Topic by Jake Johnson

  • Avatar
  • Jake Johnson
  • Member
  • Offline
  • From: Atlanta
  • Registered: 04-05-2007
  • Posts: 4,491

Topic: Five lectures on the Acoustics of the Piano

Many people here have doubtless seen or studied this well-known site\set of lectures before, but I ran across it again recently, led to it by following a reference by Philippe, and thought I should share it:

http://www.speech.kth.se/music/5_lectures/contents.html

The first lecture is by itself wonderful--The pages include photographs of Cristofori's piano and a replacement soundboard, and a series of pictures of an experiment in which a sand-like substance was poured on a soundboard (of a unstrung piano) that was then vibrated at its various modes so that the "sand" reveals the vibrating areas. And all of the lectures offer much more information than I can claim to fully understand.

Last edited by Jake Johnson (30-12-2008 01:51)

Reply by mwinthrop

  • Avatar
  • mwinthrop
  • Member
  • Offline
  • Registered: 04-07-2007
  • Posts: 61

Re: Five lectures on the Acoustics of the Piano

Thank you for posting this site link.  I did not know this site existed.  I am working my way through the lectures and I am learning alot I did not know!

On this forum, I have notices many people are concerned with having very low latency with Pianoteq.  I have looked at a diagram in the lecture series (http://www.speech.kth.se/music/5_lectur...asure.html) and saw there was about 30ms measured between the time a person touches a key and a sound comes out of the piano that was tested in the article.  If I read the graph correctly, I was wondering if someone might help me understand how a desired low latency on an electronic instrument compares with the latency of a real piano?  I have noticed some people are trying to achieve less than 5ms latency with Pianoteq and not achieving this can be an issue.  Do people who want very low latency with Pianoteq desire this because it makes things easier in the electronic world with other synthesized sounds or are they trying to more closely approximate a real piano?  Is this more a personal preference kind of issue?  I am asking this question out of curiosity.  I am just trying to get a little smarter on the subject as I have limited background with this..  Not sure if this was the right place to make this post or if it should have been in one of the other threads.  Thanks.

Reply by Jake Johnson

  • Avatar
  • Jake Johnson
  • Member
  • Offline
  • From: Atlanta
  • Registered: 04-05-2007
  • Posts: 4,491

Re: Five lectures on the Acoustics of the Piano

My impression is that in the lectures, the time delay is measured from the point at which one touches the key, while the latency, as the term is usually used in working with synths, is the time that passes between when the key touches the key bed and when a note is sounded. In other words, it's apples and oranges.

Last edited by Jake Johnson (17-01-2009 15:24)

Reply by mwinthrop

  • Avatar
  • mwinthrop
  • Member
  • Offline
  • Registered: 04-07-2007
  • Posts: 61

Re: Five lectures on the Acoustics of the Piano

Jake - I agree with your assessment.  However, I am still curious if anyone knows what the latency of a real piano is compared to a software piano (comparing apples to apples that is).  Thanks again for posting the lecture series - the information is very interesting.

Reply by Glenn NK

  • Avatar
  • Glenn NK
  • Member
  • Offline
  • From: Victoria, BC
  • Registered: 09-10-2008
  • Posts: 1,008

Re: Five lectures on the Acoustics of the Piano

First, thanks to Jake for posting this.  I'd seen this a few years ago, and lost the link/bookmark, and was looking for it so this is very timely.  Thank you.

On Latency:

The only problem I've had with Pianoteq, is when playing at the keyboard or playing a midi file in real time; I was hearing fairly frequent "clicks".

With an EMU 1820M S/C and a fairly fast dual core AMD, using ASIO, I set the latency to 5 ms because that's what I had heard that was needed for live playing or the delay between pressing a key and the issuance of sound would be annoying.

However, the clicks became more annoying with time, so I asked the guys on Multimedia Production Forums for help.  After checking everything out with the computer, one fellow suggested that 5 ms simply wasn't required and gave some convincing evidence.

So I set my ASIO latency to 20 ms.  I could not discern any delay at all (and the clicks disappeared).  I'm tempted to set it at the next available setting (50) to see what happens.

__________________________
Procrastination Week has been postponed.  Again.