales can turn to dread so easily, orderly worlds undone by the interdependencies that sustained them for centuries. </p><p>Days like this, the central limit theorem feels like a bad joke.</p>Fernando Pereirahttp://www.blogger.com/profile/05849361902113771573noreply@blogger.com1