Home » Daynotes Home » Week of 17 January 2005

Photograph of Robert Bruce Thompson Daynotes Journal

Week of 17 January 2005

Latest Update: Friday, 21 January 2005 10:36 -0500

Click Here to Subscribe Buy PC Hardware in a Nutshell, 3rd Edition:
Buy Building the Perfect PC:
[Amazon] [Barnes & Noble] [Bookpool]
[Amazon] [Barnes & Noble] [Bookpool]
Visit Barbara's Journal Page

Monday, 17 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]

10:13 - The first draft of Part I of our new article for WindowsDevCenter, Building the Perfect PC: Quality on a Budget, is posted on the subscribers page. It's not polished yet, and there may be changes after we actually build the systems, but it's in reasonably final form now. If you spot any mistakes or have any suggestions, please post them on the messageboard.

One of Barbara's belated Saturnalia gifts finally arrived, a Creative Labs MuVo Micro N200 flash-based MP3 player. Like one of our Border Collies who once caught a UPS truck by the front tire, the question becomes, "Now that I have it, what do I do with it?"

Barbara wanted a tiny, sub-$100 MP3 player that she could take to the gym. I looked for a unit that supported OGG Vorbis format, but the only ones I could find that did that were way out of our price range. So I settled for the MuVo, which like most units supports MP3 and WMA. The first question is, which format should I encode Barbara's CDs to?

I have no idea whether the MuVo supports variable bitrate MP3s, so if I settle on MP3 format I'd do some test files at different data rates using constant bitrate and variable bitrate encoding to determine which formats the MuVo supports. Obviously, after I rip all her CDs I'll have to encode them, and I'd prefer to encode them only once. I'm not concerned about minimizing file sizes for the MuVo; I want quality good enough to use for home audio as well. With constant bit-rate MP3, that probably means encoding at 320 kb/s.

I installed grip this weekend and ripped and encoded a dozen or so of Barbara's CDs to 320 kb/s MP3s, just to have something to play with. At 320 kb/s, a full CD generally occupies from 75 MB to 175 MB, which isn't unreasonable. The good news is that when I connect a USB cable to Barbara's MuVo it appears as a USB mass storage device. I can simply copy MP3 files to it as I could to any other drive, and delete them when she wants to substitute other files. The MuVo has a USB 2.0 interface, which is a blessing. I'd hate to fill it with USB 1.1. It'd take quite a while.

But before I rip and encode hundreds of CDs, I want to make sure I've made the right choice for encoding. I think I understand that WMA has better audio quality than MP3 across the board. WMA even supports lossless compression, which would be ideal. I'm not concerned about the proprietary WMA format, because these files will be only rips of the original CDs, which will remain available. I am concerned with playback, because of course we're standardized on Xandros Linux around here. If I'm going to encode to WMA, I want to make sure that there's a full-function player for Linux that supports WMA format(s). Ideally, I'd like to do the encoding on a Linux box, but if there's no WMA encoder available for Linux (or if the Windows-based WMA encoders are superior in features or audio quality), I'd be willing to do the encoding on a Windows box.

Which brings up the question of how to organize all these audio files. Barbara has hundreds of audio CDs. Ideally, what I'd like is a Linux application that would allow me just to feed in a stack of audio CDs. It would rip each CD, get the metadata from CDDB, encode each track to MP3 or WMA using whatever parameters I specified, and add each CD and track to a database. When Barbara wanted to reload her MuVo, she could simply fire up the database application and sort by CD title, track name, artist, genre, or whatever to build a custom playlist. She could then mark checkboxes for those tracks she wanted to copy to the MuVo, connect the MuVo's USB 2.0 cable, and copy away.

Any suggestions or links would be appreciated.

I did encounter one oddity while I was playing around with the MP3s. I tested the MP3s by playing them on my primary desktop system, which has a Logitech Z3 speaker system. That's an inexpensive 2.1 set, $55 or so from NewEgg. The audio quality was pretty decent, although the bass is boomy. Still, I was impressed with how well the Z3 speakers reproduced the music.

Then I went back and played the same tracks on Barbara's system, which has a $150 M-Audio StudioPro 4 2.0 speaker set. M-Audio calls these speakers monitors, and they aren't kidding. I expected the M-Audio speakers to sound better, but the difference was stunning. The bass was tight, with no boominess. The midranges and highs were crystalline, with superb separation. The M-Audio StudioPro 4 monitors are in an entirely different class, as they should be for their price.

The oddity was that when I returned to my office and again played the tracks on the Logitech Z3 speakers, they sounded as good as ever, which is to say pretty decent. I expected to be sitting there thinking, "these suck". In fact, I was sitting there thinking, "these sound pretty damned good."


Tuesday, 18 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]

09:57 - O'Reilly sent me a PDF proof of the cover for PC Hardware Buyer's Guide: Choosing the Perfect Components yesterday. I posted it on the subscribers' page.

Thanks to everyone who's responded to my questions about ripping and encoding Barbara's CD collection. There are a lot of useful comments over on the messageboard. And there's this.

-------- Original Message --------
Subject: Boomy speakers
Date: Mon, 17 Jan 2005 18:06:18 -0600
From: Edgerton, Paul A
To: Robert Bruce Thompson

In a prior life I was a recording engineer.  I think I can explain why your Logitech speakers seem boomy while the M Audio speakers in the other room do not.  The short answer is that bass frequencies have very long wavelengths. They are much more affected by the physical dimensions of the room and coupling between the speaker and nearby surfaces. Those factors have much less effect on higher frequencies.

You mentioned that your Logitech speakers are a 2.1 system.  I'm guessing that you have the subwoofer near a wall, on the floor or both.  Many speakers are designed to radiate into a spherical space, but most subwoofers were intended to rest on the floor, radiating into a hemispherical space. If you put them too close to a wall, this looks like a quarter-sphere at bass frequencies, so the bass would seem louder.  If you stuck the subwoofer in a corner, you are radiating into only 1/8 of sphere.  The walls of the room act as a kind of a horn.  (Remember Klipschorns?)

The implication of this is that you can fine-tune the amount of bass by moving the subwoofer closer to the walls or further out into the room. There are a couple of gotchas, though.  First, your little subwoofer probably isn't putting out much real bass; I would guess most of it's output is around 100 Hz.  This is near the top of the range of "real bass," and if you put the subwoofer too far away from the satellite speakers, the sound image sort of becomes "unglued" in the sense that the subwoofer appears to be a distinct sound source rather than a reinforcement of the satellite speakers.

Another problem occurs when you have a small room with parallel walls and not much low-frequency absorption.  All rooms have pronounced resonances (or eigentones) that reinforce some parts of the frequency range but not others. The smaller the room, the fewer and further apart these resonances are. Unless the dimensions of the room are selected to avoid this problem, several different parts of the room may resonate at the same frequencies, especially in rooms whose dimensions are integral multiples.  This buildup of resonances can be exacerbated by putting the subwoofer under your desk.

Barbara's speakers are better to begin with, and I suspect that they are in a more favorable acoustical environment.  But as I suggested before, you can experiment with subwoofer placement to make some improvement in your other set up.  Don't expect too much, though.  Those little subwoofers make up for their lack of real, chest-pounding bass by putting out more sound in the boomy range from 100 - 200 Hz.  If you tame the boominess, you may kill what little bass they produce.  As in everything, TANSTAAFL.

I see that I 'm seriously past due on my membership renewal, which I'll take care of right away.  Thanks for all you do.

-- Paul Edgerton

Thanks. You're right about the Z3s. The subwoofer sits on my desk (which is actually a solid 3'0" door) backed right up against the wall. I'd have it on the floor, but the cables that connect it to the satellites are very short and have proprietary connectors. I don't expect much real bass from a $55 set of speakers with an 8" woofer. I suspect if they provided a real frequency response curve to even -3 dB that it'd run out around 100 Hz.

I do appreciate real bass, though. Back when I was in college, I dated a girl who was going to Oberlin. They were re-doing the common area in her dorm, which had built-in speakers. They'd ripped a bunch of stuff out, including the speakers, and were going to discard it. I salvaged two huge speakers. They were 24" and had about 60-pound magnets. Those suckers could move some air. I built them into cabinets and drove them with a Phase Linear amplifier, which IIRC provided 350 W per channel. When I cranked that thing up, it literally vibrated books off shelves. I ended up selling the amp and speakers when I needed money, but I wish I'd held onto it. I don't think Phase Linear is even around any more.


Wednesday, 19 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]

10:27 - It's been cold the last few days, with lows from 23°F down to 10°F (-5°C to -12°C). I know the folks in Embarrass, Minnesota, where it got down to -56°F the other day, would consider that a heat wave, but around here it's no joke. Many homes aren't insulated for temperatures much below -10°C or -15°C, nor are most furnaces large enough to cope with very low temperatures. Fortunately, when we had our furnace replaced a few years ago, I specifically told them to super-size it. Between that and our 40,000 BTU/hour natural gas logs, we'd probably be able to stay comfortable down to about -40°, which isn't going to happen around here.

Yesterday, for the first time, I was cold in my office. I actually put on an LL Bean chamois shirt. During the winter, I always close the heating vent in my office to keep warm air away from the computers. I usually have enough systems running to heat the office all by themselves. So I was surprised when I started to feel cold. Then I realized that there was only one computer running in my office, my main desktop system. At times, I've had as many as a dozen systems running in here, which is probably the equivalent of one of those 1,500W quartz heaters.

I talked to O'Reilly yesterday about sending out review copies of PC Hardware Buyer's Guide: Choosing the Perfect Components. They're going to do the same thing they did last time, offer copies to 50 of my subscribers with the understanding that those who receive review copies will post reviews on Amazon.com and anywhere else they like.

When we did this for Building the Perfect PC, I just announced it to my subscribers and took the first 50 who responded. A few people got their requests in late and were disappointed, so this time I'm going to do things a bit differently. I'll ask any subscriber who wants a review copy to email me. I'll collect those emails from now until next Friday at noon. If more than 50 people respond, I'll hold a random drawing to choose the 50 winners.

For BPP, O'Reilly sent out a review copy and an O'Reilly t-shirt. This time, alas, they're short of t-shirts, so it'll be only the book.

I got email from Roland Dobbins, saying that 320 kb/s was an insanely high bit rate for encoding MP3s. Thinking about it, I believe he's right. At 320 kb/s, only about 107 minutes of music fits on Barbara's 256 MB MuVo N200 flash MP3 player. At the 160 kb/s encoding Roland recommends for most music, she could fit twice that. But 160 kb/s isn't adequate for critical listening on home audio equipment or good headphones.

Thinking about it, I realized that just one 160 GB hard drive can store about 250 audio CDs in .WAV format. That means I could store Barbara's entire CD collection on a couple hard drives without using any compression at all. So I think what I'm going to do is rip her CDs, encode them to 160 kb/s MP3s (or perhaps VBR MP3s at that average bit rate), and keep the ripped .WAV tracks. That gives me the best of both worlds at the small cost of some disk space. Later on, if she ends up with a player that supports OGG Vorbis, I can simply re-encode the .WAV files without having to re-rip everything.

14:59 - Another one bites the dust. I just installed Xandros 3 on my old Compaq Armada E500 notebook. Xandros detected all the hardware except the D-Link 802.11g+ Cardbus wireless adapter. That wasn't unexpected. Linux generally doesn't have great support for 802.11g yet. Some people have reported getting similar cards running with NDISwrapper, but I haven't tried that.

Interestingly, the Armada "feels" faster under Xandros than under Windows 2000. I've noticed before that Xandros feels the same or slower than Windows 2000 on newer, faster equipment, but on older, slower equipment Xandros feels faster than W2K. I'm not sure why that is, or even if it is. These are just subjective impressions. I haven't run any benchmarks to confirm relative performance.

It's not just a question of boot time. Windows "cheats" by displaying the desktop before it has completely loaded. Depending on how fast the system is, Windows may display the desktop for 30 seconds or more while services are still loading in the background. Linux loads everything and then displays the desktop, so at times it seems to be slower loading. What I'm talking about is relative performance once the OS and apps are completely loaded. For example, OOo scrolls a large document faster under Xandros than under W2K.

This system has only a 750 MHz Pentium 3 and 256 MB of RAM, so possibly it's a question of Windows wanting more hardware resources and Xandros being happy with what it has.

I used the Xandros installer's option to resize the Windows partition, so right now the Armada dual-boots Windows 2000 and Xandros. I think I'll leave it that way for a while so I can continue to use Cartes du Ciel under Windows if I want to. The Kstars planetarium app for Linux is pretty, but has only a small fraction of the functionality of CdC. We could probably get by with just Kstars, though, because it does what we really want it to do, which is help us locate deep-sky objects. It allows searching by common name, NGC/IC number, and catalog name. So, for example, if we're looking for the Andromeda Galaxy, we can locate it with Kstars by searching for "Andromeda Galaxy", "M 31" (for Messier 31), NGC 224, UGC 454, and so on.

Also, the 1.0 version installed from Xandros Networks with Xandros 3 appears to be better than the 0.9.1 version available on Xandros Networks for Xandros 2. I'll have to look at the newer version and see what they've added and improved. But I could probably re-install Xandros 3.0 right now and just tell it to take over the whole disk. I don't really need Windows on the notebook, and I could probably find an old 802.11b WiFi card that Xandros 3.0 would recognized. Perhaps I'll keep it dual-booting for a while, though, to verify that power management works properly.


Thursday, 20 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]

08:20 - I'm off to the dentist to have my fangs cleaned.

Things are likely to be sparse around here for the rest of the month. I have a couple systems to build, an article to write, chapters to work on for the new edition of PC Hardware in a Nutshell, and a bunch of other stuff to get done. Also, Xandros beta testing is starting again. I can't give any details, but if you keep an eye on the Xandros web site I think you'll be surprised soon.

10:56 - Back from the dentist. I survived again.

My apologies for the truncated page this morning. N|vu has some annoying bugs, and one of the most annoying is that when it publishes a page it sometimes claims it has published successfully when in fact the transfer did not complete.

The UPS tracking site tells me I should receive two Antec SLK2650BQE cases today. Those are the last missing components for the budget systems. I plan to build the AMD budget system this weekend.


Friday, 21 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]

09:28 - Magistrate Judge Brooke Wells handed SCO a big win yesterday. Despite Groklaw's attempt to downplay it, this is a real win for SCO. It allows them to go on the fishing expedition they've wanted all along. More important, it allows SCO to delay things yet again. Ms. Wells showed her complete lack of understanding of the case with the following remark:

"Finally, the court wishes to note the importance of the code to both parties. SCO has much to gain by showing that any so called homegrown code allegedly within the purview of the contract ended up in Linux. In equal respect, IBM's case will be strengthened tenfold if IBM can show that notwithstanding possible contract protections, homegrown code provided no basis for the code that IBM eventually contributed to Linux. . . ."

"Homegrown code" refers to code that IBM wrote independently and later incorporated in AIX. SCO's bizarre theory (now, not originally) is that "homegrown" IBM code, once it was incorporated in AIX, came under SCO's control. For example, SCO argues that if IBM wrote a journaling file system that they later incorporated in AIX, IBM was not entitled to donate that "homegrown" JFS code to Linux because it had already been incorporated in AIX. Nothing in copyright or case law supports that argument, and yet Wells appears to have swallowed it hook, line, and sinker, despite the fact that IBM has clearly pointed out that SCO's attempt to unilaterally extend copyright law has no basis.

SCO gains nothing by showing that IBM donated "homegrown" code to Linux. IBM does not and never has denied that it contributed "homegrown" IBM code to Linux. IBM argues, correctly, that their "homegrown" code is theirs to do with as they please. Wells has completely missed the point, and by doing so has subjected IBM to the costs and delays that will be required to provide this needless discovery to SCO.

Looking at things as a layman who is reasonably familiar with the facts of the case, it seems to me that Ms. Wells should be dismissed and sanctioned for incompetence. With so-called judges like Wells on the case, I expect this one to go all the way to the Supreme Court. If, that is, SCO doesn't run out of money first. Now that SCO's lead law firm has agreed to fee caps, it'll be interesting to see just how hard they pursue the matter.

A pair of Antec SLK2650BQE cases arrived yesterday, so we'll be building at least one system this weekend, assuming the forecast ice storms don't kill our power.

10:36 - Thanks to everyone who's subscribed or renewed recently. If you're not a subscriber and you'd like to support this place, click here.

This from Svenson.

-------- Original Message --------
Subject: daynote
Date: Thu, 20 Jan 2005 23:26:05 -0800
From: Jan Swijsen
To: Robert Bruce Thompson "

[quote]10:56 - Back from the dentist. I survived again. [/quote]

You did but the dentist ... ?  [G]

Kind regards,
Sjon Svenson

Heh. I went unarmed.

Years ago, one of my friends who routinely carries a Colt Combat Commander pistol went to a new dentist for a filling. He declined to remove his jacket, but when they reclined the chair his jacket fell back and they saw his pistol.

He's a very large, intimidating-looking guy, anyway, but apparently the pistol really worried them. He said they were very solicitous--"Are you really, really sure you're numbed up enough?" He thought it was funny. Nowadays, they'd probably call Homeland Security.


Saturday, 22 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]


Sunday, 23 January 2005

[Last Week] [Monday] [Tuesday] [Wednesday] [Thursday] [Friday] [Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com Forums]


Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005 by Robert Bruce Thompson. All Rights Reserved.