Author Topic: External FPS control via Audio input (Genlock/3D/Speed Ramping)  (Read 7517 times)

Tom C. Hall

  • New to the forum
  • *
  • Posts: 10
External FPS control via Audio input (Genlock/3D/Speed Ramping)
« on: October 28, 2013, 10:51:46 PM »
Would it be possible for the audio input frequency be used for live FPS override? Kind of like a Sync plus generator but over sound instead of an electrical input.
http://en.wikipedia.org/wiki/Video-signal_generator#Sync_pulse_generators_.28SPG.29

A sound file patterned off the trilevel sync wave would be pretty easy to generate and then playback as an MP3.


Possibilities would be interesting, such as 3D Sync, or any sort of multi cam shooting, as well as by ramping the wave from go from very low FPS to the highest FPS possible. Using any live pitch adjustment tools you could speed ramp the FPS in real time. Would be like ramping the motor on old ARRIFLEX S for sudden slow motion. The movie Hotfuzz used the trick a lot in the final shootout.

It could be played back from any sort of audio device, like an Iphone, or sent wirelessly via normal audio transmission.

Of course it would disable any sound recording. Or maybe it just uses one side of the stereo, and the second stereo channel would still be open to normal audio. Also it would have to be limited to what the cameras are capable of doing frame rate wise.

a1ex

  • Administrator
  • Hero Member
  • *****
  • Posts: 12564
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #1 on: October 28, 2013, 10:58:16 PM »
Technically possible. I don't have any use for it myself, so feel free to research it.

Tom C. Hall

  • New to the forum
  • *
  • Posts: 10
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #2 on: October 29, 2013, 04:30:57 PM »
Another idea i had today would to have a function if the camera has a headphone jack to play this sync signal out as master clock, with the other cameras as a reciever. Also an function to adjust the timing on this signal input to fine tune the delay to get perfect sync.

Tom C. Hall

  • New to the forum
  • *
  • Posts: 10
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #3 on: October 29, 2013, 10:13:47 PM »
I was thinking of alternates to running the pulse for every frame. It may make it difficult in the case of power failure that the camera may behave in unexpected ways. Knowing from experience shooting with DSLRs in Stereo myself that they drift a whole frame pretty consistently over 5 minutes. By pure bullshit guess work that makes me feel like the first 24 frames if started from the same start signal would be for all purposes locked.

Would it be possible to wait and check an external audio clock for their start pulse, then recheck every 24 frames by holding and waiting for the next pulse to drop?

Of course that limits the option to speed ramp or use alternate frame-rates. (other than ones that are multiples of 23.976 or 29.98) but that's also how it works on a system like Red Epic. I'm also reading about hardware options to convert a standard genlock signal into something that can go into MIC LEVEL. That way off the shelf gen lock devices could lock the cameras. Of course just playing an MP3 from an Iphone is also a very appealing aspect.

I am not a programmer myself but I am trying to read code and convince my programmer friends to help me with this project. I have access to half a dozen professional 3D beamsplitters at any given moment but getting two epics or alexas for low budget projects is always more trouble than it's worth. Being able to shoot with DSLRs would really open up 3D to the low budget market.

S3Dcentre

  • New to the forum
  • *
  • Posts: 6
  • Emily Carr University's Stereoscopic 3D Centre
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #4 on: November 06, 2013, 11:52:51 PM »
Hey Tom,

Get in touch with us.

Patryk

g3gg0

  • Developer
  • Hero Member
  • *****
  • Posts: 3182
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #5 on: November 08, 2013, 12:49:50 PM »
the main problem i see is to measure the delay of the audio driver etc (DMA, I2S, DAC and the other way respectively ADC, I2S, DMA).
i dont know how to *reliable* quantify the delay in every single step.
we barely know how they work and getting a clear picture of all the delays would take about 2 weeks i guess.
plus we would need measurement equipment (e.g. a picoscope)

when knowing all this, all we can do with software and the current understanding would be some kind of soft sync, which is not a real genlock.
softsync means, we measure how much we are off the source clock and try to correct the delays so the exposure start is close to the audio trigger.
still the real exposure start can also jitter and we dont know how much that is again.

audio signal --> ADC --> I2S --> DMA --> software --> exposure timer --> software --> exposure start

and every single element in that queue is probably running asynchronously with unknown delays and jitter.
Help us with datasheets - Help us with register dumps
magic lantern: 1Magic9991E1eWbGvrsx186GovYCXFbppY, server expenses: paypal@g3gg0.de
ONLY donate for things we have done, not for things you expect!

ItsMeLenny

  • Hero Member
  • *****
  • Posts: 928
  • 550D
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #6 on: November 09, 2013, 03:40:33 AM »
Hey Tom,

Get in touch with us.

Patryk


Hey Tom,

Come hop into my van, I have all the candy you want.

Lenny

Tom C. Hall

  • New to the forum
  • *
  • Posts: 10
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #7 on: November 16, 2013, 12:01:15 AM »
the main problem i see is to measure the delay of the audio driver etc (DMA, I2S, DAC and the other way respectively ADC, I2S, DMA).
i dont know how to *reliable* quantify the delay in every single step.
we barely know how they work and getting a clear picture of all the delays would take about 2 weeks i guess.
plus we would need measurement equipment (e.g. a picoscope)

when knowing all this, all we can do with software and the current understanding would be some kind of soft sync, which is not a real genlock.
softsync means, we measure how much we are off the source clock and try to correct the delays so the exposure start is close to the audio trigger.
still the real exposure start can also jitter and we dont know how much that is again.

audio signal --> ADC --> I2S --> DMA --> software --> exposure timer --> software --> exposure start

and every single element in that queue is probably running asynchronously with unknown delays and jitter.

Soft sync sounds much better than the current turn over and hope tha we have now. Unfortunately I have been very busy lately so I haven't made much progress. I am interested to hear what the ECUAD people have done, Patryk.

Tom C. Hall

  • New to the forum
  • *
  • Posts: 10
Re: External FPS control via Audio input (Genlock/3D/Speed Ramping)
« Reply #8 on: January 06, 2014, 10:29:44 AM »
the main problem i see is to measure the delay of the audio driver etc (DMA, I2S, DAC and the other way respectively ADC, I2S, DMA).
i dont know how to *reliable* quantify the delay in every single step.
we barely know how they work and getting a clear picture of all the delays would take about 2 weeks i guess.
plus we would need measurement equipment (e.g. a picoscope)

when knowing all this, all we can do with software and the current understanding would be some kind of soft sync, which is not a real genlock.
softsync means, we measure how much we are off the source clock and try to correct the delays so the exposure start is close to the audio trigger.
still the real exposure start can also jitter and we dont know how much that is again.

audio signal --> ADC --> I2S --> DMA --> software --> exposure timer --> software --> exposure start

and every single element in that queue is probably running asynchronously with unknown delays and jitter.

 if the delays are constant it would be pretty straight forward to dial in the offset, maybe with a menu function. The trick I used to test sync is a swinging weight on a string, the difference in timing is very easy to see as one camera will see it going one way and the other camera will see it changed direction and going the other way.

Would something like a PID work for calculating the difference? It also might not be nessesary to run it over sound, what about through the shutter port?