Feaure audio going direct to camera

[Chris Pickle]

Here you seem to indicate that there is really no difference between plugging straight into the camera versus using a sound mixer (human) and sound mixer (device) and as a matter of convenience, it is better to just go straight in. I take exception to this analysis. What it leaves out, are all the duties of a sound mixer that are critical to getting quality audio. Things such as... ensuring proper gain staging, pre fader / post fader levels, detecting phase anomalies from a lav and boom mic, for example, adjusting boom or boom op and other monitoring duties... something that the AC or whoever is listening to camera audio is going to be too busy to do. Most camera people I know would prefer to not even really listen to audio and just leave it up to sound dept.



Yes it can be, but it can also be easy. In my opinion, those who have problems in post are either inexperienced, have not put proper attention into timecode or even basic cinema sound synch details, or continue to butt their heads against the same basic workflow issues without addressing them. Being inexperienced is not bad, everyone is at some point, and I'm not implying that you are. But if we assume that professional tools that are designed for professional results require some professional expertise and competence to operate - there should be some leniency allowed for the process. What makes synching sound on RED more difficult that any other film camera? In fact, RED has TC input / output, film cameras do not. RED operates at the same speed, no matter what the AC mains speed is. No longer do you have to deal with AC pilot tone resolvers and adjusting motors for both camera and audio reel to reel recorders. Not everyone can afford a timecode workflow. If you can, there are thousands of ways you can mess something up, but a very little bit of research and education on TC goes a long long ways and if everyone in the workflow understands their responsibilities, it really can't get much more automated and easy to work with. You going to make mistakes, sometimes big ones, the first couple of times you work with timecode. I know I did. Just make sure those mistakes are made on non-critical test runs and not real deal productions. For those that are on a budget and can't afford decent lights, let along $2k smart slates and $2k to $14k audio recorders with calibrated master clocks, just use proper discipline with a dry erase marker, old fashioned slate, an audio mixer with slate mic, tone generator, and the people willing to adhere to proper methodology. Your DIT can slate the scene with proper notations on the slate, your editor will see the scene and take information on his screen. The mixer can enable the slate mic while the scene info is being called out and now your editor can hear the scene information at the beginning of the audio file. Mixers such as the Sound Devices 302 can precede the slate mic with a test tone generator pulse. This pulse shows up as a nice square constant level spike on the audio file's WAV pattern display. It is unmistakeable and allows one to easily get the audio recorder's file and the camera's scratch track audio to line up... you just zoom in and make the squares line up. The actual slate's clap will be an even more distinct marker that can allow even more exact alignment, its just that the tone's signature will be more visible in the timeline. I'll mention this one last thing, because it is probably the root of many people's issues. In the US, When most of us talk about 24p, we are really talking about 23.976 and when talking about 30p, we really mean 29.97. I see a lot of indy filmmakers who know they want to shoot 24p, actually set their cameras to 24p rather than 23.976. The only people who should really use 24p are those that intend to do a filmout and even that is debatable, as the content will almost certainly be destined for DVD, Blu-Ray, or broadcast release too. When someone runs true 24p frame rates, as an audio recorder, we have to pull up our master clock's speed so instead of recording 48 KHz, our clocks actually have to run at 48.048 KHz. You also have to correctly set all drop frames / non drop settings on all cameras, recorders, and pull up / pull down in the post workflow when the editor ingests the content into a workstation. A lot of people choose the 24p / 30p setting, because it sounds simpler, but in reality, most people should really be choosing 23.976 / 29.97 just to keep things in synch with regular NTSC standards and timings.



I know that there are lots of major projects that ADR absolutely everything. I also know that there major projects, even feature films, that have ADRd only mistakes or in the more rare case, nothing at all. Sometimes they will take the audio mixer's 2-track mixdown and use that untouched. This is of course becoming more rare and the duties of the sound mixer are changing towards more ISO'd tracks and less actual mixing. If you want me to get useable audio next to a highway, there is probably something I can do to make that happen, although it is going to take time and specialized tools.



I absolutely agree. One is simply not able to realize 24 bit resolution in even a studio environment. A typical audio system won't be able to take advantage of around 16 to 18 bits worth of audio fidelity. Really high end systems might get up to 20 bits. Most modern equipment is automatically 24 bit capable, so might as well take advantage of the off the shelf converters that use this word length, but at the end of the day it is the analog circuits in front of and behind the digital chips that really determine sound quality.



If you mean by recording audio line level via a sound mixer, I concur, without the basics, you would be blowing your budget on external recorders, unless you hire a mixer who rocks one anyways. If you mean by plugging the mics directly into the camera, I would have reservations. If there is anyway that you can do it another way, I would highly suggest that you do so. It's not only because of the fact that the RED has substandard mic preamps, which I believe it does, but also the fact that recording audio should not be a set and forget activity, that someone should really be adjusting gain levels for each situation, if not continuously as necessary.

I wasn't suggesting not using a recordist to monitor levels, get good sound, etc. I just meant that instead of sending that sound to a second system, send it to the camera. A great sound person is still very vital.

Syncing in post is a nightmare if you didn't use some sort of time code approach and then use some sort of automated tools. The issue is marrying the audio to the video. In film, the transfer house would match it to your dallies and the audio would be part of the video at ingestion. that's no longer the case with Red. It's actually a massive step backward. You take your video files and then manually sync the audio to the video, then some how get the audio to always be matched to the video when ingested to FCP. A friend wrote us a widget to match files, and to wrap them together, but we first had to trim the audio to the frame with the video and name the audio clips to match it's video. Big big headache. Then, technically you'd have to render out the file to a self contained. We didn't do this, and instead just had them reference in the QT wrapper. This however, gave us headaches in FCP as the app had to look at the video then find the matching audio that was linked in the QT wrapper. Yes some better TC tools would help, but so would recording directly into the camera.

Clearly the sound capabilities on the Red sucks big time. It's a joke really.... not to sound too hard on Red. The idea that you could argue that film has always used a second system so we shouldn't complain about Red's shortcomings is offensive. My old DVX handled sound better. Why not strive for better then how film handled sound, not the same, or even worse then a DV camera.

There are lots of uses for the Red other than big budget films that have massive sound departments. It appears that the Scarlet and other new Red offerings will have a better sound monitoring plan, so it looks like they knew this needed to be better. At least they listne to us.

But as mentioned, on important projects, regardless of where the sound is being recorded to, you will alwyans need a great location sound person.

Chris

 

[Tom Visser]

I have high hopes for the rev2 audio board. Based on my experiences with rev1, I have to agree with you, audio quality is lacking. The main problem seems to be that the line inputs lack about 10dB of headroom. The larger problem is that if you simply pad the input down -10dB and then send -20dBU tone (padded to -30dBU) to the point that should be -20dBFs on the RED, it seems that you can actually clip the RED well before 0dBU on your mixer. This then means that I have to calibrate for the clipping point, or 0dBFs on the RED, rather unusual, since one can not have confidence by calibration in the normal fashion. I have also noticed that RED cameras seem to have wildly different input sensitivities between inputs. I don't know if this is because of damage or defect in the specific cameras that I have worked with, or something more widespread... I would never use the preamps on a RED, either rev1 or rev2 board, but I wouldn't use the preamps on any other camera that I have worked with either.

With dolly moves, jib, and crane shots, you really must face the reality that as a sound person, I simply don't want to be near the camera, in their way, or tethered for sound. As a sound person, I must invest heavily into wireless hops and the new Lectrosonics D4 system looks quite interesting. (a little bit over $3k for a set, I'd probably have to shell out closer to $5k for an extra receiver, as I'm mostly seeing A & B camera crews on most of my work) Although I'd prefer on board audio used mostly for dailies, I'm also fine going with camera audio if used hard wired or with a very good wireless hop, sending a mix to the camera from my mixer, while at the same time, recording iso'd tracks for safety. These iso tracks would be submitted to post, but only used when necessary.

 

[Joel Kaye]

I have high hopes for the rev2 audio board.

Rev 2 is fine. With Rev 1 I used an external phantom power unit and fed that into the cameras.

I've recorded a lot of very usable stuff directly into camera using a Sennheiser 416. With a Y cable you could record to 2 tracks set at different gain levels and mix it all later. Video guys have been doing that forever. Perhaps not optimal - but you can produce a lot of good stuff that way if you need to.

 

[Tom Visser]

Rev 2 is fine. With Rev 1 I used an external phantom power unit and fed that into the cameras.

I've recorded a lot of very usable stuff directly into camera using a Sennheiser 416. With a Y cable you could record to 2 tracks set at different gain levels and mix it all later. Video guys have been doing that forever. Perhaps not optimal - but you can produce a lot of good stuff that way if you need to.

That's a nice little secret technique. Spliting a mic just once shouldn't be a big deal from an impedance standpoint. Just a word of warning, for those that wish to try this, only engage phantom power from one of the pres.

 

[Mark Weingarten]

Feature Audio: Wireless Audio Feed to the Red Camera?

Feature Audio: Wireless Audio Feed to the Red Camera?

Hi All,

I'm putting it out there to ask about feeding the Red Camera audio wirelessly.

I have tried several different receivers and have yet to achieve stellar results.

And yes I have tried the argument that they should forget about the in camera audio and post sync all of mine, but.....

Anyone having success feeding a Red Camera wirelessly?

What's working?

Which brand / model number wireless are you using?

Are there any custom cables and or mounting brackets being used?

If so what are those, and where did you get em?

Thanks, Much Mark.

 

[Mark Phelan]

Mark,

I've had good success with the following setup:

TRAM TR50 lavaliere microphones
Lectrosonics SMa transmitters
Lectrosonics UCR411A receivers

The receivers plug directly into the camera and are attached to an ET Battery Plate with velcro. The plate is mated to the V Dock system from ET, works great. If possible, I will tend to route the wireless through the SD 442 mixer for better results. I will also swap the two Lectros with a Sound Devices MM1 to use as a preamp for when I'm using a Schoeps CMC641 on a boom.

 

[Jon Lillie]

Anyone having success feeding a Red Camera wirelessly?

I have tried it on the last student film I worked on. (I'm a student myself, but the only one with any audio experience and desire to do it for the next 30-40 years.)

Which brand / model number wireless are you using?

I used a sen G2 300 series

Are there any custom cables and or mounting brackets being used?

Just needed the red audio cables, and then the 1/8" to XLR-M for the pack. Didn't mount it to anything, just set it on top of the camera, between the rails. Stayed put the entire shoot.

I know it's kind of ironic plugging in a 200 dollar wireless into a 40k camera(mat box, filters, etc), but it's what I had access to, and it seemed to work.

I set gain by running tone into my mixer, and then matched the meters in the camera to my mixer, and to my laptop (motu traveler, with boom recorder) so I had to watch only one set of meters.

I'm in post now... actually taking a break from editing the dialogue. The editor didn't have any problems editing, and so far I haven't really had to go to my ISOs.

Hoped this helped.

On a side note, if anyone is looking for an assistant or intern... I would love to get coffee or lunch.

 

[AaronPicot]

attach mic for quick scratch audio??

attach mic for quick scratch audio??

little off topic, but related,

i want to attach a mic to the camera to record some pick up stuff we are getting. audio need not be quality at all (mixed), i just want it for reference but it shouldn't be ungodly awful either. think news ENG stuff. i'll probably just pick up a cheap-ish mic from a local shop to do this. i have the mini xlr adapter cable. what else do i need and what should i know going in? it is even possible? thanks.

 

[Jon Lillie]

little off topic, but related,

i want to attach a mic to the camera to record some pick up stuff we are getting. audio need not be quality at all (mixed), i just want it for reference but it shouldn't be ungodly awful either. think news ENG stuff. i'll probably just pick up a cheap-ish mic from a local shop to do this. i have the mini xlr adapter cable. what else do i need and what should i know going in? it is even possible? thanks.

It's possible though I don't recommend it. There is to my knowledge no way to modify the pres on the camera without going into the software to do it. and the Cam Op would not enjoy that very much.

That is the caveat. Just be sure you have the Pres set right, and 48v turned on if needed, and you should do ok.

If it's just reference you should be ok...

 

[ericyoung]

It's possible though I don't recommend it. There is to my knowledge no way to modify the pres on the camera without going into the software to do it. and the Cam Op would not enjoy that very much.

That is the caveat. Just be sure you have the Pres set right, and 48v turned on if needed, and you should do ok.

If it's just reference you should be ok...

You can assign the EVF dial wheel to control one channel of audio level. Not as good as a dedicated audio level knob, but better than nothing I guess.

 

[John Neff]

I think Tom Visser's comments weigh in very well for a truly professional approach.
When I did location audio for David Lynch's "Inland Empire", he was shooting on SD Sony PD-150 cameras, and some scenes had a lot of actors, plus he likes different area mic placements in case he wants a long shot, with corresponding audio.
So, I set up a Digi 002 system with a laptop and external firewire drive, and fed it from a combination of lavs, boom and area mics, through bith the 002 preamps and a Focusrite OctoPre that has 8 channels with limiters, and ADAT out to the 002. I had an 8-channel Audio Technica wireless rig for the lavs, with the paddle antennas, and all speaking actors were lav'd. Each mic was recorded to a separate track in ProTools, then I would send a stereo mix out to the main camera via A-T wireless again, feeding the 2 XLR inputs on the camera. The second and third cameras used a shotgun or internal mic, and that audio was used just for synch purposes.
Because the audio was being sampled into the 002, made analog again for the wireless mix output, and re-sampled into the camera, there was a definite latency issue.
Each scene and take were clapped, and in post, when you line up the 002 tracks with the camera audio, you can see the distance between the clap peaks.

Drag the camera audio back to where the highest peak lines up, and you have sync.

Laborious, but the editor has camera sound to edit with, then you bring in the individual tracks for mixing, and it works rather well.

Now, with RED, you have 4 inputs, so run the wireless stereo or mono mix into the camera, pop a 416 or schoeps on the camera, and you've still got one or two inputs left.
But I find that external preamps with limiters really are the only way to go to ensure that when some goes from a whisper to a scream, you have the capability to not clip the audio.
Who wants to tell the Director you need a re-shoot because of distortion?
Not me...