Csound retunes karaoke singers on the fly

Hmph. Maybe I haven't been to enough Karaoke bars to appreciate the need
for retuning every non-12 note Karaoke singers produce... check out the
story below, which I've pasted from

http://www.reuters.com/news_article.jhtml?type=technologynews&StoryID=764437

Jon

p.s. I agree with moving the list off Yahoo--I'd rather have as little to
do with commercial websites as possible. Majordomo rules.

-------------<begin forwarded article>-------------
Japan Firm Unveils Tech for Perfect-Pitch Karaoke
April 02, 2002 05:57 AM ET
By Edmund Klamann

TOKYO (Reuters) - Some might call it Japan's biggest victory against noise
pollution since pop duo Pink Lady split up two decades ago.

Karaoke sound systems provider Taito Corp said on Tuesday it had teamed up
with a U.S. professor and chipmaker Analog Devices Inc on technology that
could give even the most tone-deaf crooner perfect pitch.

Using the "Csound" computer music language pioneered years ago by
Massachusetts Institute of Technology professor Barry Vercoe, Taito will
market a system this summer that adjusts sing-along music automatically to
the pitch and tempo best-suited to an individual singer.

"This is sound synthesis on software that until now was only used in
experiments and research," said Hidehito Kitamura, who headed up Taito's
development of the new system.

Eventually, he said, Taito may use the technology to reconfigure a
singer's errant tones to the proper pitch, without otherwise altering the
sound.

"We'll be moving from one new feature to the next."

That could be good news for the millions of regular patrons at Japan's
karaoke bars -- both those who dread the inevitable pressure to sing even
if they can't carry a tune, and those who have to listen to them.

"Karaoke is said to be an original Japanese cultural contribution, but
this could take it to new levels," Keio University professor Toru Iwatake
told a news conference unveiling the technology.

Tuesday's demonstration was limited to automatic tempo adjustment,
particularly suited for ballads, Kitamura said.

SALES PITCH

The system will also let a singer calibrate the key automatically before a
song begins, he added.

Machines now require manual pitch and tempo adjustments, which can be hard
for amateur songsters to gauge and can create distortions that are
difficult to sing along with.

For those confident of their vocal prowess, moreover, the new system will
be able to objectively assess pitch, rhythm and skill at such voice
techniques as vibrato and crescendo -- an indispensable item for the
occasional karaoke competition.

Karaoke rating systems already exist, but they can only determine how
closely the singer's voice matches the recorded original, not singing
ability.

"This assesses singing skill mathematically," Kitamura said.

MIT's Vercoe, who lauded Taito for finding a way to bring the sound
synthesis technology to market, said pitch correction with Csound had been
demonstrated long ago. It could even convert a spoken voice into melody,
in real time.

"It's in the technology. You just have to switch it on," he said.

Taito's Kitamura added that the advent of high-speed communications had
also been essential for the new system.

Ten years ago, Taito pioneered Japan's popular "telecom karaoke" services,
which pipe karaoke music and lyrics to pubs and night clubs via wireline
networks at far less cost than maintaining an on-site disc collection.

Noting the new system would require about a hundred times or more much
bandwidth than current systems, Kitamura added: "We can do this thanks to
ADSL (assymetric digital subscriber lines) and fiber-to-the-home."