OT: XLR Length and Latency

jrazz wrote on 12/19/2011, 9:08 PM
I have a Sennheiser Evolution g2 and a Sennheiser Evolution g3. I also have a couple of Sony HVR-A1u's that I pair them up with.

Any idea on how far I could extend the XLR cable before I would introduce latency into my video/audio sync? The cable that came with the Sennheiser is about a foot long (which makes sense for on camera usage). Could I go 25 feet? 75 feet? What would be the limit?

Any input or resources on this would be appreciated before I buy some cabling.

j razz


musicvid10 wrote on 12/19/2011, 9:18 PM
"Any idea on how far I could extend the XLR cable before I would introduce latency into my video/audio sync?"

If the wireless and wired mics are on the same physical plane, there will be no measurable difference at the board except for a possible phase shift in the wireless circuitry. Since the wireless mics are constantly moving, we don't bother trying to match phase with stationary mics. But if you're curious, just put the mics next to each other and record a sine wave test.

We think of 150' of balanced cable at mic level (with no ground loops) as being completely trouble free. At 250' losses start to creep in unless you have very high quality cable.
ChristoC wrote on 12/19/2011, 9:25 PM
The speed of light is approximately 3x10^8ms (exactly 299,792,458 m/s), but electric signals don't travel that fast in conductors. They travel anywhere from 40% up to 95% of that speed, depending on properties of the cable (the dielectric constant of the material surrounding the copper is a big factor).

Twisted pair copper cabling tends to be on the lower end, somewhere between 40% to 70%. So a good estimate is somewhere between 120,000,000 m/s to 210,000,000 m/s (m/s = metres per second).

Now conservatively assuming troublesome latency could begin at around 10 milliseconds, which is 1/100 of a second.... therefore a potentially troublesome length of wire would be at between 1,200,000 and 2,100,000 metres! ( = approx 3,937,000 and 6,890,000 feet, .... or 746 and 1,305 miles)

You will have zero problems at the (minute) distances you are considering!

larry-peter wrote on 12/19/2011, 9:28 PM
j razz,
No need to worry about latency. Those electrons aren't quite at the speed of light but you'll run into attenuation and noise from cable length several miles before you lose sync. In using good quality cable, I've run as far as 150 ft. with line level signals, which I'm assuming is what the Evolution receivers output. That's what my Lectrosonics output.
I've never run a phantom powered mic more than 100 ft. so I can't say if attenuation of the phantom power has an effect at greater distances, but I think you'd be quite safe with any of the lengths you're considering.
farss wrote on 12/19/2011, 9:38 PM
All of the above are correct but I wonder if the question is being asked the right way around?

If you run a 100M mic lead 100M away from the camera then the signal from it will arrive at the camera before the sound it is carrying reaches the camera. If you're using any of the camera's mics that far from the sound source then you wil have a significat difference between the timing of the two signals.

Chienworks wrote on 12/19/2011, 9:39 PM
I've run mics through 1200' of computer grade twisted pair 22 gauge cable with a foil shield and drain wire. The signal was still perfectly clean and strong. There might have been a little loss at the high & low ends of the frequency spectrum, but probably outside of the frequency range of the mic anyway.

The stuff off the shelf at the local music store i'd trust up to maybe 250'.
Geoff_Wood wrote on 12/19/2011, 9:44 PM
"Latency" has to be the most mis-used and misunderstood term of the year !

The internal digital processing in the camera or soundcard will swamp any signal 'latency' by many orders of magnitude.

You will have no problem with signal delay unless the mic is several thousand kilometres from the camera.

Geoff_Wood wrote on 12/19/2011, 9:45 PM
Certainly, about 30cm (1ft) per ms.

Easily re-aligned wil some any notable 'event' on the timeline.

jrazz wrote on 12/19/2011, 10:41 PM
Thanks all for the replies. I will plan on running an XLR cable about 100 foot then (the distance from where the camera will be place and the sound booth).

MusicVid, this will not be sync'ed with any other audio, just the video being recorded on the camera. The 2nd XLR input will not be used- the sole audio will be captured from the sennheiser wireless mic.

Geoff, I understand what you are saying- I was using latency in a loose sense referring to delay and sync alignment due to length of media (cable).

Thanks again for the quick response.

j razz
Steve Mann wrote on 12/20/2011, 10:07 AM
I think your original question escaped the response torrent. Everyone latched onto your length of the cable and missed the question that you are getting audio from two sources.

When its available, I use house sound which comes from talent microphones on the stage, plus some studio microphones at the front of the stage. (Not all talent is miked). There is no noticeable delay between those two sources. When logistics don't permit me to put the studio mics near the stage, but in the front of the house near the cameras, there is a noticeable delay, but it's less than one frame of video for about 100 ft of separation.
rraud wrote on 12/20/2011, 11:22 AM
Warning: If you are using the SK100 G2/3 portable receiver, the output is unbalanced, so long cable runs (especially at mic level) may induce noise. If you don't have a balanced line driver available, run the G2/3 receiver's AF level at it's highest, and adjust the receiving input stage accordingly. Beware.. this will be higher than mic level, so you will want to use line level... however it will still be unbalanced, so a line driver is still recommended.
johnmeyer wrote on 12/20/2011, 2:54 PM
As already noted, the cable length doesn't matter at all because the signal travels at the speed of light. You will get no delay by using a longer cable.

However, the difference in distance from the audio source to various microphones is very important, and you will have to compensate for that.

At sea level (and standard temperature, blah, blah), sound travels at 1,125 feet per second. In NTSC land, there are nominally 30 frames per second. Thus, sound travels 1125 / 30 = 37.5 feet for every frame of video. Thus, for each 37.5 feet (11.4 meters), you must shift the audio by one frame (move it to the left). For PAL, the calculation is 1125 / 25 = 45 feet for every frame of video.

The effect is pretty significant even in situations where you wouldn't think it mattered. A few years ago I shot a fashion show in a high school gym. The stage was behind one baseline. I had a feed from the soundboard which had mics on the emcee and other talent, and was feeding speakers set up at the front of the stage. I collected ambient sound from my main camera which was set up at the other end of the basketball court. The length of a basketball court is ninety-four feet, and as you can tell from my calculations above, I had to shift the audio by three frames in order to get sync and avoid introducing an artificial echo between the soundboard and the ambient sound, which I was mixing together.

I'm not sure how Steve calculated his "less than one frame of video for about 100 ft of separation." Maybe my calculations are wrong. If so, please let me know.

ChristoC wrote on 12/20/2011, 3:08 PM
Quite correct - there's an easy to remember ready reckoner: sound travels approx 1 foot per 1 millisecond.
Using that + simple finger&toe math:
@PAL 1 frame ~ 40 feet
@NTSC 1 frame ~ 33 feet
johnmeyer wrote on 12/20/2011, 3:46 PM

Nice easy-to-remember rule of thumbs. I'll use that.
jrazz wrote on 12/20/2011, 9:05 PM
Thanks John for the in depth understanding and Christo for the easy to remember "rule of thumb" :)

Maybe this will help clarify (however, I do believe I have the answer to my question already): a speaker will be filmed and recorded. There are two recievers- one that is on cam and the other that is connected to the sound board. The on cam receiver is the one I am wanting to move- off the camera and back to the sound booth (100 feet away). There are no other mics used except the 1 lapel mic on the speaker. The sound will be run through the mixer board and out the speakers in the auditorium but that will be through the other receiver that sits in the sound booth.

I hope that makes sense. 1 mic, 2 receivers, 1 camera- that's it.

j razz
ChristoC wrote on 12/20/2011, 9:18 PM
OK, then the main consideration is what rraud pointed out - if that receiver is 'bodypack' EK100G2/3 it has an unbalanced output, and the 100 foot run may be problematic for unwanted noise and interference pickup. If its the 'rackmount' receiver, then the XLR balanced output should be fine.....
Why move the receiver away from the camera?
musicvid10 wrote on 12/20/2011, 9:23 PM
"if that receiver has an unbalanced output the 100 foot run may be problematic"

Two words - Direct Box
SK receiver will output a healthy unbalanced line level (+12dBu). Lift the ground on the direct box. End of problem.

On the other hand, EW 100/300/500 rackmount receivers all have balanced out -- all you need is a TRS->XLRM cable for the EW100, or a regular XLR for the others.
jrazz wrote on 12/22/2011, 3:26 PM
Why move the receiver away from the camera?

The camera can take power directly from the outlet- the Sennheiser cannot as it runs soley off 2 AA batteries. I need to be able to change the batteries.

j razz
ChristoC wrote on 12/22/2011, 9:15 PM
There's an accessory for those bodypacks which allows you to run them from external 12V....
rraud wrote on 12/23/2011, 11:44 AM
I need to be able to change the batteries.
> Yes, but only every 12 hours or so, in my experience with rechargeable AAs.