I am wondering when media companies will be convinced that people will not pay for lousy service and crappy content. Freedom is the only way forward and people WILL PAY FOR CONVENIENCE AND NOT CONTENT. If the artists tailor their approach to this, they benefit and when the drip turns into the flood THEY WIN. As much as I support the right of artists to be compensated fairly for their work, the entire system built to support them, screws every person who does not sit in the middle, artist and consumer, people who should be peers and are not. Who has not felt raped by Ticketmaster, or Sony (embedding theft protection malware in the audio CD) or the horrors of DRM.
Some artists get this, but I fear most are too lazy or too stupid to understand that with action (concerted to be sure) on their part, we have reached the golden age.
Friday, April 03, 2009
Thursday, March 19, 2009
Adaptive Signal Processing, Polyphase Filter Banks, for the old and new
Recently I have spent enough time on adaptive signal processing algorithms to have driven my friends crazy (yes, I do have friends). Everyone wants the polyphase filter bank code. I completely sympathize. While I am not a Dionysius (a Greek king), I invite you to be Damocles and feel the sword as you sit in my chair for a few sentences and then I won't bore you any more with this.
The polyphase filter bank code is "done". It has been for months. The implementation of the filter banks in question? A completely trivial matter for someone with as much digital signal processing experience as I have suffered, err uhh, enjoyed. There is a problem. The problem is significant and I am absolutely determined to solve it. Polyphase filter banks should be everywhere, ubiquitous if you will. They should be easily used, and easy to get right. This is a major problem but it really is the only one I am interested in pursuing to the end. I find that all of the major players, from my new friends and colleagues, to those I have never had the pleasure of meeting, have failed to address an overall structure.
The structure in question is how to do all of this in a completely automated way. You give me a few design parameters, ones that are obvious, I give you back a pointer to a filter bank object and from that moment on, you give me signal, and I give you output. What is difficult about this? The filter bank's signal performance is completely determined by the filters and the filter performance is determined by the size of the computer word. If you ask me to implement a filterbank with brick wall edges and 100 dB adjacent channel suppression, the filter design and the flow of signal through the filter bank needs new work done by fred harris and I (soon to be published). If you need perfect reconstruction with these parameters (the channelization is reversible to high accuracy through a synthesis operation), I return with a "you are a fool" and tell you the best you can hope for is ...... The nature of the filters inside depends on whether or not you want linear phase or not and .......
The construction of all of this filter determination code has been the object of study for brains better than mine for decades (Julius O. Smith comes to mind) but it is not all collected into one object. I am determined to crack this nut once and for all. Irrespective of how much respect I have for my filter bank colleagues, they have written enough papers on one-off designs. They have charged enough money for one-off consulting. Through the power of doing this right and then giving it away GPL and probably publishing a treatise, I am determined to make this a sea change, once and for all.
The work done by fred harris and I needs publication approval from my masters. These are the ones that feed my children, send them to school, and allow me to drive a fancy sports car and go to Broadway twice a month. The approval is coming. The filter bank FILTER DESIGN work is being done and we will all have it soon.
Thank you for your patience. Understand that I do the other things knowing they are not as important as the filter bank code. They are nevertheless necessary. They show off the differences between these sdr signal processing chains and what is possible in traditional "narrow band high performance radios" to the SIGNIFICANT advantage of the SDR's. I may be mad, but there is method to my aggravating pace.
The polyphase filter bank code is "done". It has been for months. The implementation of the filter banks in question? A completely trivial matter for someone with as much digital signal processing experience as I have suffered, err uhh, enjoyed. There is a problem. The problem is significant and I am absolutely determined to solve it. Polyphase filter banks should be everywhere, ubiquitous if you will. They should be easily used, and easy to get right. This is a major problem but it really is the only one I am interested in pursuing to the end. I find that all of the major players, from my new friends and colleagues, to those I have never had the pleasure of meeting, have failed to address an overall structure.
The structure in question is how to do all of this in a completely automated way. You give me a few design parameters, ones that are obvious, I give you back a pointer to a filter bank object and from that moment on, you give me signal, and I give you output. What is difficult about this? The filter bank's signal performance is completely determined by the filters and the filter performance is determined by the size of the computer word. If you ask me to implement a filterbank with brick wall edges and 100 dB adjacent channel suppression, the filter design and the flow of signal through the filter bank needs new work done by fred harris and I (soon to be published). If you need perfect reconstruction with these parameters (the channelization is reversible to high accuracy through a synthesis operation), I return with a "you are a fool" and tell you the best you can hope for is ...... The nature of the filters inside depends on whether or not you want linear phase or not and .......
The construction of all of this filter determination code has been the object of study for brains better than mine for decades (Julius O. Smith comes to mind) but it is not all collected into one object. I am determined to crack this nut once and for all. Irrespective of how much respect I have for my filter bank colleagues, they have written enough papers on one-off designs. They have charged enough money for one-off consulting. Through the power of doing this right and then giving it away GPL and probably publishing a treatise, I am determined to make this a sea change, once and for all.
The work done by fred harris and I needs publication approval from my masters. These are the ones that feed my children, send them to school, and allow me to drive a fancy sports car and go to Broadway twice a month. The approval is coming. The filter bank FILTER DESIGN work is being done and we will all have it soon.
Thank you for your patience. Understand that I do the other things knowing they are not as important as the filter bank code. They are nevertheless necessary. They show off the differences between these sdr signal processing chains and what is possible in traditional "narrow band high performance radios" to the SIGNIFICANT advantage of the SDR's. I may be mad, but there is method to my aggravating pace.
Sunday, March 08, 2009
Image Rejection along A Road Less Traveled.
"I shall be telling this with a sigh
Somewhere ages and ages hence:
Two roads diverged in a wood, and I--
I took the one less traveled by,
And that has made all the difference" (Frost)
Some time ago we began the process of studying, learning, suffering, failing, and then succeeding that always accompanies the creative process. Many wrong paths were taken. Life always intervenes as ruts in any roads taken. I write this for all those who have suffered with us, demonstrating the utmost patience and most especially to tell our long suffering European friends (with HF broadcaster images) that relief has arrived.
We have known for some time the mathematics of fixing the image suppression due to amplitude and phase imbalance in the QSD. Alex Shavkoplyas has done a fantastic job of following down one road in Rocky. It is a beautiful thing and my favorite cattle prod, Phil Harman, asked me very pointedly "What is the matter with you on this?" I explained and it immediately began to gnaw at me how can we do this. My apologies in advance for it haven taken so long to reach a place of an acceptable algorithm. I wrote a paper that I put in DCC (replete with errors). I followed that up with an article for QEX. I am grateful it was never published. Now the good one will be.
Unlike Rocky, we cannot take more than seconds, much less minutes and hours to converge to a useful image suppression result. Once rocky learns on a particular piece of hardware with its ROCK-bound frequency, it only needs to tweak for changes in temperature, age of crystal components, etc. You the user would be quite unhappy if it were commanded by Big Brother (Flex, DttSP) that you could not tune your radio but once a day! I suspect that amateur radio operator Winston Smith would be in rebellion and Gerald "OBrien" Youngblood could not even torture you to accept it. I know my Julia (Shann N2HPE) has suffered the torture of this with me. I am eternally grateful for her partnership. The ministry of truth has convinced you to be patient. For this I thank you.
Knowing the theory and mathematical equations necessary to accomplish a task does not make it easy to accomplish this in an algorithm. Image present is the result of nonlinearities in the QSD, RFIC, or any SSB mixer (IQ mixer). So nonlinearity must be present in the correct fix up. Nonlinearity introduced on purpose in a system hailed by all for the linearity and signal handling capability of its front end seems at best contradictory, at worst, immolation! Nevertheless, it is required. The application of these nonlinear processes must be done with some care.
Our first attempts at deriving an algorithm that would work in DttSP and PowerSDR succeeded. But they were calibration dependent. Calibration at the factory means man hours and increasing costs to Flex the manufacturer. We accepted the less than acceptable status quo until a month ago because we just didn't have a useful plan to do otherwise. Doubts, internal as well as external, began to consume. The realist among us went "who cares". The idealist averred as to how this was unacceptable. Shouldn't there be a middle way?
Traveling to Austin to work on an implementation of the image suppression algorithm became the top of the list of myriad demands on my time. The top of the heap being the work that is being done by Flex for my employer and sponsor. Knowing the mathematics as described above allows one to write down an iterative procedure where sequential estimation is used to estimate the solution (based on approximations to Newton's algorithm in an application of the fixed point theorem to real life).
The TEN LINES OF CODE were written in an hour. They were working immediately. It was clear that we needed to use something like this but it was unclear how to use it in the system as a whole. It will cause some dislocation to manufacturing, require dislocation to calibration, and require dislocation to the user access to those pieces that the user, heretofore, had come to expect. Klaus Lohman, in expert testing, pointed out an issue which seems troubling but in the end, isn't. But this had to be addressed. We learned that some of the parameters of our search algorithm were wrong. These resulted in occasional numerical instability. We did not have a plan on how to roll this out at all, but we knew we must.
Last week I returned to Austin and the magic of working in a partnership filled with respect but constant questioning allowed by mutual trust, Eric and I worked out a way to use this in any system that would attempt to use it. It is not enough to have the right algorithm in DttSP. The correct application is dependent ON THE USER not the algorithm designer so it requires knowledge of how one correctly uses it in a system for it to be effective.
So eight weeks ago, I added the math to DttSP v3.0. Four weeks ago I put a version in PowerSDR. Last week Eric Wachsmann and I made it practical. The practical comes in asking simply how can this mathematical magic be made useful to the user. It turned out to be remarkably simple. When you tune the radio and you tune it far enough that you change the synthesizer frequency, the algorithm starts a retraining cycle. This must be made stable. So, it starts VERY fast so that the image is almost suppressed completely in about 50ms but then it begins a step by step decrease in gain in the algorithm and at the end of the about 500ms, it has converged. So we can just turn it off then right? Wrong. The algorithm even works on noise BUT its answer is off because loud signal is a better input than noise. So we could not turn the algorithm off completely. We learned we need to leave it turned on at a low convergence rate. Then we ran into the problem that again, you Winston Smith, are not going to allow OBrien dictate to you what to do and think. You are going to tune rapidly and foul this learning algorithm up and the stack would run dry! That too turned out to be relatively easy to fix. If you change the DDS, the convergence algorithm is stopped and restarted and that is done in a thread safe manner.
Now that this lengthy TMI (too much information) account is done, we encourage you to use it. This information is provided for one reason only. So you who have suffered the long wait with us understand something of the creative process and why that MUST be married with the practical.
branches/n4hy/iqtest/bin/Release contains the new code.
I have been experimenting with fixes to the ANF and NR algorithm and I have fixed the complex but separable versions in this branch as well. The ANF and NR is most decidedly not stable code as it will continue to receive work all week long until we have convergence to the acceptable algorithm there as well.
The necessary steps to see this in the trunk must be done in stages. We need resampling in the IQ processing chain. We need to turn the receive training algorithm off when you are transmitting so that transmit training algorithm will run with the benefit of the perfectly adjusted receiver. All of this must be seamless. We learn from experience that you are mostly uninterested in how, you just want the damn thing to work.
Following this, the entire IQ imbalance conversation concerning Flex, QSD, etc. should be long gone. This will be useful for RFIC's, QSD, etc. It will work on RX and TX if you have full duplex hardware (Flex 3000,5000, RFX boards for GnuRadio). It will work on RX on the oldest and meanest of our hardware applications (for example, it will work on SDR-1000) and an appliqué for Flex 5000 and Flex 3000 will allow it to train the transmitter image to the noise floor in YOUR OWN RECEIVER not to mention your ionospheric neighbors.
I will tell you that success is sweet. But I am entirely aware of how long it took. I reject any gratitude because at the end of the process I am really quite critical of my having taken so long to arrive at what, in the end, amounts to no more than 20 lines of code and I had to have excellent help getting there.
Somewhere ages and ages hence:
Two roads diverged in a wood, and I--
I took the one less traveled by,
And that has made all the difference" (Frost)
Some time ago we began the process of studying, learning, suffering, failing, and then succeeding that always accompanies the creative process. Many wrong paths were taken. Life always intervenes as ruts in any roads taken. I write this for all those who have suffered with us, demonstrating the utmost patience and most especially to tell our long suffering European friends (with HF broadcaster images) that relief has arrived.
We have known for some time the mathematics of fixing the image suppression due to amplitude and phase imbalance in the QSD. Alex Shavkoplyas has done a fantastic job of following down one road in Rocky. It is a beautiful thing and my favorite cattle prod, Phil Harman, asked me very pointedly "What is the matter with you on this?" I explained and it immediately began to gnaw at me how can we do this. My apologies in advance for it haven taken so long to reach a place of an acceptable algorithm. I wrote a paper that I put in DCC (replete with errors). I followed that up with an article for QEX. I am grateful it was never published. Now the good one will be.
Unlike Rocky, we cannot take more than seconds, much less minutes and hours to converge to a useful image suppression result. Once rocky learns on a particular piece of hardware with its ROCK-bound frequency, it only needs to tweak for changes in temperature, age of crystal components, etc. You the user would be quite unhappy if it were commanded by Big Brother (Flex, DttSP) that you could not tune your radio but once a day! I suspect that amateur radio operator Winston Smith would be in rebellion and Gerald "OBrien" Youngblood could not even torture you to accept it. I know my Julia (Shann N2HPE) has suffered the torture of this with me. I am eternally grateful for her partnership. The ministry of truth has convinced you to be patient. For this I thank you.
Knowing the theory and mathematical equations necessary to accomplish a task does not make it easy to accomplish this in an algorithm. Image present is the result of nonlinearities in the QSD, RFIC, or any SSB mixer (IQ mixer). So nonlinearity must be present in the correct fix up. Nonlinearity introduced on purpose in a system hailed by all for the linearity and signal handling capability of its front end seems at best contradictory, at worst, immolation! Nevertheless, it is required. The application of these nonlinear processes must be done with some care.
Our first attempts at deriving an algorithm that would work in DttSP and PowerSDR succeeded. But they were calibration dependent. Calibration at the factory means man hours and increasing costs to Flex the manufacturer. We accepted the less than acceptable status quo until a month ago because we just didn't have a useful plan to do otherwise. Doubts, internal as well as external, began to consume. The realist among us went "who cares". The idealist averred as to how this was unacceptable. Shouldn't there be a middle way?
Traveling to Austin to work on an implementation of the image suppression algorithm became the top of the list of myriad demands on my time. The top of the heap being the work that is being done by Flex for my employer and sponsor. Knowing the mathematics as described above allows one to write down an iterative procedure where sequential estimation is used to estimate the solution (based on approximations to Newton's algorithm in an application of the fixed point theorem to real life).
The TEN LINES OF CODE were written in an hour. They were working immediately. It was clear that we needed to use something like this but it was unclear how to use it in the system as a whole. It will cause some dislocation to manufacturing, require dislocation to calibration, and require dislocation to the user access to those pieces that the user, heretofore, had come to expect. Klaus Lohman, in expert testing, pointed out an issue which seems troubling but in the end, isn't. But this had to be addressed. We learned that some of the parameters of our search algorithm were wrong. These resulted in occasional numerical instability. We did not have a plan on how to roll this out at all, but we knew we must.
Last week I returned to Austin and the magic of working in a partnership filled with respect but constant questioning allowed by mutual trust, Eric and I worked out a way to use this in any system that would attempt to use it. It is not enough to have the right algorithm in DttSP. The correct application is dependent ON THE USER not the algorithm designer so it requires knowledge of how one correctly uses it in a system for it to be effective.
So eight weeks ago, I added the math to DttSP v3.0. Four weeks ago I put a version in PowerSDR. Last week Eric Wachsmann and I made it practical. The practical comes in asking simply how can this mathematical magic be made useful to the user. It turned out to be remarkably simple. When you tune the radio and you tune it far enough that you change the synthesizer frequency, the algorithm starts a retraining cycle. This must be made stable. So, it starts VERY fast so that the image is almost suppressed completely in about 50ms but then it begins a step by step decrease in gain in the algorithm and at the end of the about 500ms, it has converged. So we can just turn it off then right? Wrong. The algorithm even works on noise BUT its answer is off because loud signal is a better input than noise. So we could not turn the algorithm off completely. We learned we need to leave it turned on at a low convergence rate. Then we ran into the problem that again, you Winston Smith, are not going to allow OBrien dictate to you what to do and think. You are going to tune rapidly and foul this learning algorithm up and the stack would run dry! That too turned out to be relatively easy to fix. If you change the DDS, the convergence algorithm is stopped and restarted and that is done in a thread safe manner.
Now that this lengthy TMI (too much information) account is done, we encourage you to use it. This information is provided for one reason only. So you who have suffered the long wait with us understand something of the creative process and why that MUST be married with the practical.
branches/n4hy/iqtest/bin/Release contains the new code.
I have been experimenting with fixes to the ANF and NR algorithm and I have fixed the complex but separable versions in this branch as well. The ANF and NR is most decidedly not stable code as it will continue to receive work all week long until we have convergence to the acceptable algorithm there as well.
The necessary steps to see this in the trunk must be done in stages. We need resampling in the IQ processing chain. We need to turn the receive training algorithm off when you are transmitting so that transmit training algorithm will run with the benefit of the perfectly adjusted receiver. All of this must be seamless. We learn from experience that you are mostly uninterested in how, you just want the damn thing to work.
Following this, the entire IQ imbalance conversation concerning Flex, QSD, etc. should be long gone. This will be useful for RFIC's, QSD, etc. It will work on RX and TX if you have full duplex hardware (Flex 3000,5000, RFX boards for GnuRadio). It will work on RX on the oldest and meanest of our hardware applications (for example, it will work on SDR-1000) and an appliqué for Flex 5000 and Flex 3000 will allow it to train the transmitter image to the noise floor in YOUR OWN RECEIVER not to mention your ionospheric neighbors.
I will tell you that success is sweet. But I am entirely aware of how long it took. I reject any gratitude because at the end of the process I am really quite critical of my having taken so long to arrive at what, in the end, amounts to no more than 20 lines of code and I had to have excellent help getting there.
Thursday, March 05, 2009
Beagle Board in a case
Dear reader:
As you know from my ramblings, there is great interest in achieving high performance with low power in an easily programmable chip in order to do software radio. That is, I can order it to be an HD radio, FM stereo receiver, police band scanner, HDTV receiver, etc. just by changing the software.
While this may seem extreme, rapid progress IS being made.
Texas Instruments released the OMAP family and many of us have become interested in pursuing the bleeding edge of technologies, including the OMAP family. In particular many of us have built up systems using the Beagle Board. The Beagle Board is nice, but you still have to download a bill of materials, screw driver the thing together, and pray it will do the right thing.
Stop the presses, this is no longer required. Though one can put together a system that is less expensive than the Always Innovating offerings, one would still have to buy parts, assemble, and then cross your fingers. With the Always Innovating computers, this is no longer necessary. Go check them out.
As you know from my ramblings, there is great interest in achieving high performance with low power in an easily programmable chip in order to do software radio. That is, I can order it to be an HD radio, FM stereo receiver, police band scanner, HDTV receiver, etc. just by changing the software.
While this may seem extreme, rapid progress IS being made.
Texas Instruments released the OMAP family and many of us have become interested in pursuing the bleeding edge of technologies, including the OMAP family. In particular many of us have built up systems using the Beagle Board. The Beagle Board is nice, but you still have to download a bill of materials, screw driver the thing together, and pray it will do the right thing.
Stop the presses, this is no longer required. Though one can put together a system that is less expensive than the Always Innovating offerings, one would still have to buy parts, assemble, and then cross your fingers. With the Always Innovating computers, this is no longer necessary. Go check them out.
Saturday, February 28, 2009
Imogen and Web 2.0
A long awaited album entitled Polyfilla is almost finished for Imogen Heap. Thank goodness she is active and sharing eye candy and other tunes such as
Blanket (with Urban Species)
One gets the feeling in the undercurrents in her Web 2.0 submissions on blog, twitter, myspace, facebook, twitpix, 12seconds and so much more that she thrives on the slightly more than tenuous connection and interaction with her fans. For us it is a LOT better than a quick autograph from a sweaty hand, tired from the act, only wanting to escape our clutches unscathed. She has volunteered all of this surely for her own motives. But so what? It serves some of our needs. That is the give and take. It is interactive to the extent they wish it to be.
The clear message in the sharing with her followers is that she derives not only pleasure or amusement, but some kind of connection of her psyche to those who interact and follow. Without the pain of supporting the demands and needs of a lover or demands for time from a friend or acquaintance, this large collection of other psyche's constitutes a sort of muse or maybe just a release valve; helpful in either case it seems.
She is not the only artist, writer, actor, politician or just good old public figure to find this an inspirational or emotionally helpful interaction. Stephen Fry, Neal Gaiman, and even Jane Fonda all seem to derive exactly the same thing from twitter. All of them using the laptops or smartphones can twitter, blog, and share pictures with all who follow them minute by minute, blow by blow.
This connection, as tenuous as it is, is like the torrent going over Niagara Falls in comparison to the silent unheard voices of what surely used to be their isolation inside their circle of handlers or neurotic needy friends and associates. They seemingly do it because they appear to derive something they need from it and further, they personally control the level and degree without a NO NO NO publicity agent, lawyer, company, their phobias or neuroses, etc. meddling with their heads and needs.
We have watched Stephen Fry twitter his way to great acclaim in several venues, lose weight, and become 10 years younger before our eyes. Neil Gaiman shares freely in his achievements, one after another, and his family life with his daughter, loyal assistant and even all the way down to the sufferings and redemption of his own favorite white German shepherd. Gaiman feeds our voyeuristic cravings for the tidbits of our idols and in return, gets one of the greatest of marketing tools ever. He uses it as such with a true sense of decorum by twittering, blogging and more, aided by the word of mouth of fan boys like me Have you seen what is going on with Gaiman?" Their exhibitionism and our voyeurism, sanitized with these tools, like a condom if you will giving excellent safe fan-dom, allows us to share in their blow by blow to fulfill our needs. In turn, we derive entertainment, information, inspiration, or something else entirely as needed. We shower them with praise, sympathy, adoration, disgust, whatever ... to our hearts content, so long as it is 140 characters or less! No long winded, uncontrolled, fan girl squealing in their ears and seeping gushing noxious fumes of suffocating fandom tinged with jealousy.
If it all becomes too much or they lose focus, the Imogen solution is a simple click away: "Sorting out video/ live/ budgets/ schedule/ meetings...shite! I haven't finished it yet! Need to refocus, shut it out and make music today." After feeding the battery, it is charged and can be put to use off the mains without connection to the fuel the masses have been providing until it is time for recharging again.
Watch this and see the creation process in action on Polyfilla.
This sharing and the feedback which follows in other media mentioned mitigate in favor of the premise put forward here quite clearly. Then it dawns on you, this is just one artist (as fabulous as she is). Then you remember that even septuagenarian actresses are doing the Web 2.0 dance. Sharing with us their illnesses, joys, glee at those other famous friends attending her new play . Hmm. This is getting interesting. So..... What about ..... so and so? Hit Google and find them. Search for them on twitter, facebook, myspace, or whatever. It is a revolution and a great one. But wait, I am doing it here (not that many care but it is fun)!
Web 2.0 is the potential of the internet finally beginning to reveal itself in a glorious way. It is oh so much better than television because it is interactive to a much greater degree. It is not (yet?) quite as satisfying as feeling the insides of Raskolnikov or Humbert Humbert rot before you in your mind's eye. And it certainly is not based on decades of writing, no, struggling whilst neglecting Nora and the kids, just so Bloom can live a single day in our imagination for the rest of our lives. You don't need to read well, have a great vocabulary, or even just the time to read the novel .... but it is fascinating, entertaining. It is getting there.
Blanket (with Urban Species)
One gets the feeling in the undercurrents in her Web 2.0 submissions on blog, twitter, myspace, facebook, twitpix, 12seconds and so much more that she thrives on the slightly more than tenuous connection and interaction with her fans. For us it is a LOT better than a quick autograph from a sweaty hand, tired from the act, only wanting to escape our clutches unscathed. She has volunteered all of this surely for her own motives. But so what? It serves some of our needs. That is the give and take. It is interactive to the extent they wish it to be.
The clear message in the sharing with her followers is that she derives not only pleasure or amusement, but some kind of connection of her psyche to those who interact and follow. Without the pain of supporting the demands and needs of a lover or demands for time from a friend or acquaintance, this large collection of other psyche's constitutes a sort of muse or maybe just a release valve; helpful in either case it seems.
She is not the only artist, writer, actor, politician or just good old public figure to find this an inspirational or emotionally helpful interaction. Stephen Fry, Neal Gaiman, and even Jane Fonda all seem to derive exactly the same thing from twitter. All of them using the laptops or smartphones can twitter, blog, and share pictures with all who follow them minute by minute, blow by blow.
This connection, as tenuous as it is, is like the torrent going over Niagara Falls in comparison to the silent unheard voices of what surely used to be their isolation inside their circle of handlers or neurotic needy friends and associates. They seemingly do it because they appear to derive something they need from it and further, they personally control the level and degree without a NO NO NO publicity agent, lawyer, company, their phobias or neuroses, etc. meddling with their heads and needs.
We have watched Stephen Fry twitter his way to great acclaim in several venues, lose weight, and become 10 years younger before our eyes. Neil Gaiman shares freely in his achievements, one after another, and his family life with his daughter, loyal assistant and even all the way down to the sufferings and redemption of his own favorite white German shepherd. Gaiman feeds our voyeuristic cravings for the tidbits of our idols and in return, gets one of the greatest of marketing tools ever. He uses it as such with a true sense of decorum by twittering, blogging and more, aided by the word of mouth of fan boys like me Have you seen what is going on with Gaiman?" Their exhibitionism and our voyeurism, sanitized with these tools, like a condom if you will giving excellent safe fan-dom, allows us to share in their blow by blow to fulfill our needs. In turn, we derive entertainment, information, inspiration, or something else entirely as needed. We shower them with praise, sympathy, adoration, disgust, whatever ... to our hearts content, so long as it is 140 characters or less! No long winded, uncontrolled, fan girl squealing in their ears and seeping gushing noxious fumes of suffocating fandom tinged with jealousy.
If it all becomes too much or they lose focus, the Imogen solution is a simple click away: "Sorting out video/ live/ budgets/ schedule/ meetings...shite! I haven't finished it yet! Need to refocus, shut it out and make music today." After feeding the battery, it is charged and can be put to use off the mains without connection to the fuel the masses have been providing until it is time for recharging again.
Watch this and see the creation process in action on Polyfilla.
This sharing and the feedback which follows in other media mentioned mitigate in favor of the premise put forward here quite clearly. Then it dawns on you, this is just one artist (as fabulous as she is). Then you remember that even septuagenarian actresses are doing the Web 2.0 dance. Sharing with us their illnesses, joys, glee at those other famous friends attending her new play . Hmm. This is getting interesting. So..... What about ..... so and so? Hit Google and find them. Search for them on twitter, facebook, myspace, or whatever. It is a revolution and a great one. But wait, I am doing it here (not that many care but it is fun)!
Web 2.0 is the potential of the internet finally beginning to reveal itself in a glorious way. It is oh so much better than television because it is interactive to a much greater degree. It is not (yet?) quite as satisfying as feeling the insides of Raskolnikov or Humbert Humbert rot before you in your mind's eye. And it certainly is not based on decades of writing, no, struggling whilst neglecting Nora and the kids, just so Bloom can live a single day in our imagination for the rest of our lives. You don't need to read well, have a great vocabulary, or even just the time to read the novel .... but it is fascinating, entertaining. It is getting there.
Sunday, February 22, 2009
Saturday, February 21, 2009
The King (Honda S2000) is Dead. Long Live the King
Honda has decided to stop producing the S2000.
You may see here this is before I put dustless pads on it (dust from standard brake pads is seen to discolor the beautiful rims).
The S2000 story is a bit of fairy tail apocrypha. The anecdotes go that the engineers wanted to take the V-Tec engine and make a hand built serious sports car for the Honda enthusiasts. They were allowed to and in 1999 they sold a few. There grew a ground swell of demand for the award winning, high performance, fantastic driving, hand assembled car that sold for UNDER $40000.
My favorite single day in it was driving to see my mother in Alabama. I was coming from Orlando and this was a "pit stop", only slightly out of the way, on my way back to NJ.
I hit an infamous many mile stretch of back road that is about as straight as they come. I hit 150 mph and I was still accelerating and not at red line. It felt like I was doing 70 on the interstate.
I love this car as I never believed I could love a car. The story goes that Honda lost money on every one they ever built. So it makes sense that it is a 10 year vanity statement for Honda that needs to go, especially in these times.
110000 were made and sold, last month they sold 1000-ish world wide. Of the total, over half were to the USA (surprise!). Mine will rot into the ground before I give it up since that is how I have used every automobile I've ever purchased. Those would be the ones that I gave my kids when they wanted a new BMW convertible!
You may see here this is before I put dustless pads on it (dust from standard brake pads is seen to discolor the beautiful rims).
The S2000 story is a bit of fairy tail apocrypha. The anecdotes go that the engineers wanted to take the V-Tec engine and make a hand built serious sports car for the Honda enthusiasts. They were allowed to and in 1999 they sold a few. There grew a ground swell of demand for the award winning, high performance, fantastic driving, hand assembled car that sold for UNDER $40000.
My favorite single day in it was driving to see my mother in Alabama. I was coming from Orlando and this was a "pit stop", only slightly out of the way, on my way back to NJ.
I hit an infamous many mile stretch of back road that is about as straight as they come. I hit 150 mph and I was still accelerating and not at red line. It felt like I was doing 70 on the interstate.
I love this car as I never believed I could love a car. The story goes that Honda lost money on every one they ever built. So it makes sense that it is a 10 year vanity statement for Honda that needs to go, especially in these times.
110000 were made and sold, last month they sold 1000-ish world wide. Of the total, over half were to the USA (surprise!). Mine will rot into the ground before I give it up since that is how I have used every automobile I've ever purchased. Those would be the ones that I gave my kids when they wanted a new BMW convertible!
Many roads to go down, I chose one, and that made all the difference
May Frost forgive me.
I have written in these pages about the Intel ATOM 330 and its instantiation in an Intel Motherboard, the D945GCLF and D945GCLF2. Both of these boards have really good performance for a mobo costing $90 or less!
Always, there comes along a better mousetrap. Nvidia has done itself proud it would seem with the ION. The ION addresses what are, in my opinion, the worst shortcomings of the Intel mobo offerings
The ION has DDR3, 2 GB built in. The D945GCLFx has a slot for memory and will only use DDR2.
The ION has GeoForce 9400M. The Intel has Intel 945 graphics chip. The 9400M is by all sorts of measurements "ten times faster".
The ION mobo and small box have much better connector support for IO and have only external SATA drive support. This is better.
Some other recent work shows that the JFS appears to seriously outperform ext3 for Linux file systems in many ways. Since I want to upgrade my mass storage drive on my (current) main Linux development system to 1.5+ TB from 250 MB and an older drive using EXT3, I will do the work there to test. Should that work, I will back up the home directories and other /usr/local and /opt directories and copy the main drive to another large disk using JFS. There appears to be much better dynamic control over inodes in JFS rather than attempting one size fits all (pun intended).
We are back in old man winter after a brief respite. This makes it easy not to do outside work that needs doing and to read another novel. I could not remember much of A Separate Peace by John Knowles so I decided to polish off this novel(la?) yesterday and today.
Soeaking of today, Tom Rondeau and I made a pitch to Dr. Saltman, director of CCR and Dr. Boyack, director of computing CCR about two labs we need to efficiently do our software radio and cogntive radio work. They bought all of the arguments and told us to draw up a list of things we need, procedures for transferring code to and fro, etc. It was a nice meeting.
Frank Brickle gave a nice talk today at work on SDR, CR, VR, ... etc. It was great.
I have written in these pages about the Intel ATOM 330 and its instantiation in an Intel Motherboard, the D945GCLF and D945GCLF2. Both of these boards have really good performance for a mobo costing $90 or less!
Always, there comes along a better mousetrap. Nvidia has done itself proud it would seem with the ION. The ION addresses what are, in my opinion, the worst shortcomings of the Intel mobo offerings
The ION has DDR3, 2 GB built in. The D945GCLFx has a slot for memory and will only use DDR2.
The ION has GeoForce 9400M. The Intel has Intel 945 graphics chip. The 9400M is by all sorts of measurements "ten times faster".
The ION mobo and small box have much better connector support for IO and have only external SATA drive support. This is better.
Some other recent work shows that the JFS appears to seriously outperform ext3 for Linux file systems in many ways. Since I want to upgrade my mass storage drive on my (current) main Linux development system to 1.5+ TB from 250 MB and an older drive using EXT3, I will do the work there to test. Should that work, I will back up the home directories and other /usr/local and /opt directories and copy the main drive to another large disk using JFS. There appears to be much better dynamic control over inodes in JFS rather than attempting one size fits all (pun intended).
We are back in old man winter after a brief respite. This makes it easy not to do outside work that needs doing and to read another novel. I could not remember much of A Separate Peace by John Knowles so I decided to polish off this novel(la?) yesterday and today.
Soeaking of today, Tom Rondeau and I made a pitch to Dr. Saltman, director of CCR and Dr. Boyack, director of computing CCR about two labs we need to efficiently do our software radio and cogntive radio work. They bought all of the arguments and told us to draw up a list of things we need, procedures for transferring code to and fro, etc. It was a nice meeting.
Frank Brickle gave a nice talk today at work on SDR, CR, VR, ... etc. It was great.
Sunday, February 08, 2009
White Space (NOT a rascist remark)
White Space is the term broadly applied by technologist's and the FCC to the vast spectrum spaces, especially in urban areas, which are wastelands. For now, we will concentrate on the frequencies currently allocated solely to television stations. Since there are only a few channels occupied in even the most populous cities, coupled with the fact that the primary propagation mode for these frequencies is line of sight, much of the spectrum for TV channels is completely unoccupied even in these cities.
We need spectrum to increase the amount available to wireless devices. The FCC realizes this and in November of 2008, issues rules allowing this. CTVR at Trinity College Dublin has been interested in this concept for a while. A prominent DARPA project manager, Preston Marshall, a leading person in DySpan has long been interested, personally and professionally in White Spaces (a Dynamic Spectrum hot spot issue, pun intended). Preston is getting his Ph. D. from CTVR.
Tom Rondeau, hired by my employer, with my being the primary instigator, finished his Ph.D. thesis for Charles Bostian at Virginia Tech while working at CTVR. He then did a short post-doc before joining us. While Tom was at CTVR, with Keith Nolan, also of CTVR did a cross country tour following the SDR Forum meeting in Denver Colorado in Oct. 2007 and did a survey of white space while doing their drive.
This will be of increasing interest in the coming years with a lot of support for this coming from a consortium including Microsoft and Google.
We need spectrum to increase the amount available to wireless devices. The FCC realizes this and in November of 2008, issues rules allowing this. CTVR at Trinity College Dublin has been interested in this concept for a while. A prominent DARPA project manager, Preston Marshall, a leading person in DySpan has long been interested, personally and professionally in White Spaces (a Dynamic Spectrum hot spot issue, pun intended). Preston is getting his Ph. D. from CTVR.
Tom Rondeau, hired by my employer, with my being the primary instigator, finished his Ph.D. thesis for Charles Bostian at Virginia Tech while working at CTVR. He then did a short post-doc before joining us. While Tom was at CTVR, with Keith Nolan, also of CTVR did a cross country tour following the SDR Forum meeting in Denver Colorado in Oct. 2007 and did a survey of white space while doing their drive.
This will be of increasing interest in the coming years with a lot of support for this coming from a consortium including Microsoft and Google.
Friday, February 06, 2009
Coraline
Neil Gaman's Coraline opens in theaters today in 3D if your theater does it. I paid for 50 3D glasses with "lovely" designs all over them for a viewing party for family, friends (strangers, drunks, .....). Thank goodness they only cost $20 (plus shipping and handling of course).
It is a beautifully written book with enough illustrations to allow you to get an idea of Gaman's imagination when he sees his story. I recommend the movie release version of the book because it contains an interview with Gaman, Henry Selick (director) who did Nightmare before Christmas and James and the Giant Peach. This version also contains excerpts from the movie screenplay, illustrating some differences from the book (such as a new character). I love stop action animation done with class. Selick is clearly capable.
Enjoy!
It is a beautifully written book with enough illustrations to allow you to get an idea of Gaman's imagination when he sees his story. I recommend the movie release version of the book because it contains an interview with Gaman, Henry Selick (director) who did Nightmare before Christmas and James and the Giant Peach. This version also contains excerpts from the movie screenplay, illustrating some differences from the book (such as a new character). I love stop action animation done with class. Selick is clearly capable.
Enjoy!
Wednesday, February 04, 2009
MEP
The Microwave Engineering Project, referred to in earlier entries, has pushed ahead with a multi-pronged approach building hardware for the testing of the overall design elements of the communications system to flow from the project.
Beagle Boards and Intel ATOM computers, using ARM/NEON/TMS320C64X and Intel Atom 330 respectively, will vie for control of the user interface and digitization and playback of user signals. The BB is lower power and much more compact using podcast and is here directly.
Michelle Thompson, W5NYV (MEP fearless leader) has 3 each beagle board and ATOM 330 mobos to play with.
Beagle Boards and Intel ATOM computers, using ARM/NEON/TMS320C64X and Intel Atom 330 respectively, will vie for control of the user interface and digitization and playback of user signals. The BB is lower power and much more compact using podcast and is here directly.
Michelle Thompson, W5NYV (MEP fearless leader) has 3 each beagle board and ATOM 330 mobos to play with.
Suitsat 2
Suitsat 2 has moved in fits and spurts and not even close to a straight line towards its goal. During the past few days, we have pushed on and received movement in several areas. I believe working hardware for this can and will be produced and that leadership will exert direct influence on this path. I am NOT in a leadership position but a grunt trying to get some of the pieces done.
I worked a plan with project management tonight (Feb. 4) to provide some serious documentation leadership so we can get the tough interconnection pieces done in a timely fashion and then glue the experiments together.
Suitsat was a big success the first time and it was fun being the first to "get" the picture and identify what it was. It was decided that Suitsat 2 should be undertaken but be more ambitious in its scope as can be seen here and here. The project is based on the use of Microchip's dsPIC33 as a low power SDR capable dsp engine. ARISS is an international program who mission is to promote use of amateur radio on the International Space Station.
The transmissions from this package will be receivable by anyone with some simple equipment and my fingers are crossed for its eventual success.
I worked a plan with project management tonight (Feb. 4) to provide some serious documentation leadership so we can get the tough interconnection pieces done in a timely fashion and then glue the experiments together.
Suitsat was a big success the first time and it was fun being the first to "get" the picture and identify what it was. It was decided that Suitsat 2 should be undertaken but be more ambitious in its scope as can be seen here and here. The project is based on the use of Microchip's dsPIC33 as a low power SDR capable dsp engine. ARISS is an international program who mission is to promote use of amateur radio on the International Space Station.
The transmissions from this package will be receivable by anyone with some simple equipment and my fingers are crossed for its eventual success.
Filter Design Book recommendations
Followers of the work of Frank Brickle and I know that Frank has for years followed and promoted the work of Julius O. Smith. I subscribe to this support. Fortunately, JOS has done a lot of publishing his work online for years. And he is winnowing down into books such as
Introduction to Digital Filters with Audio Applications (at Amazon)
or
Introduction to Digital Filters with Audio Applications (at Barnes and Noble)
as well as his awesome treatment of Mathematics of Discrete Fourier Transforms belongs on the signal processor's shelf.
I have also looked through quite a few other books and have purchased these and wish to recommend them:
Introduction to Digital Signal Processing Filter Design by B.A. Shenoi
and
Digital Filters Basics and Design by Dietrich Schlichthärle
With these books on my shelf, I am able to make very good progress on my longer term goals of rending "fdatool for free as in beer" for Octave and SciPy.
Introduction to Digital Filters with Audio Applications (at Amazon)
or
Introduction to Digital Filters with Audio Applications (at Barnes and Noble)
as well as his awesome treatment of Mathematics of Discrete Fourier Transforms belongs on the signal processor's shelf.
I have also looked through quite a few other books and have purchased these and wish to recommend them:
Introduction to Digital Signal Processing Filter Design by B.A. Shenoi
and
Digital Filters Basics and Design by Dietrich Schlichthärle
With these books on my shelf, I am able to make very good progress on my longer term goals of rending "fdatool for free as in beer" for Octave and SciPy.
The Wrestler
Filmed in a grainy looking format with what feels like a hand held digital camera, the cinematography brilliantly sets up an ambiance of decay, seediness, and failure.
Mickey Rourke enters into a semi-autobiographical role and does not pull a single punch in showing a life of one time success, but a complete failure of meeting responsibilities to those around them, and ultimately, to himself. Rourke could not possibly play this role better. The toughest part of his performance is knowing that he has lived this life already. You can hear in his breathing the respiratory degradation he currently suffers. In his face, not covered with a single bit of makeup, you can see the marks of a life poorly led.
Marisa Tomei plays her role beautifully. As an aging stripper, with a kid to raise, and Rourke's love interest, dealing with facing the facts that she is no longer capable of playing the role she must play as stripper, she takes her multilayer character and shows us all the layers with what seems perfect identification with each.
I do not know if Rourke can win over Penn and the other great roles nominated, but certainly, his role deserves the Academy Award nomination it has received.
Mickey Rourke enters into a semi-autobiographical role and does not pull a single punch in showing a life of one time success, but a complete failure of meeting responsibilities to those around them, and ultimately, to himself. Rourke could not possibly play this role better. The toughest part of his performance is knowing that he has lived this life already. You can hear in his breathing the respiratory degradation he currently suffers. In his face, not covered with a single bit of makeup, you can see the marks of a life poorly led.
Marisa Tomei plays her role beautifully. As an aging stripper, with a kid to raise, and Rourke's love interest, dealing with facing the facts that she is no longer capable of playing the role she must play as stripper, she takes her multilayer character and shows us all the layers with what seems perfect identification with each.
I do not know if Rourke can win over Penn and the other great roles nominated, but certainly, his role deserves the Academy Award nomination it has received.
Tuesday, February 03, 2009
Travel and Weather
I can get just a bit harried and I did not check my tickets carefully this morning. I had specifically asked for train tickets to and from Princeton Jct. on Amtrak. I watched as the train pulled away from Pct. Jct. 15 minutes before I was supposed to board ...... in Trenton.
I will check more carefully from now on.
The skies are opened and white stuff is falling but the temperature is above freezing and nothing adding to the road.
Why is having this lousy weather renewing my interest in a touring motorcycle? Weird.
I am working remotely on the dual cell broad band engine server at LTS. I am also continuing to hack on DSP code for all sorts of things. When I run out of mental fuel for one, I jump to the other. It has always worked and appears to continue to. Go figure!
Some views from today's "near miss" Nor'easter:
I will check more carefully from now on.
The skies are opened and white stuff is falling but the temperature is above freezing and nothing adding to the road.
Why is having this lousy weather renewing my interest in a touring motorcycle? Weird.
I am working remotely on the dual cell broad band engine server at LTS. I am also continuing to hack on DSP code for all sorts of things. When I run out of mental fuel for one, I jump to the other. It has always worked and appears to continue to. Go figure!
Some views from today's "near miss" Nor'easter:
Anyone for a swim?
If it weren't so much trouble and mess, you could like this beauty.
My BigIR vertical and part of my NVIS array under the weight of the snow
Sunday, February 01, 2009
Reading Steinbeck, Working on DSP/SDR, Playing with toys on a warmer day
Just finished Steinbeck's "The Moon is Down". One of the most intimate portraits of occupation by and resistance to an occupying army. May I suggest that this be mandatory reading in all war colleges all over the world?
Continuing today on filter design software and more, both stand alone, and for application to polyphase filter banks. This is sheer tedium but it must be done, correctly, and in GPL software so people like me do not have to waste their bloody time ever looking for it again. I am determined to make fdatool for Octave and scipy.
I am playing with different coffees these days. My most recent "trial" was on Jacu bird coffee from Brazil. It is euphemistically listed as "bird selected". It bears no relation other than that to the horrible and expensive Kopi Luwak which is "civet selected". My favorite continues to be Sumatra Lintong. I have found another green bean company besides SweetMarias.
I also like several Ecuadorian, Panamanian in addition to the expensive Jamaican and Hawaiian coffees. For every day, the latter are just a bit over the top in price. My coffee roaster, an Iroast2, has easily paid for itself many times over. The green beans are considerably cheaper per pound, even in small quantity, than "grocery store" ground swill. My brewer is a Technivorm and is carefully controlled to brew at exactly 93C/200F. Ummm, Mmmm, GOOD!
I recently purchased one of the "Chinese electronic wonders" that cost 10% of what the same American or German made instruments cost 10-15 years ago. I purchased a Vector charging system. It cost $120. The cheapest thing I had seen before it that claimed to do what it can do was nearly $1000. It will charge at 2, 10, 20, or 40 amps. It will also pulse rapidly at high amperage/voltage to recondition badly handled batteries. I have 320 amps hours in series 24 AGM deep discharge batteries. I neglected them horribly for a year. They were all in bad shape. The difference nowadays is they cost well over $200 each. I recovered two of the cells to over 1/2 capacity with two treatments, up to five are useful. One of the cells is beginning to show life after two treatments, so I hold out hope. One is dead. So the unit has easily payed for itself in one set of savings. I paid 1/3 what these batteries cost now.
Winter is letting go a bit. It is 10C/50F. Quite comfy in a sweater, and sunny as well.
Continuing today on filter design software and more, both stand alone, and for application to polyphase filter banks. This is sheer tedium but it must be done, correctly, and in GPL software so people like me do not have to waste their bloody time ever looking for it again. I am determined to make fdatool for Octave and scipy.
I am playing with different coffees these days. My most recent "trial" was on Jacu bird coffee from Brazil. It is euphemistically listed as "bird selected". It bears no relation other than that to the horrible and expensive Kopi Luwak which is "civet selected". My favorite continues to be Sumatra Lintong. I have found another green bean company besides SweetMarias.
I also like several Ecuadorian, Panamanian in addition to the expensive Jamaican and Hawaiian coffees. For every day, the latter are just a bit over the top in price. My coffee roaster, an Iroast2, has easily paid for itself many times over. The green beans are considerably cheaper per pound, even in small quantity, than "grocery store" ground swill. My brewer is a Technivorm and is carefully controlled to brew at exactly 93C/200F. Ummm, Mmmm, GOOD!
I recently purchased one of the "Chinese electronic wonders" that cost 10% of what the same American or German made instruments cost 10-15 years ago. I purchased a Vector charging system. It cost $120. The cheapest thing I had seen before it that claimed to do what it can do was nearly $1000. It will charge at 2, 10, 20, or 40 amps. It will also pulse rapidly at high amperage/voltage to recondition badly handled batteries. I have 320 amps hours in series 24 AGM deep discharge batteries. I neglected them horribly for a year. They were all in bad shape. The difference nowadays is they cost well over $200 each. I recovered two of the cells to over 1/2 capacity with two treatments, up to five are useful. One of the cells is beginning to show life after two treatments, so I hold out hope. One is dead. So the unit has easily payed for itself in one set of savings. I paid 1/3 what these batteries cost now.
Winter is letting go a bit. It is 10C/50F. Quite comfy in a sweater, and sunny as well.
Friday, January 30, 2009
Interesting projects
These days with RFIC's coming along, interesting ways to make stable microwave oscillators, there are a couple of interesting projects for the radio amateur. Both are moving a little slowly but with help from you gentle reader they could move a lot faster:
Microwave Engineering Project adroitly led by Michelle Thompson, W5NYV and uWSDR led by several good folks. Both are worth investing some time with to see if you can help/enjoy this kind of development. Both are peopled by very strong engineers, hardware and software.
Cheers, N4HY
Microwave Engineering Project adroitly led by Michelle Thompson, W5NYV and uWSDR led by several good folks. Both are worth investing some time with to see if you can help/enjoy this kind of development. Both are peopled by very strong engineers, hardware and software.
Cheers, N4HY
Thursday, January 29, 2009
Bayesian Nets, MDL, MCMC, Visual Studio 2008
Today my employer and I agreed to engage Frank Brickle as a consultant for the purposes of working on the application of Bayesian Nets, Minimum Description Length, Monte Carlo for Markov Chains to the problems of cognitive radio. A large bibliography is being prepared for our efforts by several people.
I helped Eric Wachsmann of Flex Radio to port their PowerSDR from MS Visual Studio 2003 .NET to Visual Studio 2008 and to use .NET 3.5.
I helped Eric Wachsmann of Flex Radio to port their PowerSDR from MS Visual Studio 2003 .NET to Visual Studio 2008 and to use .NET 3.5.
Wednesday, January 28, 2009
OMAP processor/ Beagle Board
The Beagle Board is beginning to gain some push from inside TI. That is good news. In addition to the recently release free version of Code Composer Studio for these OMAP parts, we find ARM releasing library support, including NEON code (bottom of page) and TI releasing library support for the TMS320C64x . I am assured that more is soon to follow.
Good news! These processors are more than capable of running DttSP SDR core and that port will begin.
Good news! These processors are more than capable of running DttSP SDR core and that port will begin.
Subscribe to:
Posts (Atom)