In the car, often the road noise will make it to hard to hear the quiet tracks, so I normalize it to avoid having to adjust the volume when it plays, and then re-adjust after it is done for the next track.

You guys seem to think that normalizing somehow magically makes all your music play at the same volume. It doesn't. All it does is measure the single loudest peak in your song, then adds an equal amount of volume to the whole song so that the peak really does hit 100 percent (or 98 percent or whatever the setting is).

The problem is that most CDs you buy are already normalized, so setting the normalization flag in your ripper rarely does anything at all.

You could normalize every album in your collection and they would still sound like they were all played at different volumes. The reason is that every album is produced/mastered differently, and they all have different dynamics (relationships between loud and quiet parts). There are other threads on the BBS which discuss this subject, known as "dynamic compression".

In the meantime, I unconditionally normalize all my albums to 100 percent, just in case I happen to run into a CD that's not already normalized. This has actually happened to me a couple of times. Recently I got my hands on an old CD that had some songs peaking in the 70% range. In that (rare) case, normalization is very useful and helpful.

Doug "tanstaafl" Burnside has been struggling with normalization recently because his collection is a mixture of classical and rock/techno. The classical stuff is extremely dynamic, which also makes it seem to have a relatively quiet percieved volume compared to the rock and techno stuff. One of the ideas he was toying with was to take the rock recordings and normalize them DOWN to a low percentage, while leaving the classical pieces at 100 percent. Doug, what did your final solution end up being?

___________
Tony Fabris

I have sience found AltoMP3 Maker
Since you use Alto MP3, I had a question. On the options, CD-ROM screen is the option to Delete Silence Automatically with 2 input boxes (Leading Silent Frames and End Silent Frames). Are those boxes for how many frames to delete (which could result in chopped songs) or to include when encoding to MP3? I'd assume the latter but just want to be sure.

Thanks.

Squid2k1
---
18GB of Green Mk II Fury - Fast As A Shark

... take the rock recordings and normalize them DOWN to a low percentage, while leaving the classical pieces at 100 percent. Doug, what did your final solution end up being?

I ended up leaving them the way they were, all normalized to 98%. It turned out that I had to take a lot of normalization out of a rock song to get a really noticeable reduction of apparent volume compared to my classical songs, to the point where I was effectively compressing the dynamic range of the song by more than I was comfortable with.

I encoded Rush's "2112" at 100%, 80%, 60%, and 40% normalization and there was amazingly little difference in the apparent listening volume.

The normalization seems more effective at bringing quiet tracks up in volume than in taking loud tracks down.

tanstaafl.

"There Ain't No Such Thing As A Free Lunch"

Has Anyone Looked at the WMA Format and what drawbacks or advantages it may have over MP3???

I try to keep all my music normalized. Most of the non-classical albums I have already seem to be normalized, although I think in some cases I have had to go back and re-rip with normalization.

For classical, normalization isn't enough because of the dynamic range, so I wrote a short program to compress out the extreme dynamics and make sure the entire track is as loud as possible. It works pretty well, although the background noise is much louder during the quiet passages, and some of the more dramatic parts of the music are washed out a bit because the extremes in volume are removed, causing some strange effects sometimes (although you still can easily hear the contrasts between "quiet" from "loud" parts). Purists would be horrified. But I definitely think it's worth it, because classical music is unlistenable in the car otherwise.. Half of it is inaudible and the other half is way too loud.

BTW I use musicmatch 4 to rip and encode at 128kbps.

I wrote a short program to compress out the extreme dynamics and make sure the entire track is as loud as possible.

Richard Lovejoy of this BBS wrote a custom kernel which does this in real-time on the Empeg. See the FAQ section of this BBS, under the heading "how do I put custom logos or kernels into the empeg?" for more information.

Hey, Doug, have you tried Rjlov's kernel yet?

___________
Tony Fabris

Hey, Doug, have you tried Rjlov's kernel yet?

No. I don't think rjlov's kernal will be the way to go for me, because the classical music I play relies so much on dynamic range for its effectiveness. I mean, just imagine what Ravel's "Bolero" would be like played through rjlov's volume-adjusting kernal. [shudder]

I am fortunate enough to have a pretty quiet automobile with a lot of pretty good speakers, and the difference between cabin noise at 30 MPH and, say, 70 MPH, is such that I only have to raise my player volume by two or three dB to compensate.

tanstaafl.

"There Ain't No Such Thing As A Free Lunch"

A quick question on loading Rjlov's voladjust kernel with 1.01: is it just the same as 1.0? It was such a while ago I did it, but the studded snow tires go on soon and a moderately quiet car becomes a moderately loud car.

-Zeke

just say you weren't paying much attention...

A quick question on loading Rjlov's voladjust kernel with 1.01: is it just the same as 1.0?

Since it is a modified kernel, it is vital that you have the same version number as the current player software. If you only have his 1.0 kernel, you must download his 1.01 kernel if you wish to use it with the Empeg 1.01 player software.

There are no "on the surface" differences between 1.0 and 1.01, but he has added some "under the hood" hooks which allow you to link to a chooser program. I have been able to run this, and although it's useful, it's not quite ready for public consumption yet because the chooser doesn't display anything on the screen and it requires some fancy mucking about at the bash prompt to install it. Also, it requires the developer build in order to work. All the chooser program does is allow you to choose between "off", and "on" with two possible amounts of compression at boot time.

___________
Tony Fabris