Last FPCR distinctives sermon

Status
Not open for further replies.

NaphtaliPress

Administrator
Staff member
The last in the sermon series on FPCR distinctive practices was posted last week. It is on Worship Patterns and covers some various topics including the no instruments in public worship distinctive. I have started posting higher quality MP3s of FPCR sermons on the main page of fpcr.org. Since these tend to be rather large they will not be archived but only the most recent available there. The lower quality files are still archived by date and in the search engine (though that is only current through 2004). This sermon will be up for a few weeks and the temporary direct link is below. Or see the lower quality file at the link below that.
Worship Patterns sermon (35.5 mb)
See OOW and smaller archived files here.
 
Originally posted by NaphtaliPress
The last in the sermon series on FPCR distinctive practices was posted last week. It is on Worship Patterns and covers some various topics including the no instruments in public worship distinctive. I have started posting higher quality MP3s of FPCR sermons on the main page of fpcr.org. Since these tend to be rather large they will not be archived but only the most recent available there. The lower quality files are still archived by date and in the search engine (though that is only current through 2004). This sermon will be up for a few weeks and the temporary direct link is below. Or see the lower quality file at the link below that.
Worship Patterns sermon (35.5 mb)
See OOW and smaller archived files here.

Chris,

What are your settings at? 35MB seems kind of high. I was getting good quality at about 10-12MB per hour. You can hear the quality on my website.
 
Fred,
Do you have a link for your website? The sermon was 75 minutes, 44100 frequency, and 64 KB. I noticed some unpleasant artifacts on lower settings; maybe I could go lower than this but this sounds pretty good.
 
Fred,
I note you used 32 KB and 22050 frequency. It has some slight slurring. I archive everything at a lower rate than even that, 11025 at 16KB, but wanted something really clear for temporary availability on or 'front page'. I tried dropping it some more but the slurring and artifacts start to creep in pretty quickly. Hence my choice. I'll play with it some more this week.
 
Originally posted by NaphtaliPress
Fred,
I note you used 32 KB and 22050 frequency. It has some slight slurring. I archive everything at a lower rate than even that, 11025 at 16KB, but wanted something really clear for temporary availability on or 'front page'. I tried dropping it some more but the slurring and artifacts start to creep in pretty quickly. Hence my choice. I'll play with it some more this week.

Yes, but generally the quality has been pretty good for me. The 22050 is mono. I'll take a listen to yours to see how it sounds in comparison.

I agree that quality can drop off quickly. When I tried 24kb or 16kb, it was extremely "hollow" and "tinny." I usually find most sermonaudio files very bad (some are 6kbs)
 
True. We've sacrificed and gone with tinny and hollow simply because of the size of the file archive, now over 3600 audio files. I'm way behind on keeping the search engine up to date but maybe now that CPJ 2 is out I can take a few days to fix that oversight.
 
You really only need 44kHz sampling rate for music. 22kHz is fine for voice as most of the speech spectrum is below 11kHz.
 
That's interesting. There is a little musical intro and exit that gets hit very slightly on quality at 22KH, but I note that the file size is exactly the same when compressed at 44K. I used the same 64 Kbps each time.
 
File size with mp3s depends on bit rate and length of track. The track is divided up into small time segments. A time segment that contains more information can "borrow" against future time segments that don't require as much information to describe. Music usually requires more information than speech or silence, so the encoder would use some of the bits allocated to these time segments for the music segments. When you set the sampling rate to 44kHz, the music section will be forced to borrow a larger share of bits from the rest of the track. Depending on the relative length of the music and the speech, it may not amount to anything noticeable.
 
I realized that my previous post wasn't terribly helpful. The time segment part isn't really important. What is important is target bitrate and sampling rate. When you set the bitrate, you are giving the encoder a target to match. It will do as much or as little compression as it has to to match the target. Not all audio tracks are going to sound as nice at a given bitrate. One reason speech can sound better at a lower bitrate is it has a narrower frequency spectrum than music. When you downsample the track you want to encode to 22kHz, you are effectively doing some preprocessing to make the encoder's job a bit easier. Before downsampling can occur, the track goes through a low-pass filter that removes all the high-frequency components. At a 22kHz sampling rate it would remove everything above 11kHz. In speech this frequency range isn't terribly important (in fact telephones remove everything above 4kHz), so it can be safely discarded. So now when you encode the file, the encoder doesn't even have to take the mostly unused high-frequency components into consideration because they are no longer there. Granted, most encoders are going to throw out that part anyway because they will detect the low energy of the track at those frequencies.

Of course, as you noticed, the music part suffers from the low-pass filter.

One thing you may want to do is try different encoders. Many encoders are optimized for music at relatively high bitrates and may not be equally strong for speech at lower bitrates.
 
I was thinking, shouldn't you have a non-instrumental EP intro for this sermon anyway?
 
Thanks very much for the info btw. I may look at other encoders but if there is not size difference in the file then I may stick with what I'm doing.
 
Originally posted by NaphtaliPress
Thanks very much for the info btw. I may look at other encoders but if there is not size difference in the file then I may stick with what I'm doing.

Chris,

What are you using for an encoder or encoding program?
 
Originally posted by NaphtaliPress
Fred,
dbPowerAMP. Nominal cost (might have been free?). You have any opinion on it?

I believe it is free unless you need the "Power Pack." This is what I use as well - it works better than anything else I have seen, except Adobe Audition, which is more of a sound editing program than decoder program.

But I guess the more apt question is what Codec are you using? Lame? Fraunhofen? Blade?
 
Originally posted by SRoper
It might be. LAME, if it is not licensed with Fraunhofer, violates US patent law.

From:
http://lame.sourceforge.net/tech-FAQ.txt

6. Does LAME use any MP3 patented technology?

LAME, as the name says, is *not* an encoder. LAME is a development
project which uses the open source model to improve MP3 technology.
Many people believe that compiling this code and distributing an
encoder which uses this code would violate some patents (in the US,
Europe and Japan). However, *only* a patent lawyer is qualified to
make this determination. The LAME project tries to avoid all these
legal issues by only releasing source code, much like the ISO
distributes MP3 "demonstration" source code. Source code is
considered as speech, which may contain descriptions of patented
technology. Descriptions of patents are in the public
domain.

Several companies plan on releasing encoders based on LAME, and
they intend to obtain all the appropriate patent licenses. At least
one company is now shipping a fully licensed version of LAME with
their portable MP3 player.

Note that under German Patent Law, §11(1) a patent doesn't cover
private acts with non-industrial purposes. Probably interesting for
developers is that a patent doesn't cover acts with experimental
purposes, that aim at the object of the patented invention (§11(2)).
 
Yeah, if it came with that program. As I understand it, you can't download the source from the LAME site and compile it.
 
Status
Not open for further replies.
Back
Top