The UK Home Automation Archive

Archive Home
Group Home
Search Archive


Advanced Search

The UKHA-ARCHIVE IS CEASING OPERATIONS 31 DEC 2024


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: DIYHA Digital Jukebox - Issues


  • To: <ukha_d@xxxxxxx>
  • Subject: RE: DIYHA Digital Jukebox - Issues
  • From: "Keith Doxey" <lists.diyha@xxxxxxx>
  • Date: Sun, 19 May 2002 14:17:43 +0100
  • Mailing-list: list ukha_d@xxxxxxx; contact ukha_d-owner@xxxxxxx
  • Reply-to: ukha_d@xxxxxxx

Hi Ian,

Agreed, trying to tackle the whole 58000 tracks would be an impossible task
and I wouldnt expect anyone to try. The problem with such a huge repository
is that you could NEVER listen to everything in there and I supspect that
the majority of people also wouldnt want to listen to a lot of the stuff in
there. Everyone has different tastes which is why each persons collections
started out differing so much.

My jukebox can handle as many different versions of a song as I want but it
only references one of the versions against an album. Here is a list of
songs containing "Horse" in the title.

Ride A Wild Horse  - Dee CLARK
They Shoot Horses Dont They - RACING CARS
White Horse (90's Funk Mix) - LAID BACK
White Horse (Original Extended Mix) - LAID BACK
White Horse (White Bitch Remix) - LAID BACK

The last three being different vesions of the same song, again this was a
requirement that I had from my DJ days where I often had both 7" and
12"
versions of the same song.

Using the example you gave I would have 3 entries....

1. Queen - Somebody to Love
2. Queen - Somebody to Love (Live Version)
3. Queen and George Michael - Somebody to Love

"A Day At The Races" and "Greatest Hits" would both
point at #1 which would
be the digitally remastered version as it sounded the best. That also means
that when you buy one of these compliation CD's where "to obtain the
best
possible quality some tracks have been recorded using one or more members
of
the original group" (rough translation - sounds nothing like it
should) you
can reference another copy of that track that you have in your collection
that is taken from the original recording.

As I have said before, I have NEVER been an Albums person, apart from
Compilation and Greatest Hits albums 99% of my purchases have been singles.
I choose music because I like the song, not because it happens to be by a
particular artist. I get bored very quickly listening to the same artist
for
more that 2 or 3 tracks.

My Jukebox was developed because nothing out there seemed to cater for what
is basically a very large Singles collection. Entering the data is quite
slow but a whole album can be entered in about 4 or 5 minutes even if none
of the Artists or Songs exist within the database. As I progress through
I find that more and more the Artist is already there. Once you select the
artist you are presented with a list of existing songs by them. If it is a
duplicate you click on the song and that reference is stored. If its a new
song you enter the details, if its a different version you click "New
Version" and then enter the data that is different eg the duration or
12"
mix, Live, Dub etc.

Being a Singles collector means that the ratio of Artists/Songs is totally
different to what most people probably have in their collections. For
example, the extract I have from the Archive has approx

35000 tracks
3000 albums (average 11.66 tracks)
1300 artists (average 2.3 albums, 26.92 tracks)

My collection only partially entered but fairly representative has

1691 tracks
86 albums (average 19.66 tracks)

but this consists of
1305 songs
714 artists (average 1.82 songs per artist, 0.12 albums per artist!)

The fact that I have 50% of the number of artists you have despite only
having about 4% of the number of tracks probably indicates why none of the
software out there worked for me.

As I see it the main problem with the archive stems from the way it was
created. All CD's were blindly ripped by people with fast machines and they
didnt worry about duplicate entries. Tags were either manually entered or
grabbed from CDDB which gave inconsistant format and filenaming
conventions.
People then joined several large collections together because they could.
The main problems now are the sheer size of the archive and the inability
to
easily find what you want easily. Coupled with the different playback
methods eg Winamp, Rio, DDAR, Audiotron etc each of which has its own
particular way of deciding what to play and in what order.

My Jukebox demo was to show what is possible. If it can work for others
then
so much the better, if not, then it still does what I want.

Obtaining data is easy.... there is data everywhere.
Obtaining MEANINGFUL data that enables powerful applications means either
painstaking data entry or lots of Data Cleansing, both of which take time.
If done progressively it isnt too bad but to tackle a huge volume of data
is
daunting to say the least.

Keith

www.diyha.co.uk
www.kat5.tv


-----Original Message-----
From: Ian Lowe [mailto:ian@xxxxxxx]
Sent: 19 May 2002 09:09
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] DIYHA Digital Jukebox - Issues


I have some comments:

1) Naming
In a great many cases, the naming info provided by the record companies is
*wrong* from album to album. Perfect example: Faith Hill's first album
"Faith Hill" has a song titled "This Kiss", which I
rather like. I have
several copies from various compilations and the album itself. The song is
incorrectly named as "The Kiss" on her own album, or, at least,
the one
published in the UK. it's fine in the states.

Correcting naming errors is a *mind numbing* exercise. I have manually
corrected over THREE THOUSAND filenames and tags in the main archive. The
thought of doing all of them or manually inputting that data in some way
makes my head spin.

I mean, be honest Keith, how long in hours has it taken you to input what
you had? and thats for a thousand and a half plus tracks? The archive I
have
here is sitting at about 58,000 tracks. To manually enter *one word* into a
database for the current archive would take hundreds of man-days.

2) Duplicate removal.

How do you decide if two tracks are "the same" ? as an example:

I have four copies of "Queen - Somebody to Love", all at the same
bitrate
etc and of similar sizes.
*BUT* One is the original, from "a day at the races", One is from
Greatest
Hits and is digitally remastered, and sounds *much* better than the
original
which was a fairly bad CD transfer. The Third one is a live version, and
the
fourth is an absolutely fantastic version with George Michael singing as a
tribute to Freddie Mercury!

The only way to figure that out is to listen to them all, and again, that
doesn't scale too well!

Not to mention, can the database actually handle multiple versions of the
one song?

Ian.

-----Original Message-----
From: Keith Doxey [mailto:lists.diyha@xxxxxxx]
Sent: 19 May 2002 00:24
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] DIYHA Digital Jukebox




For more information: http://www.automatedhome.co.uk
Post message: ukha_d@xxxxxxx
Subscribe:  ukha_d-subscribe@xxxxxxx
Unsubscribe:  ukha_d-unsubscribe@xxxxxxx
List owner:  ukha_d-owner@xxxxxxx

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/




Home | Main Index | Thread Index

Comments to the Webmaster are always welcomed, please use this contact form . Note that as this site is a mailing list archive, the Webmaster has no control over the contents of the messages. Comments about message content should be directed to the relevant mailing list.