Joe,
sometimes there is no alternative to a human being making a decision.
Lining up music in a Rivendell [ or other playout library ] can be
computer controlled; but not perfectly; in either music or acoustic terms.
Lets deal with your first question:
There are vendors who claim that their library has all tracks
'normalised' and all cue points set. Which data is available in a form
suiting the system it was first curated on, and may be translated into
Rivendell. They are not cheap, and you have no real way of determining
the provenance of each track unless you have a CD original. There are
some perfectly horrible acoustic levelling algorithms out there. Best
avoided.
There are a couple of vendors who supply broadcast wav libraries who
previously supplied CD libraries to stations. Some are 'original' audio
and some have been 'processed'. They don't come with cues but may have
intro and tail data. I have recently seen posts of comparative audio
screenshots contrasting the original [CD or Vinyl] and the Library
version to illustrate how the dynamic range has been tweaked. This may
be acceptable for some.
Your Not For Profit wont have the funds to buy a commercial library. I
have a quote for a commercial station I am constructing and a 'basic'
library will cost $US10,000 [$2 a track x 5000 cuts] Probably a fair
price but beyond all but a commercial endeavour.
Now to the second part:
You can import songs into Rivendell in bulk from CD or other formats,
and you can set the segue points.
By default Rivendell 'normalises' each cut so that the electrical level
of each track is the same.
This is not a perfect process, its within 3db, and in no small part is
determined by the sound card you have in your input machine.
Start and finish cues are set, and again depend on the recording as they
are set electrically. You would be surprised how much studio crud there
is between tracks on some CD's
# man 1 rdimport
is a good read
You can force a segue when the level falls to X db or % at the end of
the track. [I do not recommend this]
There is no argument that the apparent 'loudness' of a cut is determined
by the audio content as well as the electrical peak. Frequency range,
combination of instruments, use of compression, double tracking, and
various other tricks can make some songs sound a lot louder in
comparison to others.
I have a bunch of 78's with varying loudness across the lot. This is a
not a new problem.
In pre automation days someone sat at the on air desk, listened to what
went out and set levels using a meter to monitor the electrical signal,
and their ears to determine loudness. Quiet passages got cranked up,
loud passages were mixed down a bit. [I was there]
Various devices were created that could split the audio spectrum into
two, three or more bands which could be individually levelled and
tweaked so all output from a station sounds more or less the same which
in the extreme case eliminates any nuance of dynamic range in the
original recording.
[Audimax & Volumax, Harris Tri-Bander, Prism, et al] You may have one at
the end of your on air chain.
They are only as good as what they are presented with.
Faced with the problem of trying to maintain a consistent level on your
station from a system that will be automated sometimes, and live others,
setting the level and cue points of each individual cut is unavoidable,
and part of music library management.
In "Edit Markers" you can set the cut gain so that each song pays out at
more or less the same listening level, and you can set the cue points so
that [for example] the segue occurs 'on the beat' of the beginning of
the fade out. Sometimes the music demands a tight segue, sometimes you
need to allow that ending to decay before playing the next song. You
need to consider Tempo. This is a MUSICAL decision that it's currently
not possible for a computer to make. Setting the segue at so many db
below the average level of the cut is possible; but it takes no account
of what has been playing.
Yes it's a lot of work.
I would be using some smart delegation to get it done.
Have your Jocks and Jockettes come in with their music, which will be in
every format possible, and set up a workstation where they can ingest
the audio. You may have to do the "copy files to a temp directory from
USB and use rdimport to bulk import them into the MUSIC or whatever
group you use to hold tracks before they are graded" bit but they can
rip CD's easy enough.
If you copy files from their device into a temp folder, anything that
won't copy is stuff you don't want. They may have a whole lot of folders
with album cover pics and such. Mount the device and use 'find' to dig
out the mp3's or MP3's [ find /path/to/location -name '*.*3' -e cp {}
/path/to/temp/folder \;]
Be ready to have to clean out the mp3 tags to get stuff to import. Thats
why I like working on copies. Do rdimport on the mp3's with
--delete-source and if any fail do "eyeD3 --remove--all" on the left
over files and try again.
You can use filenames to input metadata. Be prepared that the metadata
*will* need editing.
Rivendell can handle most formats, stuff it rejects you generally don't
want to air. If it's that important to a programme become familiar with
SOX. It can help fix a lot of problems.
If audio came from mp3 or other compressed format I identify that in the
'User Defined' section of the entry for each cart. Later if I get a CD
or direct wav from Vinyl dub I can replace the compressed version.
Then you show them how to go through and set the markers and the playout
level, using the same monitors at the same level setting for everything.
They will also have to check the Artist Title and other data because
even stuff from an on line database will have errors.
Not every Jock person will be competent and confident, but in any group
you will find you have several who can be easily taught how to ingest
music, and set it up. Once they know how they will enthusiastically get
music in the system for a while then as they tire you can line up others.
Somewhere out there is the geeky kid who wants to be in radio who will
come in after school and do a perfect job. Train and Trust.
It's important that you have a database backup somewhere, and a copy of
the existing library somewhere. I have 4 copies of /var/snd in 3
locations updated daily. You need at least one to keep things Jock-Proof.
Joe, the music library will be a never ending labour, and there is no
easy way. There has been discussion here and elsewhere about processing
of wav files and setting playback levels based on calculated loudness. I
think this is a 'no undo' process but I have not bothered with it
because I am definitely 'Old School'. Rivendell lets you set up the
playout levels and it does take time.
I have a way of listening to the library in segments while I am doing
other stuff and I pick up faults in tracks. Mostly momentary glitches in
audio which may have happened anywhere along the chain.
All of my playout systems have a limiter compressor at the end of the
chain to enhance average level and to protect the transmitter against
random peaks. I have used JAMMIN with some success. The compressor is
set soft and the limiter hard. Every track is graded so it hits the desk
at about the right level. I actually get comments about our audio
because it has dynamic range close to that on the original recording.
It is possible to process all the audio and set standard segue points so
you have rock solid modulation and nogapsbetweensongs but somehow I
don't think you would like to listen to it.
regards
Robert Jeffares
On 28/02/17 04:43, Joe Thompson wrote:
Finely got this little Non-Profit LPM station to purchase a new server
and I got it up and running at least on the rivendell side of
things per a previous post.
Now my question I got is related to the audio library. Is there
anyplace in particular that one would recommend that they source their
music library from that would already have the cue, segue points etc
coded in. This way they don't have to do it for every song
manually they ingest.
The second part of my question is: is there a way to configure
rivendell so that it will automatically set those points as its
ingesting audio. Knowing that there may need to be a few tweaks done
here and there because not all songs are equal?
I'm asking because currently they are considering having each jock
bring in his or her own library and input it into rivendell. I can see
that becoming a really huge mess and time consuming process. They
don't really have a set format which I and others who have been in the
business for a number of years have frowned upon.
_______________________________________________
Rivendell-dev mailing list
[email protected]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
--
*Communication Consultants*
64 Warner Park Avenue
Laingholm
Auckland 0604
09 8176358
0221693124
06 650 6087
_______________________________________________
Rivendell-dev mailing list
[email protected]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
