Commercial MP3 Encoder

Archived from groups: rec.audio.pro (More info?)

mcp6453 wrote:
> Is there such a thing as a commercial, multipass MP3 encoder?

Sure.


> in those instances when we have no option
> but to encode to MP3, is there anything better than LAME 3.96

At lower bitrates, many users report that the Fraunhofer MP3 codec
sounds better than LAME does.

Preprocessing can hugely improve the quality of encoding. Hardware
companies like Audioactive put a lot of effort into this.

Archived from groups: rec.audio.pro (More info?)

Kurt Albershardt wrote:
> mcp6453 wrote:
>
>> Is there such a thing as a commercial, multipass MP3 encoder?
>
>
> Sure.

Are you aware of any that you can list by name?

>> in those instances when we have no option but to encode to MP3, is
>> there anything better than LAME 3.96
>
>
> At lower bitrates, many users report that the Fraunhofer MP3 codec
> sounds better than LAME does.
>
> Preprocessing can hugely improve the quality of encoding. Hardware
> companies like Audioactive put a lot of effort into this.

What kind of preprocessing? Rolling off everything below 80 Hz, 100 Hz,
120 Hz. Rolling off everything above 8 KHz, 10 KHz, 12KHz? Certainly
having consistent levels would be beneficial. Should I start for
22.5KHz, 44.1KHz, 96KHz wave files? My present plan is to use 44.1KHz
mono wave files as the source files. Any suggestions for preprocessing
that would hugely improve the MP3 result would be appreciated.

Archived from groups: rec.audio.pro (More info?)

mcp6453 wrote:
> Kurt Albershardt wrote:
>
>> mcp6453 wrote:
>>
>>> Is there such a thing as a commercial, multipass MP3 encoder?
>>
>> Sure.
>
> Are you aware of any that you can list by name?

Already did (Fraunhofer.)

What is your target bitrate? That makes a difference.




>> Preprocessing can hugely improve the quality of encoding. Hardware
>> companies like Audioactive put a lot of effort into this.
>
>
> What kind of preprocessing? Rolling off everything below 80 Hz, 100 Hz,
> 120 Hz. Rolling off everything above 8 KHz, 10 KHz, 12KHz?

Yes. Depends on the bitrate and also the type of content. There's also
dynamics to consider, and masking effects. These differ somewhat among
the various encoders.



> My present plan is to use 44.1KHz mono wave files as the source files.

That's a good idea, since starting with oddball samplerates often
creates MP3s that will have trouble on some playback systems.



> Any suggestions for preprocessing
> that would hugely improve the MP3 result would be appreciated.

It's a trial-and-error processs, or at least it has been for me. If you
have the coin, check out <http://www.omniaaudio.com/> (Audioactive)



As Arny pointed out, Hydrogen Audio is a good place for more info.

Archived from groups: rec.audio.pro (More info?)

My target bitrate (not mine, the client's) is 40KHz mono for voice only.
It turns out that a program I own has Fraunhofer built in. So my next
step will be to get a six pack and start doing some listening tests.
Audience applause seems to be very telling with regard to conversion
quality. Also, for what it's worth, it seems that not maintaining at
least some headroom (3 dB below DFS, per Bob Katz?) the digital noise
goes crazy.

Thanks for pointing me in the right direction. I've picked up a lot of
helpful information.



Kurt Albershardt wrote:
> mcp6453 wrote:
>
>> Kurt Albershardt wrote:
>>
>>> mcp6453 wrote:
>>>
>>>> Is there such a thing as a commercial, multipass MP3 encoder?
>>>
>>>
>>> Sure.
>>
>>
>> Are you aware of any that you can list by name?
>
>
> Already did (Fraunhofer.)
>
> What is your target bitrate? That makes a difference.
>
>
>
>
>>> Preprocessing can hugely improve the quality of encoding. Hardware
>>> companies like Audioactive put a lot of effort into this.
>>
>>
>>
>> What kind of preprocessing? Rolling off everything below 80 Hz, 100
>> Hz, 120 Hz. Rolling off everything above 8 KHz, 10 KHz, 12KHz?
>
>
> Yes. Depends on the bitrate and also the type of content. There's also
> dynamics to consider, and masking effects. These differ somewhat among
> the various encoders.
>
>
>
>> My present plan is to use 44.1KHz mono wave files as the source files.
>
>
> That's a good idea, since starting with oddball samplerates often
> creates MP3s that will have trouble on some playback systems.
>
>
>
>> Any suggestions for preprocessing that would hugely improve the MP3
>> result would be appreciated.
>
>
> It's a trial-and-error processs, or at least it has been for me. If you
> have the coin, check out <http://www.omniaaudio.com/> (Audioactive)
>
>
>
> As Arny pointed out, Hydrogen Audio is a good place for more info.

Archived from groups: rec.audio.pro (More info?)

mcp6453 wrote:
> My target bitrate (not mine, the client's) is 40KHz mono for voice only.

Ouch. That's getting down into the range where MP3 artifacts are rather
noticeable.



> It turns out that a program I own has Fraunhofer built in.

I've had pretty good luck with the Fraunhofer encoder (from the Nero
MP3pro plugin) doing voice at 48 kbits using 16/44.1k mono files with a
parametric LPF around 6-8k.

Archived from groups: rec.audio.pro (More info?)

I asked:
<< Will using the standard iTunes encoding routines give me the best results? >>

And Kurt replied:
<< Probably, but they won't be MP3s. The AAC encoding that iTunes uses is
much better at low bitrates than MP3 is. >>

iTunes lets you encode in mp3, as well (accessible through the "importing"
pane in iTunes preferences).

I realize AAC may be a higher-quality format, but I need something completely
"generic" and accessible to anyone...

- John

Archived from groups: rec.audio.pro (More info?)

Lorin David Schultz wrote:
>
> As for "joint stereo," *DON'T* use it. The encoder will try to reduce
> file size by finding stuff that's common to both channels. That seems
> like a great idea, but in practice it usually results in weird, phasey
> artifacts and wandering of the stereo image.

Joint Stereo is pretty much essential if you really want to save
bandwidth. For this particular application (mono speech) it would be
irrelevant.

Archived from groups: rec.audio.pro (More info?)

In article <1122718649.157505.209610@o13g2000cwo.googlegroups.com> moby@kcbbs.gen.nz writes:

> Had a museum application which I was migrating from minidisk to MP3 on
> CF cards which had embedded 18kHz tones to trigger lighting effects and
> it stone cold refused to work with VBR. FBR was fine.

18 kHz tones to control lighting? What ever happened to 18 Hz?

--
I'm really Mike Rivers (mrivers@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo

Archived from groups: rec.audio.pro (More info?)

moby@kcbbs.gen.nz wrote:

> One thing I have found is that VBR, even at 320kBPS, screws HF
> response.
> Had a museum application which I was migrating from minidisk to MP3 on
> CF cards which had embedded 18kHz tones to trigger lighting effects and
> it stone cold refused to work with VBR. FBR was fine.

If you use Lame VBR you can control the upper corner of the
lowpass traded against more or less bits in the result.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein