Audio delay in headphones

[Pete Cofrancesco]

When I plug a microphone directly into my Panasonic UX90, there’s a delay in my headphones making it difficult to monitor. If I plug the mic into a mixer and send a line feed to the camera, the delay is eliminated. The audio is fine when I imported into my computer.

Edit: I found the answer to my own question

Originally Posted by Len Rosenberg

Re the headphone audio delay, I found this answer posted by Barry Green at DVXuser:

"Don't think you can do anything about the video lots of processing before it is on the LCD, but on page 190 in the manual you can set the audio to be a live sound or processed in sync with the video to make it easier to monitor when recording a live event.

The LIVE option is to avoid echo in the headphones. The RECORDING option is to match the output.

There is some lag, yes. It takes a little while to route the video through all the processing and then to the outputs. If you listen to the headphones in RECORDING mode, it'll delay the headphone audio to match the video, but then you'll hear a distinct echo between the live spoken words and the headphone delayed words, which can cause a noticeable and distracting echo. So I always put the headphones in LIVE, to avoid that.

As far as delay on the LCD, it's about 3.5 frames. I set up a timecode generator on my phone, and filmed it through the UX180, and then took still frames of both the camera's LCD and the phone's display, and it showed a consistent 3 to 4 frame discrepancy between them.

3 frames is enough to notice a lip sync error, yes. So 3-4 frames, you probably would have been seeing video out of sync with the audio if you were viewing with LIVE set on the headphones. You can choose RECORDING; the only issue is whether you'll hear the echo or not. If you have headphones that do a great job at sealing out outside noise, then RECORDING may be the ideal choice in that situation."

[Paul R Johnson]

Forgive me, but that's a simply terrible 'feature' for a camera to have. It's bad enough the delay going to screens on many HDMI outputs, but for a camera to have internal processing delay so you have a choice of sound and LCD in sync or sound in sync to live that's awful. Does this also mean that when you pan on this camera, the LCD lags behind? I couldn't work with a laggy viewfinder. It's not mega expensive, but neither is it a cheap camera. That's pretty bad!

[Pete Cofrancesco]

I have to agree for a camera to default to audio delay to the headphones is pretty bad. I’ve never noticed a video delay. So this has been going on for a while but I didn’t realize what was happening in stage performance I had a line feed from the board in one channel and either shotgun or internal mic for ambient in the other. I believe the internal has no delay and the shotgun I was assuming was the delay based on time difference between the long distance from stage, while that probably was true I had no idea the camera was adding its own delay.

[Christopher Young]

I notice this is becoming more of a recurring question with late model digital cams although I think it is correct to offer this capability to delay or not delay audio for the following reasons.

If the audio going in to be recorded is analogue there is virtually no processing in the signal path. With video recording every frame that is recorded is one frame behind real time. First off the frame has to be scanned by the sensor then processed and then exported to the recording medium. Hence a one frame delay. Then there are more delays in processing the signal to a VF or video output. This delay doesn't happen with the audio path in EE mode so it can go straight through with only the minimal delay involved within the audio circuit.

Normally visually while shooting and watching an interview and listening to the audio in EE mode all seems right with the world and all is in sync. If you delayed the audio path to match the video frame processing delay of the recording process your ears and brain would be saying that what you were hearing is out of sync with what your eyes are seeing as you look at your interviewees lips.

Different story though if you watch closely on your camera VF, screen or a monitor and listen to the EE audio you will detect that the audio is actually ahead of the delayed processed video that is being displayed on your VF, screen or monitor. This can be up to 4-5 frames with some cameras, especially certain mirrorless cameras. For the cameraman it is far less disconcerting to listen to the EE audio as it is in sync with what his eyes see of the subjects lip movements.

Next scenario. Have a bunch of cameras in a studio or on an OB with crew in a control booth. The control room crew can't see your subjects lips except on their monitors. If they listen to an EE audio feed coming from audio mics in the studio it is going to look out of sync. In that scenario you introduce an audio delay with a delay unit to match the video timing before it goes to air and recording. They need to hear audio that is in sync with the vision that they see on their monitors in the control room.

On the other hand if they want to hear audio coming directly from your camera headphone output such as you are listening to then the camera needs to have the ability to delay the audio to match the output vision they are seeing. If the audio is coming down the line as embedded audio in an SDI feed then there isn't any issue as that is re-clocked and synced with the vision output of the camera. On a lot of the late model Sony cams there is now an audio menu item selection labeled: LIVE or LIP SYNC which can be selected depending on the viewing situation, eyeball to eyeball or in the control room.

Some of you with home theater setups know that you can end up with audio ahead of video, same with certain set top boxes because of the delays in the video processing path. Most decent systems have the ability to delay the audio to match what you are seeing on your screen. Exactly the same process is happening there as I've described above.

When it comes to event recording, such as a live shows and with sports like tennis or football you have to remember that sound travels 342 meters (1120 feet) per second. When that sound is arriving EE in the van or studio from mics and or line and wireless feeds from down near the stage or sports ground all the distances from the sound source to the mics plus any delay in the sound path back to the studio or van must be calculated. That is all part of the whole equation of mixing commentator sound, tennis ball on racquet sound, crowd sound and any other nat sound sources. The result of that sound equation then has to be matched to the vision timing to keep all in sync for the viewer.

Constantly with live recordings you have re-time the live mic sources with the line feeds and if you know the distances involved in the sound source to mic sound paths it easy enough to calculate the time delay and from that the vision frame delays involved using the 342 meters a second for sound transmission through air figure.

Usually what I do for a simple stage show recordings in post is to calculate the distance to the camera mic and then advance or delay that audio track or tracks to match the vision and then re-sync any audio desk feeds to the recalculated live audio track/s.

Chris Young

[Paul R Johnson]

Maybe I'm too old, but I'm finding living with video delays extremely annoying. I've worked on so many shows this year with totally unacceptable delays, and it's impossible for camera ops to follow fast action and snap focus if the viewfinder lags? What is the point of 4K if finding the focus point means you have to hunt as you constantly overshoot the mark? Even worse - most of my stuff is music based.so hearing a drummer and seeing a delayed one really grates. HMDI video, often stacked through the distribution chain can mean unworkable delays. sticking with SDI is so much better - but with cameras adding unacceptable delay internally, maybe latency needs adding to the spec like in audio products.

[Christopher Young]

" ...it's impossible for camera ops to follow fast action and snap focus if the viewfinder lags?"

Understood. I shoot a lot of broadcast football plus live entertainment shows.

My first real experience of this lag/delay problem was when I first worked with a Sony FS700. The HDMI delay was totally unacceptable. As far as I'm concerned for stage shows and fast moving material like sports the one frame delay on SDI I'm used to on broadcast cameras is the max one should have to work with. I've noticed that some of the lower end cameras even with SDI can have a couple of frames delay. Why this is I don't know.

I think the problem is even more noticeable with DSLR/Mirrorless cameras because they were never designed from the get go to be video cameras in the professional sense and so don't have the necessary processing power to process and pass through the signal within a one frame window. It was never a priority in their design requirements. HDMI is another issue as it was designed as a domestic semi-permanent installation connection. With HDMI so what if its signal processing from the network or streaming source is is few frames delayed. Different when you are trying to use it in a professional shooting application.

Chris Young

[Don Palomaki]

Bottom line is audio leading video is unnatural. In the real worlds, the sound wave never leads the light wave, it lags it by the distance/speed of sound. Our ear-brain system understands this.

FWIW: Digital wireless mic system have an audio lag, perhaps 1/3 frame or so depending on the system.

[Paul R Johnson]

The good thing is that for practically all purposes, the latency is low enough nobody notices - BUT - in live sound situations, it actually reduces the tendency for feedback to start, although when it does, it comes in quickly and savagely!

[Christopher Young]

Sometimes the delay is noticeable. Broadcast requires NO frame delay on audio. Sync must be 100% on frame. The problem arises in some case when you do an interview with a boom and a wireless and try to mix them only to find out there can be a delay issue. These newer Wi-Fi range mics can be problematic. I found this to be the case especially with the Sennheiser EKP AVX 1,880 - 1,900 MHz wireless mics. The EKP AVX has a 19ms delay which is approx 1 frame in the PAL world and when editing I can pick a one frame sync issue without any difficulty. When you use a clapper board and line up the boom and one of these Sennheiser EKP AVX mics on the timeline you see the problem straight away. Also if the video was recorded on some of the earlier Canon DSLRs you will find another one frame delay from the camera added into into the mix. Two frames of delay stands out like the proverbial dogs b******s. With the Rode 2.4 GHz wireless mics we measured a 4 milliseconds delay. Just about acceptable for non broadcast work.

Chris Young

[Pete Cofrancesco]

In this case the delay is due entirely to how Panasonic has set up the camera to monitor audio. Happily I was able to resolve it by setting audio to “live” instead of the default “recorded”.

My feeling with Panasonic and this camera line there are a number of usability issues which haven’t been thought through. I talked to a Panasonic rep at NAB but received an unsatisfactory answer, something to the effect that the way it was when film and audio were recorded separately. While that’s true it really doesn’t explain why you would setup a modern digital camera that way.