## SoundBytes article Sept 2020

| Music for Tablets – Interview with Henry Lowengard, App Developer https://soundbytesmag.net/music-for-tablets-interview-with-henry-lowengard-app-developer/

  We interview Henry Lowengard who has been developing music software since 1969 and is still going strong.

by Warren Burt, Sept. 2020  

Henry Lowengard ( https://jhhl.net/ ) has been involved with computers and music since 1969.  In the Amiga era, he wrote some of the most interesting computer music programs around, like RGS and LYR.  In the iOS era, he has written many apps that have become standard, such as AUMI, PolyHarp, Droneo, SrutiBox, Enumero, Synthicity Itself and others.  He is now beginning to develop apps which work in web browsers, as a means of making his products even more accessible.  All of his apps are characterized by clever interfaces, deep possibilities, and ease of playing.  We spoke to Henry via Zoom on 22 August 2020.

SoundBytes Magazine:  We’re here today with Henry Lowengard, who is a legendary designer of music software, as well as a number of other things. Tell us just a little bit about your musical activities.

Henry Lowengard:     Well, working backwards, today I was actually supposed to be in a show in the band that I’m in called Mamalama where I play the hammer dulcimer, glockenspiel, tenor recorder, melodica, and sing a little. That would have been in the Widow Jane Mine, which is a former cement mine here in Ulster County. It’s run by a not-for-profit educational and we do performances in there as benefits. I’m the president of their board of trustees.

SM:  Oh, fantastic.

HL:  I was playing in there before I became the president. I played with that band and with my other band, which is the Catskill Mountain Gamelan and both of these sound really, really good in the super live acoustic world of the Widow Jane Mine. At some point, we’ll have that again. So that’s me as a live musician. As a dead musician, I’ve done a lot of computer-based electronic music because at one point I said I’m tired of the solder smoke getting up my nose, so I decided I would just write. This was back in the 80’s and, finally, the fabulous Commodore Amiga came out, which was powerful enough to do user-interface with real-time audio at a decent quality. I have roots that go further back to that. We used to play on a PDP-8 back in high school and we figured out how to get sound out of it, so I wrote some music software that ran on that. So a good forty five years or more of this kind of activity, and I’m still doing it – typing today, in fact.

SM:  Very good. Yes, I remember in the Amiga era, you had a program, for example, called RGS that was rather wonderful. It was one of the first programs I know, besides something like Xenakis’s UPIC, which of course was inaccessible to someone in America or Australia. But it was one of the first programs where you could draw spectrographs and hear them immediately realised to sound.

HL:  Yes, indeed. It was, in fact, inspired a little by UPIC by reading a fabulous issue of Computer Music Journal. I said, “You know, first of all, this program is extremely limited. Second of all, it’s hiding behind an institution, so I’m going to just write one myself that runs on the Amiga”, and I did. Not only that, but it had a couple of really interesting features, like you could take the spectrum and not only synthesise it while you were drawing it, but you could flip a switch and use the spectrum to play MIDI based on the spectrum. So that, if you happen to have a microtonal synthesiser, and I did because I had my hacked up DX7, you could get the synthesiser to, say, speak a little or imitate some timbres or whatever, just through the same program RGS that you had already gotten used to. The good thing about RGS was you could … I used to stick a piece of clear acetate on the screen and copy a spectrum that I’d analysed and then erase the screen and draw the stuff in by hand. So you would hear the squeaky, blurpy stuff in a loop and eventually it would start making sense, and you could go and erase parts of the harmonic spectrum. It would all be running in real time, as a performance.

SM:  You worked a lot with Pauline Oliveros, when she was alive.

HL:  Yes, we were very, very active together.  I developed the program AUMI (Adaptive Use Musical Instrument) for her.

SM:  I’ve had AUMI on my iPads now for many years, but I am ashamed to say that I never really starting using it. Then, a few nights ago, I said I better learn this if I’m going to talk to Henry about it. Wow, what a program.

HL:  Oh, thanks. Leave a comment.

SM:  What I noticed about it when I first got it set up was I was about three hours into AUMI and I had AUMI with MIDI out and it was controlling thumb jam with a microtonal scale. I thought, “I could put this on my iPad 4, as well” because AUMI works with anything from IOS 9 up. The iPad 4, the highest it goes is IOS 10.3. It’s a venerable old machine that’s showing its age but still works fine. Then, my iPhone 6, of course, I can get AUMI on that. They all have thumb jam, so that I have three versions of it on three machines and was just wiggling my hands. I said to Catherine [the interviewer’s spouse – Ed.], “Have a look at this.” She whipped out her video camera and videoed it. Suddenly it was out in the world and I hadn’t even learned the program properly yet.  Here’s a 30 second excerpt of that performance: http://www.warrenburt.com/journal/2020/8/22/once-upon-a-midnight-dreary.html.

HL:  Well, I mean it has many objectives that are pretty much like that.

SM:  Yeah.

HL:  The main objective is to make sure that people who have the very smallest amount of voluntary movement can play it. They may not be able to set it up, but at that point they really can’t set up too much in their lives, anyway. But there are very interesting and heartening stories about AUMI in cases where people are kind of locked in or have had accidents and really want to play some kind of music again. People who use it as biofeedback or music therapy and physical therapy, so it’s got a lot of that. Then, as you noticed, I snuck a lot of my own musical stuff in there. Of course, it’s also being used by a lot of children in situations, so there are sounds in there that are not strictly musical. So there are sounds of percussion of pots and pans, dogs and cats and barnyard animals. Then I’ve also added sounds from mutual friends of ours, like Skip LaPlante and Dave Soldier playing stuff. Then a very extensive user of AUMI is Jesse Stewart, so I have him playing water phones and a few things like that. Then a lot of me. I stick a lot of me in there, because I don’t know anybody.

SM:  Yeah, the thing I noticed as I got deeper into the manual is that you can load in, if you give your samples the right names, you can load in whole sample sets. I haven’t gotten to using it that way yet, but that sounds really exciting. You could have a sample set where the samples were tuned to some microtonal scale and then load that in and set up a 49, a 7 by 7 grid, and away you go.

HL:  Well, the newest version, version 2.1, which will come out sometime, makes that process much, much, much easier. It doesn’t load anything standard, but basically it’s a zip file that’s full of samples in whatever format you like. As long as you follow the naming convention, it will either figure out that it’s a melodic one or they’re just crazy sounds. The crazy sounds you can order any way you like. The melodic ones, actually, because they’re user sounds, you can also order them any way you like.

SM:  You just have to name them in the right order.

HL:  Yeah, you just have to name them with the MIDI note name at the end of whatever you name each sample. I run from MIDI note 36 to 96, just so you have some wiggle room on the bottom and top, because I’ve made up some instruments that have both melodic sections and crazy sections so that I can mix it up a little bit. That’s all pretty simple to understand and then you can go with it a little further. It’s full of things that are pretty … they look like they don’t do much but in fact they help you out. If you made a sample set where the beginnings of the sample are pretty much straight forward and then the end of the sample is kind of crazy, like you play a long note and then you put a trill on the end or something like that, and you make a bunch of samples like that. You can set it so the cut off is about three tenths of a second and you can play it melodically and it’ll be fine. Or, if you wanted to mix things up, you could just play the whole sample. The tail ends of it will start mixing in with the beginnings of other sounds and things like that. It can get pretty sophisticated and yet still very simple.

SM:  Yeah, I also noticed that the IOS version that you’ve really refined is a step beyond the AUMI which is for the computer, for the desktop computer ( http://aumiapp.com/ ).

HL:  Right, the desktop version is written in Max and it’s been maintained over the years by a series of students up at McGill, so mine is a completely different code base, of course. So I can work on it whenever I want and however I want and because I’m not a student – I don’t go away. But we do have a concept that I’m working on now, which is to get it off of these hardware platforms at all and put it into a browser. I actually have a good start on that. The new version has a color tracker that I wrote. It’s very simple, it’s very stupid, and because it’s stupid you can understand what it’s doing. It’s also very fast. I put the color tracker into my java script version of it. This may show up sometime next year or whatever. I have to figure out which features I will keep and which ones I will leave out, because it would be nice to be nice and simple. Since it will run in HTML5 browsers that know about cameras and know about playing samples, and probably know about MIDI, it should be pretty easy to get this out into the world, which has always been our main thing with it. Be able to run on Chromebooks. I’m trying to think of what the cheapest device that might have a browser powerful enough to do that and enough memory to hold the samples. You could probably get away with that. Of course, it would run pretty nicely on, say, a Raspberry Pi or something. You could make a dedicated AUMI device for 100 bucks or something.

SM:  Yeah, I’ve got an old ASUS netbook, which works on Windows 8.1. It’s tiny, it’s 64 mighty gigabytes of RAM on a slot in card, and that’s it, but it does so much. Xoto Pad, that’s that two dimensional keyboard set-up program, I’ve been using that ( https://feelyoursound.com/xotopad/ ).  Add the Modartt Pianoteq and it’s a whole separate little instrument. I bought it about 7 years ago, so now it’s worth nothing. It works on batteries and you can detach it from the base so you just have a little tablet to play microtonal things on.

HL:  Well, as you mentioned, AUMI runs on everything from IOS 9 and up, so it runs on my iPad 3.

SM:  Very good.

HL:  I’ve got a room full of deprecated devices and, certainly, they can talk to each other. One other feature of AUMI that I like to point out is that, because it’s made to be played in ensembles, it has a very interesting feature that I wish other programs would have, which is you can set your instance of AUMI up to be a listener. Then, another one, run by, say, an administrator can set themselves up as the sender and it can send patches to any of the listeners using the Apple game kit. So basically they’re usually in wheelchairs with an AUMI pad and an iPad in some kind of mount stuck on a table somewhere. The two people who are managing the group are running around fixing things up. But here it can instantly send a fully configured patch to everyone in the room and they can all be on the same. You can put custom ones in each one, so that can all be managed from a central spot and done really fast, as opposed to fiddling with the controls endlessly trying to get it exactly right. So that’s something that I wish more other programs would work with.

SM:  Yeah, the other thing that’s really wonderful about it is the timing element, where you can set up all sorts of rhythmic variations which quantize your playing in various interesting ways.

HL:  Yeah, I’m surprised they don’t have … there are a lot of arpeggiators in a lot of synthesisers now and it’s very tough to keep up with them. Some of them are extremely sophisticated, but I tried to keep a couple of tasty variations in AUMI. I don’t think it’s in 2.0 but in 2.1, I added a thing which is sort of like cording, so when you have chosen a note, it will literally pick the quote third one and then you’ll see what a real third is, because I have so many different kinds of scales. You can play a third in all kinds of different scales, or a fourth or a fifth or an octave or whatever.

SM:  That’s wonderful.

HL:  It plays triads and stuff like that.

SM:  This version does do that. Yeah. Yeah.

HL:  Oh, okay. Good.

SM:  2.0, yeah that works. That’s got the triads and the pentads and so on. Yeah.

HL:   In 2.1, I took the pentads out. They were too noisy. But I put in other things. I put a little arpeggio in there and things like that. There may be more to that, but obviously the infrastructure that drives all that, if I exposed it, you could make all kinds of really crazy things. But I want to keep it kind of simple. For years, I sort of wanted to have this as the AUMI, the free AUMI, the AUMI that’s for the world. But then there would be another AUMI that is basically the same. It would take out some of the features for clinical use, the reporting. But it would put in connectivity to the IOS musical infrastructure, so I probably would have to run it as an audio unit at this point, which is horrible to code for. It’s just ghastly. Then it would open up some of the more hidden sides of it and regularise it and do stuff like that. But do I have time to do that? No. I have a lot of the code for it, but I don’t have it. There are many other apps of mine that are waiting with their little app mouths open like little birds in a nest, waiting to get a little insect or whatever I will drop in there. But I really like this. By the way, I’m pronouncing it AUMI, (“OW-MEE”). I always have, everybody pronounces it differently: “AU-OOMEE”, of course. Pauline pronounced it “OMI”, sort of like om mani padmi om. There are people who call it “The AUMI”, because it stands for Adaptive Use Musical Instrument. So everybody calls it something different, but I don’t care. Let a million pronunciations bloom.

SM:  Last time I heard a politician say something like that, I ducked.

HL:  Yeah. Well, I mean, but he didn’t really mean it. That’s part of the problem.

SM:  But you do mean it.

HL:  I do.

SM:  I’ve noticed how some apps of yours appear and disappear and then re-appear and appear in different forms and that’s because you’re working in a very free, say, non-business model driven way that you work with what excites you at the moment.

HL:  Yeah, it’s such a non-business model I have a hard time even getting enough out of these apps to pay my phone bill. But, at any rate, most of them are kind of thought experiments. Some of them do hark back to things I’d done in the distant past, which I miss. Of course my Amigas have gone to the great Amiga pile in heaven, although I have my emulated Amiga running here on the MacBook. It can actually do some of that stuff, but it’s clunky and it doesn’t have the hardware or the keyboard that I’m used to and things like that. So some of those ideas come back. Of course, I have a great, huge number of half written and unreleased apps that sometimes I talk about, sometimes I have video on, sometimes I don’t. I really want to bring some of those to life because I use them. I don’t care if other people use them, but I use them and they’re really something.

SM:  There’s also PolyHarp, which is one you’ve been working on and refining for many years. The current version is very, very complex. It allows you to do lots and lots of different things.

HL:  PolyHarp was an upgrade of my old Amiga program LYR which used the mouse to be a thumb pick, basically, on the Amiga screen for an autoharp. Then I said, “Well, what if I could … ?” – I always have the autoharp of my dreams, except my dream keeps changing. So this allowed me to make a 128-string autoharp that played MIDI. LYR actually has a feature that PolyHarp doesn’t have which is a strum sequencer. Because it only had one mouse, one touch, to do the kind of finger picking I wanted to do, you would have to set that up in a sequencer. But instead of sequencing notes, it sequenced only the strums. So you would change the chord if it’s live, but the strums would come through and they were, again, plucked in that way like a kind of arpeggiator sequencer combination. It was very good. Of course, you plug it into your microtonal synthesiser and away you go, right? Because everything I do is microtonal, right? Practically.

SM:  Yeah. Of course.

HL:  All my apps, they grudgingly say, “Oh, yeah, one of your choices is to do twelve equal divisions of an octave.” Yes, you can have that choice, but don’t stop there. So PolyHarp is really insanely dense. First of all, it inverts a lot of people’s idea of how music should be organised, so a lot of people say, “Oh, take a scale. We’ll figure out what that scale is and then we’ll pull chords out of that scale.” PolyHarp goes inside out. It says, “Let’s build chords, stick them somewhere” because harmony and melody are completely different things. So we’ll mark harmonies, or something like them, or nothing like them, and call that a chord, and give it some kind of name. Then we’ll stick it somewhere in audible space, or sometimes inaudible space because I like to do infrasound. Then, when you have a bunch of them together, you will have a bunch of strings. You may have a scale if they correspond, sort of. Or you might not, but scales are really quite irrelevant to what PolyHarp is about, because they’re kind of irrelevant on autoharps.

SM:  Exactly.

HL:  But they are there. Before we began, you mentioned Larry Polansky and one of the first microtonal things I did was one of his pieces called Will You Miss Me? Here’s an autoharp just to show you one. (holds up an autoharp to the camera) This one is tuned in what I call double diatonic. It’s in the key of D major and D minor. Most diatonics are on the circle of fifths but mine is on the circle of minor thirds. It sounds great. It’s really great. But, anyway, autoharps are a great place to work off of constraints of music theory. Of course, PolyHarp, it really wants you to be able to do good stuff, so it has a 1,000 string capacity. Every string in there, every quote note unquote in there could have several strings per chord. Usually you think twelve-string guitar, there’s two strings there. There are some kinds of instruments that have octave separations on those two strings. Well, in my case you can have up to seven strings faking being a single one. I don’t know why I picked seven. It just turned out I think that worked out for some number. Really rarely I use more than three. They can be spaced randomly within a certain cents limit or equally. So you can get all kinds of beating and chorus effects and none of them are done with an effects module. The sound coming out of the thing is completely clean, because I didn’t want to get into the effects business.

SM:  When I am reviewing software one of the things I point out if a synthesiser doesn’t have microtonal capability is that they’re actually losing a lot of sound design possibilities. Because even without playing in a microtonal scale, the use of fine tuning produces all sorts of timbral things that are quite useful and wonderful.

HL:  Well, you can explore a lot of that stuff with PolyHarp but it’s a lot easier to explore it with my really old app Droneo, which is still in the App Store. In fact, SrutiBox, Droneo’s older sister, I suppose, was the first app I stuck in the store. That was there when the App Store was about a month old and it’s still there.

SM:  Ah, very good.

HL:  So if anybody does a survey of what are the oldest apps in the App Store, it will probably pop up. I haven’t touched it for a while. It’s got some really old code in there but, nevertheless, it’s still around.

SM:  I love Droneo.

HL:  People have told me they love Droneo and they use it as the quote killer app. They say they’ve bought phones just to run Droneo which is very flattering, but, as they say, if you don’t leave a comment nobody knows. It’s not like I have a publicity department here in my kitchen.

SM:  Exactly. Another one that’s really wonderful because you just keep adding things to it is Enumero.

HL:  Oh, man. Nobody knows how great Enumero is. I use it constantly because …

SM:  I was in this improv group for about seven years and as soon as Enumero came out, it became one of my major things I used in the improv group.

HL:  Excellent.

SM:  Yeah, just to drop in, “Uno”.

HL:  Great. See, first of all, I don’t think there is any record of any place anywhere that counts in Latin. I thought that would be really good. I had four years of Latin. I know how to actually pronounce it. I think maybe people who are pronouncing the V’s as a V instead of a W, that really annoys me. I put a kind of Italian twist on it, but it’s OK. I decided I would go with the idea of counting and make it as crazy as I possibly could so I put in all kinds of things that I wanted to enumerate. I’ve put in the fantastic Conet project, Deutsche Voice, which is a spy radio network that just put out mysterious numbers over the air continuously for many years. I have chemical elements in Tom Lehrer’s order if you want. I’ve got 10,000 digits of Pi. Recently, and I had a tussle with Apple over this, I put in the entire COVID-19 genome.

SM:  Oh, wow.

HL:  But I couldn’t call it COVID-19. They kept pulling it out of the store and saying, “No, we don’t want any apps that are not medical apps that do anything with COVID-19” even though this is clearly not. So it’s called Genome, so wink wink, I’ll tell you which genome that is if you send me the dollar or whatever it costs to do Enumero.

SM:  Very good.

HL:  Because Enumero does fit in with the iOS music app ecosystem, I have a great time running it through all kinds of effects, whenever I get a new one. Thank goodness I’m a beta tester for Eventide. Oh, man, those people are pretty good friends of mine and they keep coming up with great stuff.

SM:  I have the complete Eventide app set and it’s just wonderful.

HL:  It’s really great. Obviously it’s really just specified patches that you can do in one Eventide box all by itself. But, even so, breaking it up into twelve apps or whatever they have so far is still very good for just getting your head straight. So those sound really good and it goes really well with Enumero. I like to run it through spectral transfer programs, so you can have its voice speaking through an orchestra or whatever. Yeah. Very useful. In fact, the section of it called Schwa I really want to just break out and build into its own audio unit. Then also make it adaptable so you’ll be able to record or input a long, long sample. It would do the analysis to turn it into a quote Schwa data set unquote and then have it pump out Schwa nonsense. I find that lyrics, like scales, are kind of overrated, and also rhythm, by the way. I’m really past that stage.

SM:  You were mentioning some things. What’s your most, say, immediate plan for the future besides surviving the COVID-19 horribleness?

HL:  Well, I do a lot of other things, but in terms of musical stuff … well, I have a lot of things. There’s that AUMI online. I’m really interested in this web app stuff and how far it can be pushed. There’s already some really good stuff out there that’s amazing. It’s not that I want to join the crowd, but it’s like this is a technology that I see as the future, so I’d really like to work with that more. A current web app is “The Lost Chord” ( https://jhhl.net/lostchord.html ). One of my projects, as I mentioned, I’m in the Catskill Mountain Gamelan. One of the problems with Gamelan is the actual written music is very different from western written music, at least my Sundanese version. So it’s got numbers instead of notes and they’re usually laid out on a piece of paper in a way that you have to look at it four times to realise what the timing is on it. So I’m building an online music system so that I can put these scores in, validate them and, it doesn’t do it yet, but obviously it will play them with samples of our own Gamelan so that I can hear what it would sound like. But it will let me build parts. That is, I can put the whole score in and then drop stuff out so only the bonang part or the jenglong part, or stuff like that. The main reason I did it is because I’m playing the suling, which has the worst parts. Because a lot of Gamelan music is very clock-like and mechanical and right on the beat, and then the suling and the voice, the suling is the bamboo flute, that’s very free on top. To get a solo in and figure out where to put it and where the timing is and to notate it is very, very difficult, especially on the kind of notation we have, which is sixth generation copies of music or something. So these are beautiful and clean. Then I can print out and leave some space for my suling solos to go in, which I would draw a graphic score on top of it or something like that. Anyway, so that’s one project that’s kind of on hold with everything else, but I will get back to it. There are other projects, like the successor to Droneo, which I’ve been trying to write for about eight years. I keep making attempts and then Apple’s infrastructure changes and my attempt just falls apart or there’s just too much UI squished on the screen so I have to figure it out. But I can describe a little bit what the aims are. Droneo not only sits there droning nicely on any intervals you like, but obviously we would like to drop all the limitations on that, so it won’t be limited to just eight oscillators, or drones, as I think they’re going to be called. But the generative part of it really has some really interesting concepts in it, which is it’s a graph of transitions. So there are a few transitional graph musical programs out there, but I would kind of make it manageable. Most of them kind of look like a really bad subway diagram if you’re going to do anything sophisticated, which is a general problem with graphical programming. I actually will probably be stealing ideas from my ancient animation program Vapor Paint, which you may or may not be familiar with, but I wrote it on the Amiga starting in 1989 and worked on it until the Amiga platform fell over and died. In the meantime, I had ported the renderer to see and ran it and can run it on any of these things. So the models I make up in Vapor Paint, I can render today at any resolution and any size and it’s really great. It was already way ahead of its time back then so it’s still pretty useful today. So the way that it organizes its rendering tree is kind of how this will organize this. Then, the idea is, instead of having a main base frequency that all the other oscillators bow to, oscillators will be listening to other oscillators for their frequency information and their triggering information. The triggers, there may or may not be a clock in there or the clock is, again, going to be an option just like many things are options. But mostly it’ll be able to pick up live pitch detection and I hope rhythm detection, that’s much tougher to do than pitch. Then, take those signals and say, “Oh, I’ve got something going on in the frequencies from 500 to 800 so that’s my trigger. I’m going to go and use that to be my transition for some generative process.” The other thing is the timbres. The timbres technically they shouldn’t exist. Everything in this is just going to be a sine wave so I can have several thousand of them and that’s how I will get my timbres. They will be listening to each other, too. So the timbre evolution will be part and parcel of the whole evolution of the entire musical structure, which is listening to itself and to other things.

SM:  Wow, that sounds fantastic.

HL:  That’s what that’s about.

SM:  That all sounds wonderful. I’d like to say I can’t wait, but I know I will wait and I’ll be very happy when it happens.

HL:  Well, that’s what I’ve been working on is the specifiers for your intervals. It’s already listening. I made it so that one pitch element saying, “Hi, I’ve just changed my pitch” and now the other ones are listening to it and applying the interval offset to whatever this other guy’s pitch is. They can all chain each other that way in that whole messaging and chaining system. You know, Max/MSP sends messages, right? Lots of things send messages.

SM: Yeah.

HL:  To have that built in instead of having everything hard-coded, everybody is sitting around waiting to be told what to do. In fact, the app itself is deep listening. It will be listening for the cues that it wants. It’s probably going to have to have some little language in there to tell it what to listen for. Or, it might go the way of the future which is machine learning. So you will give it some clues, move its little arms and legs in the positions it needs to do to assemble that clockwork. Then it will pick it up from there because machine learning chips are now in these devices. I might also be using my own machine learning scheme if I can get it together instead, because it could be easier to figure out and the errors will be errors I like, as opposed to errors that I scratch my head about. One of the things I often tell people is I don’t like magic, which is one of the reasons I’m not too hot on object systems. They’re full of magic. This just happens magically because some side effect is going on that you will never be able to understand. So I don’t like magic, I like assembly language, obviously. I’m probably not going to get all the way there, but having my own machine learning system would be very nice. I’d also like to point out that AUMI is free and has been for about three years. Originally we charged money for it as a hope to help fund the Pauline Oliveros Deep Listening Institute and people weren’t buying it. I dropped the price and people still weren’t buying it, even though it was vastly cheaper than most adaptive software and hardware for that market. So I said, “You know, this is keeping it out of people’s hands, so phooey. So take it.” It’s super sophisticated for all kinds of crazy musical projects. That’s my suggestion.

SM:  Fantastic. Well, thank you very much. This has been quite wonderful.

HL:  Thank you. | | – |