Implement Lame 3.96.1 afterwards ?

[roger]

Re: Implement Lame 3.96.1 afterwards ?

Check the dates you have as well. Perhaps you never uninstalled the older versions when you instaleld the latest, is that possible? or did you install the latest and then went and downloaded what you thought was latest from that other site and installed that over the one from dB?

Anyway, I don't think you can go wrong with the latest versions from dB.

No, I definitely uninstalled version 10 first, even deleting the orphaned registry keys (I do that with all software uninstallations). I think I'll try that manual update of Lame and if there are problems I'll just uninstall and reinstall.

Thanks for your responses

[LtData]

Re: Implement Lame 3.96.1 afterwards ?

What's the date on the dll? mine's 11/27/2004

[roger]

Re: Implement Lame 3.96.1 afterwards ?

All dates are 12/12/2004 ie. created, modified and accessed in 'Properties'. I don't think any of these dates will ever be earlier than the day you installed the program, will they?

[ChristinaS]

Re: Implement Lame 3.96.1 afterwards ?

All dates are 12/12/2004 ie. created, modified and accessed in 'Properties'. I don't think any of these dates will ever be earlier than the day you installed the program, will they?

Exactly, the dates for the versions you installed have to be for the day you installed them. So if you just installed dB and yet you have a date older than the day you installed the latest version, that is not the right file to be looking at. It means yuo never did install that file together with the latest.

[Spoon]

Re: Implement Lame 3.96.1 afterwards ?

You can replace lame_enc.dll with any version you want, although dMC R11 should have installed the 3.96.1, if the size is 163KB then it is 3.96.1 - the version information comes from:

\dbpoweramp\versions\compression\mp3 (lame).txt

try deleting that file and let the install reinstall.

[adaywayne]

Re: Implement Lame 3.96.1 afterwards ?

Well, there's those of us who can tell the difference betwee, say, mp3 and ogg. Also, the side of my lame_enc.dll is 163KB.

There are many factors and variables that affect one's perception of audio "quality" or "fidelity". One of the more important ones is, obviously, frequency range, and especially the upper frequency cut-off point. As points of reference, the lowest note on a grand piano has a basic frequency of about 30 to 40 Hz (Hz = cycles per second), the highest note is about 4200 Hz. The highest note on a piccolo is not much higher at 4500 Hz.

So why do audio engineers strive to achieve a flat frequency response in the range of 30 to 20,000 Hz? The answer has to do with what are known as harmonics or overtones. Every basic note played on a musical instrument creates "harmonics". Relative to the "fundamental frequency", the harmonics have frequencies which are 2x, 3x, 4x etc, the frequency of the fundamental. It is the prevalence and magnitude of these harmonics which give each musical instrument its own unique sound.

Now, they claim that a healthy, young adult has a hearing range of from 20 to 20,000 Hz (that's assuming, I think, that they have not been exposed to any of today's "music.) I, myself, nearing 70 years of age, have an upper frequency limit at about 12,000 Hz. You can do a simple test of your own hearing at:

http://www.engr.uky.edu/~donohue/audio/fsear.html

I have recently constructed spectrograms (frequency versus signal level pictures) of a music wave file encoded with different encoders at different bitrates. Since I can't post graphics here, I will simply give you the upper frequency cut-off in each case. As I said, frequency range is not the only variable in audio quality, but I think I can safely say that anything cut-off below 8000 to 12,000 Hz (perhaps higher for those with bats' ears) will sound inferior in some way to the original wave file. So here are the numbers: (note, all files are stereo and 44,100 Hz)

1. Original wave: 19,000+
2. Lame V. 3.96.1 at 128 VBR: 17,000+
3. Fraunhoefer mp3 at 128 VBR: 9,000
4. Lame V. 3.96.1 at 68 VBR: 11,000
5. WMA V. 9.1, 2-pass at 64 VBR: 13,000
6. OGG at 64 VBR: 15,000+
7. MP3PRO at 64 VBR played back using mp3PRO decoder: 16,000
8. mp3PRO at 64 VBR played back with Lame 3.96.1 decoder: 8,000

I couldn't fit the numbers in the above table so below are the actual file sizes.

1. 31,339 KB..............Original
2. 2,848 KB..............Lame at 128
3. 2,982 KB..............mp3 at 128
4. 1,499 KB..............Lame at 68
5. 1,447 KB..............WMA at 64
6. 1,380 KB..............OGG at 64
7. 1,423 KB............. mp3PRO at 64 with mp3PRO playback
8. 1,423 KB..............mp3PRO at 64 with mp3 playback

I think the results are self explanatory but here are a few comments (based solely on upper frequency cut-off)

1. Lame 3.96 is far superior to a "standard mp3 encoder I had. No wonder Fraunhofer/Thomson are rushing to commercialise mp3PRO!

2. At 64 Kbps bitrate, OGG and mp3PRO are pretty close to Lame at 128 bitrate and far superior to "standard mp3 at 128.

3. WMA at 64 bitrate is inferior to OGG and mp3PRO at the same bitrate.

4. mp3 PRO played back with a "standard" mp3 decoder (instead of an mp3PRO decoder) is a total disaster. The cut-off at 8,000 Hz is predictable because mp3 encoding splits the file at 8000Hz and encode the lower portion using mp3 and the upper portion with a different encoder. Played back with a standard mp3 decoder, all you hear is the portion up to 8000 Hz.

Hope you find the above interesting. Questions welcomed.

Arnie

PS: I should add that the Fraunhoefer "standard" mp3 codec that I used was a version from at least 2 years ago. More recent versions may provide much better results.