where to order cialis
cialis cod
prescription cialis
next day cialis
cheap cialis
where to order lipitor
lipitor cod
prescription lipitor
next day lipitor
cheap lipitor

The last update on my music computer system was just over two years ago; it is time for a new update.  The main music computer is still the same AMD Athlon in the TNN-500 case but the hard drives have been upgraded.  The sound I/O card is the Lynx Two.  The OS is Windows XP SP2 updated with the most recent patches.  The programs used are basically the same as reported earlier:

Exact Audio Copy - for ripping CDs
Sony Sound Forge 7 - for recording from records and for general sound editing
FLAC - for lossless compression
Ogg Drop - for lossy compression
Tag&Rename - for editing meta fields
Foobar2000 - for playing back files
Windows XP/Windows Explorer - for organizing and accessing sound files

As the number of sound files has grown, I’ve ended up using Windows Explorer and the Windows file system to organizing the files.  I use the hierarchical file system as follows:

• Top level: Type of file (Books, Music, Lectures, etc.)
• Music Category
• Album Title
• Resolution (for high-res files only)

The following screen shot shows the hierarchy in action:

I’ve drilled down to the individual tracks (each one a FLAC file) in the 96-24 folder under Scientist_…, under Reggae~dub, under Complete Albums.

Note that I have the Windows appearance set to the old Windows 2000 style.  I don’t like the default candy-colored, toy-style that comes with Windows XP.

Custom folder icons are used to differentiate different types of music: ripped CDs, music downloads, or LPs.  Folders with no custom icon mean that the meta-data editing is not yet complete.  For high-resolution recordings, I keep two sets of data: one for the original high-res recording, typically 96KHz/24bits, and a 44.1K/16bit version for use in my iPod. (I’ve discovered that my 5th-gen video iPod does a lousy sample-rate conversion if the file it plays is not already 44.1KHz.) Remember that I am using ROCKbox on my iPod, which plays the open-source formats.

As the music category folders get large, I plan to add another level to the hierarchy: composer for the classical categories, and artist for the others.

As some of you know, I now have a second house in Vietnam.  My main computer there is a Mac Mini that can dual boot into the Mac OS or Windows XP using “bootcamp”.    A Western Digital “Passport” drive holds all the music files, currently over 100GB.  When booted into Windows, I use Foobar2000 to play music, just like on my music computer.  However, I needed an equivalent player when booted up in the Mac OS.  After trying several, I’ve decided on “Cog”.  It is straightforward, simple, and like Foobar2000, it is free.

Notice that I am not using a database-type music player, such as iTunes.  This is because all the ones I’ve looked at either don’t support the formats I’m using or make it hard to sort by composer (the classical music problem).  Using Windows Explorer is clunky, but usable for now.  I’m keeping an eye on new software, but at the moment, my main task is to move my music to the computer and enjoy listening to it!

I recently stumbled across a source of DRM-free music that I hope represents the future of music distribution. Pristine Classical sells music, mostly classical, but some jazz and blues also, in a variety of forms, ranging from MP3 downloads to physical CDs. The prices are not very cheap, especially since they are in Euros, but no protection or DRM is attached, so you can back them up or trade them with your friends forever. (A reminder of the futility of DRM is in this Ars Technica article about how Microsoft will turn of their MSN Music license key servers at the end of the year. Your licensed (not purchased) music will still run on the computer it is licensed to until something changes - i.e. OS upgrade, hardware change, etc. I’ve never had a Windows computer be this stable for more than a year.)

Pristine Classical was started in 2005 by Andrew Rose as a offshoot of his Pristine Audio business of restoring and remastering of old records and tapes. He has apparently made deals with the copyright holders of some of these classic works so that all recordings are fully legitimate. The age of the recordings go from about 1955 back to the early electrical recordings of the late 1920s. As such, you won’t hear stereo spectaculars, but the recording quality of the late 1940s and early 1950s recordings can be quite good, if only mono. An added plus is that the details of each restoration are explained for each recording. A particularly difficult restoration is explained in the notes for this 1950 Decca recording.

An outfit that sold only MP3 downloads would frankly not interest me, since I consider the sound degradation too significant, even for vintage recordings. However, Pristine Audio recently has started offering FLAC lossless encoding on many of their recordings, in a 44.1K/16-bit format. Not only that, but they have just started offering 48K/24-bit FLAC versions that reflect the resolution of their original restoration. These are nearly twice the cost of the ordinary FLAC versions, but would offer the ultimate quality. My own experiments with resolution and sample rate show that higher bit resolution is a bigger improvement than higher sample rate.

I bought several of the classical FLAC recordings while I was on the road (I still don’t have high-speed internet at home), and when I got home played them on my main system. So-called restoration of old records can sometimes take the life out of the recordings, but Pristine’s restorations were tastefully done. No major clicks or pops and the surface noise is low, but the details are there. On the whole, I liked the restorations.

The Pristine Classical site is rather cluttered and busy, but that is partly because they offer many versions and services, including cue lists and color CD insert printouts for each recording. The following screen capture shows the general pricing structure:

With our poor dollar-to-euro exchange rate, the prices seem a bit high, but considering you are getting the full musical and sonic equivalent of a CD, with the cue lists and cover art, and no DRM, it isn’t so bad. I’m impressed with their offerings, and hope they do well.

Lynn Olson recently pointed me to an article in Rolling Stone magazine that digs deeply into the problem of MP-3-quality sound and poor dynamic range in recordings. This six-page article goes deeper than the articles I mentioned in two earlier posts (What Happened to Dynamic Range? and More on Dynamic Range). Also covered are the tweaks that are done in the recording or mastering studio, such as correcting the pitch of a singer’s voice. I won’t go into the excruciating details, but recommend this article, especially since it has a nice on-line bibliography and a selection of “Good” tracks (such as Bob Dylan’s Modern Times) and “Bad” tracks (such as the Red Hot Chili Peppers Californication) so you can actually hear the differences.

With speakers getting more accurate, and high-resolution audio getting better and cheaper, the fact that recordings are being trashed while mastered and then get squeezed through MP-3 lossy compression can only be seen as a dramatic disconnect. We can only hope that a new generation comes along that appreciates and demands good sound.

Apple recently announced a new music player, the iTouch, which is similar to the iPhone, except without the mobile phone feature. At the same time, it “refreshed” the 5G iPod video player (that I described below), calling it the “iPod Classic”, with storage up to 160GB. The external appearance is quite similar, but internally is quite different. Reports have been cropping up on poor sound quality, glitchy software, and incompatibilities with various programs.

Via reddit comes this analysis by Marc Heijligers at his HiFiVoice site. After noting that the new iPod (called the 6G, for 6th generation) sounded tinny and fatiguing. He then proceeded to run tests, comparing it with the 5G iPod, and basically found the following (quoting Heijligers):

• An uplift in treble (about 0.15dB at 15KHz).
• A strong modulation with 22k, causing a peak in the frequency extreme, and intermodulation distortion.
• A group delay that depends on frequency (i.e. non minimum-phase).

The culprit seems to be a change from the Wolfson D/A converter used in the 5G to a Cirrus converter. Heijligers hints that some corrections might be made by a firmware update, but knowing these kinds of converters, I think the problem is inherent in the Cirrus chip. His site has some interesting links related to the problem, including some to Apple Discussion Groups. It’s interesting to see what I would assume to be non-audiophile listeners (at least to the level often discussed here) complaining about “narrow-ness”, “compressed sound”, “confined … and not as warm”, “one dimensional & not musical at all”, and in summary: “… a less harmonic sound, less spatial (the scene is inside your head instead of around your head), a bit too much emphasis on high treble, and ‘less silence between the notes’. It sounds more technical, and less acoustical/realistic.” And these are all comparisons with earlier iPods!

From the very good MacInTouch site is a user summarizing the operational problems:

• VERY Slow menu switching response
• Display of clock rather than song info when ?Now Playing?
• Inability to use existing AUTHORIZED 3rd party dock products (including Apple-advertised)
• Audio skipping during operation
• Slow connection to Macs and PCs
• Inability to disable ?split-screen? menus
• Lagging and unresponsive Click Wheel
• Camera connector not working
• Inability to use EQ settings without skipping and distortion

Via the fascinating Boing-Boing site comes word that the software incompatibilities are due to a bit of encryption added that only allows Apple’s iTunes to access the iPod’s database. This locks out any 3rd-party software and really leaves the Linux folks out in the cold, since there is no iTunes for Linux. Someone may find a way to crack this, but with the DCMA law around, this is legally risky in the U.S. As Cory Doctorow, who wrote about this, said: “I guess my next player won’t be an iPod after all.”

Update: The folks at ipodminusitunes (I wonder if they got the idea of this name from my post below?) have cracked the encryption and offered a solution to the Linux users. This still doesn’t help the sound problems, though.

Its been a while since I reported on my music on computers project. I completed my sister’s classical music iPod project in April - she really likes it. It was done using Apple’s music formats: AAC and ALAC. These work well but are proprietary. I didn’t want to commit a lifetime of collected music to formats that might be shut down or at a minimum aren’t accesible by sound editing tools. After exploring Rockbox, though, I realized that high-quality open-source formats could be used on the iPod. Why the iPod, and not some other “MP3 player”? Well, to put it bluntly, MP3 sucks, both in sonic artifacts and compression size. There are some players that handle other formats as well as the lossless formats like FLAC, but their build quality is not as good as the iPod. So, I got an 80GB iPod and loaded it up with Rockbox.

Rockbox is an open-source software project to replace the software on portable music players (Apple’s as well as many others) with many features, high-quality audio control, and the ability to play many different sound formats. It is in a constant state of flux - new compiles are happening all the time, so minor glitches are to be expected, and indeed, I’ve had a few strange crashes. The user interface has a few rough edges, and the “out-of-the-box” screen appearance is awful. There are many “themes” available, however, and these can greatly clean-up the appearance. The main thing is that it does play my two formats of choice: FLAC (lossless) and ogg vorbis (lossy) very well, and its built-in database function organizes recordings quite nicely by their tag information.

read more »

At the suggestion of reader Bill, I downloaded the J. River Media Center software package and tried it out on my “music computer”. I’m still in the middle of my iPod classical music project, so I didn’t want to convert everything over to this new system, but I thought I’d “kick the tires” and play around with it a bit. Media Center has a huge set of features, so while I tested the music organizer and playback features, I didn’t even look at the TV features, CD labeler, portable player synchronizing, Download manager, among other features. This is why I am calling this a mini review.

The first impression of both the J.River web site and the Media Center software itself is one of slickness and attention to detail. There are extensive help files and FAQs on the web site, and a multitude of different modes for displaying information within the program. The appearance is attractive without too much eye candy. The default screen is similar in content and appearance to the iTunes screen. Upon opening Media Center for the first time, it asked if it could bring in media files on the computer. I agreed, and it scoured the hard drives for all media, in this case mostly music files. Basically what Media Center does is set up a database that points to wherever the files are - thus not duplicating the files the way iTunes does. Metadata (artist, pictures, etc.) are stored in the database, but can be put into the tag fields in the music files. The massive file import showed a problem, though - my FLAC files were ignored. A little study of the help files showed that only the “Monkey’s Audio” lossless compression format was supported. Given the extensive set of features and the fact that FLAC is open source makes this omission odd.

Files can be sorted and played from Media Center just like with iTunes. Unlike iTunes, Media Center preserves sample rate and bit-depth, so my 96KHz recordings played back just fine at 96KHz. (iTunes seems to downsample everything to 44.1KHz). There is lots of flexibility in creating playlists, so more sophisticated groupings of songs can easily be made.

I tried out Media Editor, a separate program that is installed at the same time as Media Center. It provides basic waveform editing with the ability to change volume, add or delete segments, etc. According to the help file, “Media Editor can open and save any of the many, many formats that are supported by Media Center “. This is somewhat misleading, since all formats are converted to .wav files while edited, then reconverted to the original format when saved, which, of course degrades files that use lossy compression. You are warned about this, though. The maneuvering and editing features are a bit clunky compared to my standard sound editor, Sony Sound Forge. There is no “interpolation” command, which is essential for taking out clicks and pops from LP recordings. So Media Editor has some use, but doesn’t replace a full-function sound editor.

On the whole, Media Center is a solid, well-designed program that does most of what I want. A fly in the ointment, though, is its product activation. It is not quite as demanding as Microsoft’s or Adobe’s activation schemes, and it allows up to ten reactivations a year, but as your system ages and eventually gets replaced, it is essential that the product activation be available forever. How many small software companies last forever? For a program that may end up holding a lifetime worth of music, this is a problem. It is also only available on the Windows platform. Since this is what I currently mainly run, this isn’t too bad for me, but I have been flirting with Mac OS X, and may want to run in a UNIX environment someday. I think I’ll hold off adopting the J. River Media Center for now, and keep looking for another, hopefully open-source program.

Its been a while since I reported on my music computer project. No definitive decisions have been made on it, but I have been learning more about iTunes and thinking a lot about music organization. For a somewhat delayed Christmas present for my sister, I bought an Apple iPod (the latest 30Gb model) and have started filling it with classical music from my collection. She already has an iPod with jazz and wanted some classical music, so she is already set-up and familiar with the iPod. I am running iTunes 7 on my “music computer” WinXP system. During the course of this project, my thoughts on the music computer have been clarified.

One of the first things I decided was that all music, wherever possible, would be stored using lossless compression. This significantly cuts the storage capacity, compared to Apple’s compressed storage algorithm, but on a 30Gbyte drive this would still allow about 110 hours of music. Since my sister primarily plays her iPod at home through her stereo, the preservation of sound quality should be noticeable. iTunes’ default setting for importing music is to compress it to AAC, Apple’s compression standard. The default can be changed, but the selection is quite buried (edit > preferences > advanced > importing). I set it to “Apple Lossless Encoder”. Thus all CDs are now ripped automatically to the lossless mode. For the cases where I imported a .wav file, right-clicking on the tune in iTunes gives a option to encode to lossless. There is no option for importing or playing FLAC files, so in the case where I had only FLAC files (e.g. from purchases from the Philadelphia Orchestra website), I had to manually decode to .wav files, then import them into iTunes. The Apple Lossless Encoder compresses about as well as FLAC - a bit better than half size for classical music and a bit worse than half for pop music.

I made some interesting discoveries on what the iPod would handle. All my music I have been transcribing from LPs to my computer have been sampled at 24 bits/96KHz. I could import these into iTunes, but when I tried to synchronize the iPod, it complained that it couldn’t handle the format. I then downsampled the file to 48KHz, and the iPod was happy. It seemed to handle the 24 bit mode, but I don’t know if it is actually playing all 24 bits. At some point I will record some test tones onto the iPod, put it on the test bench and see how well it performs. But the good news is that it handles 48KHz just fine. Update: Itunes plays back all files on my system at 44.1KHz, despite storing the files at the full data rate! Sample rate conversion is fraught with problems and is hard to do just right. As noted in my comment below, this invalidates iTunes as a high-resolution media player. I don’t know what the iPod does.
read more »

From a recent posting at Boing-Boing, I found out about a public-domain repository of music: Musopen. They accept all kinds of music, but most on their site is classical. The files I checked were MP3 encoded.  The only constraint is that it be in the public domain (i.e. not covered by copyright). Since in the USA only pieces copyrighted before 1923 are totally in the clear, either the piece must be placed in the public domain by the owner or, for works before 1964, had never been copyrighted or the 28 year renewal was never done. Interestingly, the Musopen site doesn’t mention the latter occurrence, but there are plenty of books and movies that never renewed their copyrights, so undoubtedly there are recordings made before 1964 that are in the public domain. The problem is that it takes checking on the Library of Congress database, or for the really old stuff, literally looking through the copyright card catalog in Washington D.C. I’ve done that several times for technical books and magazines.

In any case, its good that something like this exists. I hope enough people support this and contribute to make it worthwhile.

To kick off the new Research Reports part of ClariSonus, here is the result of some measurements I did earlier this year. The intent of this study was to evaluate PC sound cards for use in a PC-based audio test system. The sound cards are getting pretty good, and it is no longer necessary to get expensive test equipment to do good audio measurements. As mentioned in an earlier post, what is still needed is a good “front-end” to these cards, i.e. something that amplifies, attenuates, and protects the inputs of the card. Any comments, corrections, and questions on the report should be posted as comments to this blog posting.

No - this has nothing to do with graffiti - it is about the techniques used to attach meta-data to music files. This metadata typically stores: title, artist, album, year, and genre. More advanced tagging can include date of recording, album cover art and even lyrics. This short posting will cover the implementations of tagging, not the use of it by editors, players, and file systems, nor will it cover the gathering of tag data from databases.

Initially all music files were simply binary files composed of the music samples. To make them more usable, formats such as AIFF (Macintosh) and WAVE (Windows) have imbedding information specifying things like sample rate, and bit depth. There are methods for imbedding MIDI code for synthesizers. There are extensions for adding more metadata, but especially in the case of .wav files, the standard is confused, so many programs act as if tagging can’t be done to .wav files.

The need for a more standardized and extensible method of adding metadata arose when many people started moving record and CD collections to computer files. The desire to intelligently catalog the music led to the need for better tagging. Several standards have grown up around the various music file formats and still exist today. Several are propriety, such as Microsoft’s WMA container-based system and Apple’s AAC files (which, by the way, have no publicly-published description). In the open-software area, the main tagging formats are the ID3-type systems, developed for MP3 files, and the Ogg container-based system used by Ogg-Vorbis and Ogg-FLAC.

read more »

One of the decisions I made early in the design process of my computer-based music machine was to not use lossy compression schemes. This partly due to my disappointment in the quality of MP3 and other schemes and a purist’s desire to keep the music original and intact in any way possible. Ten years ago this would have been near-impossible, due to the limitations of disk storage. However, with the ability to buy a terabyte of fast disk storage for only a few hundred dollars, quite a lot of uncompressed music can be kept online. By adding lossless compression, the amount of music stored in the same space can be essentially doubled.

Some people have been confused by lossless compression - doesn’t compression imply a loss of data? Well, no. Think about your Zip files on your computer. They are compressed, but the original data is retrieved perfectly. Winzip, Stuffit, gzip, etc., are all lossless compression schemes that reduce the file size by eliminating redundancy in the data. These programs can be used to compress music files, but they are slow and don’t take advantage of the particular types of redundancy in digitized music. Good quality music-oriented lossless compression programs can decompress in real time and reduce the music file size by about a factor of two.

Some of the early lossless music compression programs were proprietary ones for commercial products. Meridian Lossless Packing (MLP) is an available option in high-resolution DVD-Audio recordings. Apple Computers uses Apple Lossless Audio Compression (ALAC) as an option when ripping tracks for its iPod. Microsoft Windows has its Windows Media Audio 9 Lossless scheme. The problem with these schemes for my music computer project is that they are proprietary and either are closed or require high licensing fees. As an example, Minnetonka Software makes several programs to generate and burn DVD-Audio disks on a PC. Their high-end DiskWelder Chrome product ($2995) allows the importing of MLP-coded files. To get the MLP coder costs another $1995, for a total of $4990!

read more »

I’ve been researching the software and hardware needed to implement my own high-resolution computer music server. There are many pieces needed to solve this puzzle, and they are constantly evolving. The requirements are basically:

1. Runs under Windows (at first) or, eventually under Linux/Unix
2. Can handle 24-bit/96KHz high-res files as well as CD red-book (16 bit/44.1KHz)
3. Good track and album organization
4. Good user interface

I’ve already got good hardware - what I call my “music computer” - an AMD Athlon CPU with a Lynx 2 sound card installed in a Zalman TNN-500 totally-quiet case. I have been “ripping” LPs into .wav files using Sony’s Sound Forge 7 software at 24-bit/96KHz. However, as I started accumulating music files and I started to consider putting my CDs onto the computer, I realized I needed a good way to index and play the music files.

read more »