28th September 2011
|
#1 | | Gear nut
Joined: Jan 2008
Posts: 145
Thread Starter  | Impossible super-8 mystery!
I was asked to remaster the soundtrack of a super-8 movie from 1979. In the movie a woman is playing a Beethoven sonata. Well, it looks like she is playing, but actually a track from an LP was used when the film was made. Youtube recognized the wave, so I could easily get hold of the right recording in order to replace the poor mono soundtrack with a hifi stereo track. Here is where things got weired! It's the same recording, no doubt about it. It not only sounds the same, but the waveform matches. However, the track plays a lot faster than in the original super-8 movie. This would make sense (since the projector used back in 1979 could run to fast) if the two tracks were out of pitch. BUT THEY ARE NOT! Both tracks are perfectly tuned in relation to each other! I simply can't figure this out!  | 
|  | |
28th September 2011
|
#2 | | Lives for gear
Joined: Apr 2010 Location: London, UK
Posts: 734
| Quote:
Originally Posted by Franz  I was asked to remaster the soundtrack of a super-8 movie from 1979. In the movie a woman is playing a Beethoven sonata. Well, it looks like she is playing, but actually a track from an LP was used when the film was made. Youtube recognized the wave, so I could easily get hold of the right recording in order to replace the poor mono soundtrack with a hifi stereo track. Here is where things got weired! It's the same recording, no doubt about it. It not only sounds the same, but the waveform matches. However, the track plays a lot faster than in the original super-8 movie. This would make sense (since the projector used back in 1979 could run to fast) if the two tracks were out of pitch. BUT THEY ARE NOT! Both tracks are perfectly tuned in relation to each other! I simply can't figure this out! | maybe it was a slightly different recording? have a friend who is a Standard-8 and a Super-8 genius, will ask him!
| |
| | |
28th September 2011
|
#3 | | Gear maniac
Joined: Aug 2008
Posts: 177
|
I think the issue you have is a film speed versus NTSC speed issue. The difference is .1%. At this rate they sound really really close in terms of pitch, but they will drift against each other.
Have you calculated what the drift is yet?
Randall Quote:
Originally Posted by Franz I was asked to remaster the soundtrack of a super-8 movie from 1979. In the movie a woman is playing a Beethoven sonata. Well, it looks like she is playing, but actually a track from an LP was used when the film was made. Youtube recognized the wave, so I could easily get hold of the right recording in order to replace the poor mono soundtrack with a hifi stereo track. Here is where things got weired! It's the same recording, no doubt about it. It not only sounds the same, but the waveform matches. However, the track plays a lot faster than in the original super-8 movie. This would make sense (since the projector used back in 1979 could run to fast) if the two tracks were out of pitch. BUT THEY ARE NOT! Both tracks are perfectly tuned in relation to each other! I simply can't figure this out! | | |
| | |
28th September 2011
|
#4 | | Lives for gear
Joined: Apr 2010 Location: London, UK
Posts: 734
| Quote:
Originally Posted by quadraphonics I think the issue you have is a film speed versus NTSC speed issue. The difference is .1%. At this rate they sound really really close in terms of pitch, but they will drift against each other.
Have you calculated what the drift is yet?
Randall | Except Super8 runs at 16 or 18fps most usually.
| |
| | |
28th September 2011
|
#5 | | Lives for gear
Joined: Dec 2005 Location: San Francisco area
Posts: 2,880
| Quote:
Originally Posted by tom_lowe Except Super8 runs at 16 or 18fps most usually. | We did lots of Super 8 at 24 too. Or I should say "sort of 24" since almost all of the cameras available were pretty inaccurate speed wise, both overall and in relation to where in the roll you were at a given moment.
I'd say the most logical explanation is a different performance of the same piece with both tuned to standard concert pitch.
phil p
| |
| | |
29th September 2011
|
#6 | | Lives for gear
Joined: Dec 2005 Location: London, UK | | |
| | |
29th September 2011
|
#7 | | Lives for gear
Joined: Dec 2004 Location: Finland
Posts: 3,949
|
The solution?
I would adjust the video pic track to the audio, just slightly
Well to make them sync
Matti
| |
| | |
31st October 2011
|
#8 | | Gear nut
Joined: Jan 2008
Posts: 145
Thread Starter |
Thanx to you all!
I´m leaning towards the explanation that the piano piece was recorded twice, at the same recording session. Cause even if the pitch
can occure close when one frame is edited out, in this case the pitch is exact! The film is made in 18 f/s by the way. And yes, even
Super 8 had the opportunity to go 24!
| |
| | |
31st October 2011
|
#9 | | Gear interested
Joined: Sep 2011
Posts: 13
| Quote:
Originally Posted by Franz Youtube recognized the wave, so I could easily get hold of the right recording in order to replace the poor mono soundtrack with a hifi stereo track. | Franz, curious as to where on YouTube you can get a track analyzed and matched?
| |
| | |
8th November 2011
|
#10 | | Gear interested
Joined: Feb 2009
Posts: 8
| Quote:
Originally Posted by jmsound Franz, curious as to where on YouTube you can get a track analyzed and matched? | Me too!
| |
| | |
8th November 2011
|
#11 | | drake.ch
Joined: Jan 2008
Posts: 505
| Quote: |
It's the same recording, no doubt about it. It not only sounds the same, but the waveform matches.
| Quote: |
However, the track plays a lot faster than in the original super-8 movie.
| these two things are mutually exclusive
| |
| | |
8th November 2011
|
#12 | | Gear nut
Joined: Aug 2009 Location: Toronto
Posts: 148
| Quote:
Originally Posted by jimdrake these two things are mutually exclusive | That was my thought. Then I thought, "What if someone took the original recording and time-compressed it for their YouTube video, and that's the only copy YouTube has?"
__________________
Sonny Keyes
Ricochet Audio
Toronto
| |
| | |
8th November 2011
|
#13 | | Lives for gear
Joined: Sep 2008 Location: Melbourne
Posts: 545
| Quote:
Originally Posted by tom_lowe Except Super8 runs at 16 or 18fps most usually. | I used to shoot super 8, I always shot 24fps
| |
| | |
12th November 2011
|
#14 | | drake.ch
Joined: Jan 2008
Posts: 505
| Quote:
Originally Posted by sonnykeyes That was my thought. Then I thought, "What if someone took the original recording and time-compressed it for their YouTube video, and that's the only copy YouTube has?" | then the waveform would not match and it would not sound the same.
| |
| | |
3rd May 2012
|
#15 | | Gear nut
Joined: Jan 2008
Posts: 145
Thread Starter | Quote:
Originally Posted by jimdrake these two things are mutually exclusive | A bit late perhaps, but I´ve been busy... 
Indeed the waveforms can't look different and sound the same. But I have not looked at the waveforms. I wrote that the waveforms "are the same" because
Youtube analyses waveforms in order to find what music is used in the clips.
This answers the other question. You don't have to do anything to get the soundtrack analysed. This is done automaticly, using ferry dust and black magic...
(Or some advanced device matching every soundtrack with everything that was ever recorded.)
| |
| | |
3rd May 2012
|
#16 | | drake.ch
Joined: Jan 2008
Posts: 505
| Quote:
Originally Posted by Franz Youtube analyses waveforms in order to find what music is used in the clips. | If youtube is automatically detecting songs (i didn't know they did this) then they will do it in the same way that shazam does.
Have you ever used shazam? You can hold a phone up in a room when a song is playing and it will tell you the song. It's crazy that it works, and it will be tolerant of pitch shifts, time stretching, massive distortion.
it looks at things in the frequency domain, not time domain. Hence it is not analysing the waveform.
| |
| | | |
New Reply 
