Latest | Mon | Tues | Wed | Thurs | Fri

Slogging Through the Land of Ogg

I finally decided to start saving some of my music CDs to my hard drive so I could listen to selected cuts when I wanted in the order I wanted. This is not rocket science, so I reckoned it would not be that difficult to do.

A weekend later and you can still see the radiant glow from the mushroom cloud over our house. But first, a note about the configuration. My main PC at home is based on an Intel 815EEAL motherboard with a Pentium III 933 and 512MB of RAM. There is a Plextor 40X CDRW and a Pioneer DVD drive. The motherboard has built in sound, video, and NIC. I have Windows 2000 Pro, SP2 as the operating system running on a 40GB IBM Deskstar drive.

It started easily enough by deciding what format to save the files. Since I wanted to stay clear of any copyright issues I decided to go with the Ogg Vorbis (hereinafter ogg) compressed audio format (see their site here). At their site they have links to third-party software that can rip (i.e., read the audio file from the music CD) and then encode to ogg or MP3 (see it here).

My first choice, again because of it was Open Source, was CD-DA X-Tractor (see their site here). As with most recent Windows applications, the install is simple enough. Download the file and run the install. No problem.

But once installed, try as hard as I could, I just couldn't get it work with my Plextor. I kept getting an ASPI error message 0x0004. The help file had no information on this and neither did the the forum at their web site. However, from a quick review of the postings, it is not an uncommon error. Unfortunately, there were no replies to the people asking for help on this error.

So I went back to the ogg site to try something else. The next one to grab my attention was called AudioGrabber (see it here). It is, unfortunately, a commercial product but it has the ability to delete the silences at the beginning and end of a track, something I find particularly grating. So I downloaded the free trial version and again, had no problems installing it.

But as before, I had problems getting it to work with my configuration. I did finally get it to read my Plextor by changing the AudioGrabber Settings/General/CD-ROM access method from "ASPI" to "Analog." From that point onward, I could at least read the drive, if only at 1X. Aaaargh.

So I thought I was home free and began to rip files. But when I tried to play them through WinAmp (see it here), all I got was some low frequency noise, and nothing else. Sigh. I couldn't figure out what the problem was. The software was reading the drive and it was creating these huge files. I could save in MP3, WAV, and ogg with no problems. But none of the files actually had any music in them.

I finally decided to take a look at Windows /Control Panel/Sounds and Media/Audio/Sound Recording. Clicking on the Volume button brought up the slider controls for the various inputs. I noticed that the "CD Audio" input was checked as the default. But I remembered that the Plextor's output came through the "Aux" input, not CD Audio. So I checked that box and immediately noticed that AudioGrabber was now getting a signal. Yippee. Although, for some reason, AudioGrabber seemed to be changing the default back to CD Audio every time it ran. So I went to the input side and muted the CD-Audio slider. I don't know why I needed to do that, and perhaps you won't need to, but I did. So now I can rip and save audio files from music CDs.

Still to be determined is why I am ripping at 1X, and whether ogg can be configured to give good sound, but create smaller files. This last problem is serious because I can record a five-minute song to MP3 and end up with a 5MB file. But that same song, saved in ogg at 128Kbits/s is several orders of magnitude larger (maybe 15MB or larger). While I certainly want to stay within the law, I don't have unlimited drive space. So, for now, I'm encoding to MP3.

I should have quit while I was ahead. At this point, I lost my mind and decided to try to install similar software on to my Linux box. This box has an Intel 815EEAL2 motherboard, 850MHz Celeron, and 256MB of RAM. It is running Red Hat 7.3 with the 2.4.19 kernel. The CD-ROM drive is a no-name Mitsumi Matsushita 36X.

The Ogg site had links to various third-party programs. For no particular reason, I chose something called "Grip" (see it here). It runs under the Gnome desktop and requires libghttp, which they have a link to. Now, those of you who have read these pages before know what's coming next and can skip to the end of the post. For everyone else, be aware that there be dragons here.

You must download, unzip and untar, configure, make, and install libghttp. But. Before you can do that, you have to have other software that it is dependent on. For my configuration, it was about a 100 dependencies. Being that some, or all of those are probably dependent on other programs, I quickly decided that was not going to work. In fact, this was the straw that broke the Linux experiment's back. This box is now running Windows 2000 Pro because I just got fed up with trying to install or update programs, and end up in a never ending battle with dependencies. I will not go there. I want software that works for me, not against me. For those of you thinking that installing Red Hat 8.0 would solve all these problems, think again. From what I can see, Red Hat 8.0 would have even more problems because apparently all support for MP3 was removed. Good luck trying to resolve the dependencies on that one.

Aloha!

Abby Nirmul

After all these years, very little is as easy as it should be in computing. Continuing the saga from yesterday, I got the former Linux box converted over to Windows 2000 Pro. From there I installed AudioGrabber and WinAmp (see the links from yesterday) to start ripping some music.

The software installed fine and AudioGrabber ripped at 14X! Hallelujah. No more staring at the monitor while the software slogged at 1X. So I created an MP3 and had WinAmp play it. Only, it sounded like chipmunks where singing and playing. Sigh.

{rant_mode: high; pulling_of_hair: extreme} Why is this so difficult? Why can't people get together and support standards, both for hardware and software, that make it easy for customers to use their products?

I'll have to do some Googling to find out why the speed difference. What is really strange is that when I transfer the files to my other Windows box the files play just fine. I guess the old saying about what doesn't kill you makes you stronger may apply in this situation. Of course, the saying is silent about going after the people responsible for this...

NetObjects ConFusion

Sometimes, why a decision is made is as important as the decision itself. Let me give you an example, the Judiciary supported web site creation and management software is NetObjects Fusion (see their site here). This product became the standard because it wasn't created by Microsoft. No, that's not a very good reason, although there are advantages to avoiding MS products (potentially less bloat, more security, lower price, conformance to open standards, etc.).

While I am no fanatic about what software to use, as long as it gets the job done, avoiding something based on reasoning bordering on religious zeal seems potentially counterproductive. I think this is the case in this situation.

I've been using NetObjects version 5.0 for several years now. It can create a simple site using predesigned templates that will give a clean, professional look. It will take care of linking pages together and updating menus. And when you are ready, it will upload the site to your host provider. But like FrontPage, the HTML it produces is not standards compliant. Further, the style, if I can use that word in this context, of HTML it produces is concerned more with mimicking WYSIWYG than creating efficient code readable by all.

Version 7, the latest release, isn't much better. In addition, one of the new "features", the page counter, appear to be links to external sites, rather than internal code. One wonders why the few lines of code needed to create a counter couldn't be included? While this is pure speculation on my part, I wonder if there isn't a financial relationship between NetObjects and the producers of such services?

In either case, I've used both FrontPage and NetObjects for sometime and I'm not sure either one is what I would choose to create and maintain a site. YMMV.

Surplus Ideas

What a difference an election makes. During the presidential debates, both candidates talked about what they would do with the projected budgetary surplus. In fact, much of the debate centered around just how many billions there would be and how to spend it.

The Democrats said it was best to pay down the national debt. The Republicans said yes, that is a good idea, but a better idea is to give the money back to the people since it was their's to begin with.

The Republicans won the election and so tax cuts and refunds were put into effect. Unfortunately for most people, the cuts and refunds went to the most wealthy political contributers, not the average Joe on the street.

Don't get me wrong, a case for supporting such actions can be made. We all want to be rich, right? There are incentives to being rich. One of those incentives are the tax breaks available only to the top one or two percent of the population. Nothing wrong with that because if there are disincentives to being rich, no one would spend the time and effort, creating jobs along the way, to become rich.

The problem, as I see it, is not that there are tax advantages for the few, but that there is not the needed balance for the needs of the many.

Fast forward to the present. All of the money is gone. Money that I was forced to give to the federal government, to pay for services, has been redistributed and given to the wealthy. So, what does the Republican administration want to do? Give even more tax dollars to the rich. But this time, they aren't even trying to throw a few crumbs our way. They are sending a clear message that the middle class does not matter. That it is better to fund these tax cuts on the backs of the middle class now, and into the future in the form of deficit spending, than to pay down the national debt.

So, what does the Democratic party do? Do they ask everyone to sacrifice a little now so that the following generations won't have to pay for our profligate spending? Of course not, because we, the voters, won't let them. We want it all now, but we don't want to pay for it. We want lower taxes, smaller government, but more services.

There are great opportunities for the Democratic party, the question is, do we have the will to seize them?

Winding Down

As the year winds down, it is traditional to look back during the year and reflect on the important events that made a difference in our lives. But before you go running for the door, all I want to say to my eleven loyal readers is that I appreciate your spending a few minutes reading my Miscellaneous Ramblings. May you and yours have a Happy and Prosperous New Year!

Aloha!

Happy New Year!

With a Little Help From My Friends

A Big Mahalo to fellow Daynoter Phil Hough (see his site here). Phil helped me get AudioGrabber up and running by giving me the Advanced SCSI Programming Interface (ASPI) files needed to run the software. You can download the file from ZDNET here. While some web sites say Windows 2000, the OS I'm running, does not need a separate ASPI driver, I have not been able to get AudioGrabber to work at anything over 1X without it. Note, while the letters ASPI include the acronym SCSI, that does not mean the driver works only with SCSI drives. My own system here at work is a Dell with IDE and the driver works fine with it.

With the driver loaded, AudioGrabber, running on my Dell OptiPlex GX1 PII-333 with 384MB of RAM and a Toshiba XM-6202B CDROM rips at about 8.5X. Not exactly blindingly fast, but a heck of a lot better than 1X!

As an aside, I've been comparing two encoders, LAME and BladeEnc (get both here from the AudioGrabber site) for output file size. I wish I had a spectrum analyzer to check the fidelity of the output files against the original (singer Nora Jones: "Don't Know Why", 3 min. 5 sec.) but others have done so and seem to find significant falloff above 16kHz when ripping at 128Kbit/s. They indicated ripping at 256 all but corrects the falloff. The results displayed below show files sizes at 128 and 256 Kbit/s. I also tested the LAME variable bit rate (VBR) option. VBR is touted as a good compromise between high fidelity and smaller file size. You can take a look at the results and decide for yourself. Phil indicates he rips at 160Kbit/s and finds this to be a good compromise. Only you and your golden ears can decide what is acceptable to you. YMMV.

Encoder	128Kbit/s	256Kbit/s	VBR-1
LAME	2,852KB	5,708KB	4,059KB
BLADE	2,851KB	5,702KB	N/A
Ogg Vorbis	9,778KB	9,778KB	N/A

Aloha!

It's Friday!

The Beat Goes On

I've updated the table I created yesterday (see above) to include the Ogg Vorbis encoder as I reckon people might be interested in the file size comparison. As noted earlier, Ogg creates a file much larger than MP3. But what is really interesting is that as I increased the sample rate, the file size stayed the same. I really don't know what to make of this as I am not an expert in this field. Even going up to 350Kbit/s still gave me the same file size as sampling at 128! I'm not sure if that is correct or if something is misconfigured. But there you go.

Whichever the case, MP3 still gives a substantially smaller file size versus Ogg. And using MP3 VBR-1 saves 1.7MB over the corresponding MP3 file recorded at 256Kbit/s, while at the same time providing a similar, if not higher fidelity. YMMV.

I've also got things working at home on my PIII-933. Ripping is occurring at between 20 to 25X which means a four minute song can be ripped, normalized, and MP3 encoded in about a minute. Not bad. I think I'm going to like doing this but I'm going to need to get a larger hard drive... <G>

Have a Great Weekend Everyone - Aloha!

© 2002, 2003 Daniel K. Seto. All rights reserved. Disclaimer