Mixing Audio from SDI and External XLRs?
 

Is there a device that does this: connects to a camera's HD-SDI; has at least 2 balanced XLR inputs; and records the same timecode as the camera so the extra 2 XLR inputs have identical timecode and sync up in a NLE with no problem.

I am using an EX1 and will need to record up to 4 wireless mic's and I need control over each channel when editing (Premiere Pro).

Thanks

Why SDI? Certainly there are high-end equipment like Sound Devices, et.al. that record multiple sound channels plus TC.

a camera's hd-sdi would be an output. yes the device is an audio embedder, and could add up to eight channels of audio to the hd-sdi stream, which might already contain the input audio channels from the camera. then what? i don't see where you expect to go from there...an external vtr? or are you trying to divine the TC from the hd-sdi for an audio only recorder?

Why HD-SDI? For an EX1, isn't that the only source for timecode and/or to trigger recording?

I am basing my assumptions off of the nanoFlash video recorder which uses HD-SDI and can be triggered to record thru HD-SDI once it notices the camera has begun recording. The nanoFlash can be setup to use identical timecode as the camera's internal recording.

I need to be able take my video and additional audio tracks and lay them on the timeline without any fuss. If there is no such device, then my next option would be a 4 channel mixer and output to the EX1's 2 inputs. But, like I said, I prefer to have full control over each channel while editing.

The timecode enbedded in the video stream is Vertical Interval Timecode (VITC) while that used for audio recording is Linear Timecode (LTC). Different critters. When you have a camera with TC in/out and Genlock in, you're jamming the codes between the generator in the camera and that in the recorder so both are reading the same, not converting one into the other per se. But be that as it may, current file-based audio recording systems DO NOT have their recording or playback speeds controlled by timecode in the first place - the code only serves to establish a single position reference point for file lineup and doesn't provide any speed reference for maintaining sync of the duration of a take. In fact it doesn't do anything more or better than an old fashioned clapper slate does, except it does it in a little more automated and marginally more convenient manner.

A quality 4-channel or better field recorder such as the Sound Devices 744T or 788T models jamming their code to a smartslate photographed at the start of each take will do exactly what you need. Drop the video into the NLE, find the slate frame, slip the timeline so it reads the same TC value as displayed on the slate, and drop in the audio file - no need for the video VITC at all, really. The 788 can lock its sample clock to incoming video to provide a speed reference so there's no drift over longer takes but neither of them reads the timecode embedded in a video signal. If you want, you could get something like the Horita VLT50 VITC-to-LTC converter to read the code in the camera's stream and send it to the audio recorder but that really doesn't buy you a lot and you still should slate the takes anyway.

Steve, thank you for that easy-to-understand explanation. However, those Sound Devices units cost too much.

So, anyone with experience recording 4 talking heads via wireless lav's with a camera that only has 2 mono XLR inputs?

Everything will be recorded in our studio so what about using cheaper (ie <$1000) studio mixers and taking those 4 inputs and outputting into the 2 on the cam? http://www.bhphotovideo.com/c/product/650182-REG/Mackie_ONYX_1620I_Onyx_1620i_16_Channel.html
Another idea I had was to record the extra channels directly into our PC via M-Audio Profire 610 but a few people said there would be sync problems.

Timecode is timecode. Linear and Vertical-interval are merely two different methods of encoding/transmitting it. Back in the analog world vertical-interval was a common method of encoding timecode into the video track, especially where you didn't have any accompanying audio tracks. And, of course linear was the primary method of recording timecode for audio tracks.

That said, however, timecode is not mandatory to maintain sync. If there are only four talking heads, one solution could be the old-fashioned method of mixing it live. Another could be simply recording on a multi-track recorder without timecode, and recording a reference mixdown on the camera sound track. It really is almost trivial to slip a separate audio recording into sync on any modern editing application.

Having done it myself, I have not found any significant sync problems from recording on a separate recorder (such as a computer). If you have any doubts about it, just do a test recording and see for yourself.

How do I record a reference mixdown on the camera sound track?

Thanks

so again...are you recording audio+video on the nano flash? or trying to record on the camera vtr?

if you are recording audio+video on the nano flash--take your hd-sdi camera output and loop it through one of these on the way to the nanoflash input...

http://www.bhphotovideo.com/c/product/620833-REG/Blackmagic_Design_CONVMCAUDS_Mini_Converter_Audio_to.html
this will take either four balanced line level inputs, or four aes pairs (8ch), and embed them into your video stream. you will still need some kind of preamp/mixer to get the mics to line level. the embedder output will have your cam video + 4 audio channels and will be recorded on the nano flash-and then you can import the one thing into your NLE.

personally, i would get a mixer with 4 direct outs (or 4 AES for EIGHT channels!) to feed the embedder inputs for the nanoflash recording, and then mix your mics together and record the mixed audio via the camera audio inputs for reference on the EX1 recording. you can even set the embedder to use channels 5-8 for it's own inputs, reserving 1-4 from the camera and get your reference mix in both places (assuming the cam embeds it's own audio-but most do). lots of options...

Greg, sorry for the confusion, but I am not using a nanoflash. I have been looking into getting one and through my research into it, I learned about the various methods the nanoflash records, including using identical timecode as the cam's internal recording. However, thanks for the info about that BM mini converter - if I get the nanoflash, I will definitely get the converter also.

One option I am considering is the Aja Kona but recording uncompressed is not an option; cineform has too many problems and capturing to DNxHD requires Avid's $10k+ hardware.

So, what mixer do you recommend? I linked to a Mackie 16/2 firewire mixer which is within our budget, and I have used Mackie's control surfaces before when I was using Propellerhead's Reason. If I get the Mackie mixer, would I need a M-Audio ProFire 610 (or similar device)? The Mackie Onyx 1620i appears to do everything I need: 1) connects to the PC; 2) 4+ XLR inputs going to 2 XLR outputs to the cam; 3) balanced TRS output to Mackie monitors.

Actually, the Onyx 1220i should work also (4 XLR & 4 line inputs) or the 820i with 3 XLR & 5 line inputs (but how do I take the wireless receiver output to line, from either a Sony UWP-V1 or Seinn G3?)

It could be as simple as letting the on-camera microphone (which typicaly isn't good for much of anything) pick up the ambient "mix" in the room. Since you won't be using it for anything but a sync reference, it doesn't have to be very good. Of course we don't have any idea how far away the camera will be or what the ambient acoustics are, so you will have to make your own judgement call on-site.

Or you could take the four microphone signals, feed them to some sort of multi-track recorder, and do a static (dumb) mixdown of the signals to feed to the camera(s) as the reference ("scratch") track.

the mackie 1620i in your link will work fine to mix your mics down to a 2ch mix, connect the analog L/R out to your camera inputs for the reference mix. you could also record individual channels through the firewire port to a computer to remix later. i don't have sonys, but the G3 can be adjusted to either mic or line level output, just match the level to whatever input type you choose, and use a xlr>1/4" trs adapter for the line inputs on the smaller mixer. the profire would seem to be redundant, as it's functions are built into the onyx series.

if you record audio separately, try to get a clapboard/slate at the beginning of each take-and get a slap at the END of the clip also. this gives you two sync points in post. line up the slap/clap at the head, then scroll down to the end and stretch or squeeze your audio to line up with the clap at the end-viola! your audio is now in sync for the duration of that clip.

Greg, the Sony wireless (UWP-V1) units have a switch on the bodypack for mic & line for input. Is this the same as the G3?
So, I can take the male XLR from the receiver and connect it to a female XLR to TRS cable, right?

Steve, Richard & Greg: I really, really appreciate your help!

in the menus on the G3 receiver you will find an item called "AF OUT". set this to the lowest number for mic level output, the highest (or next to highest for me) is good for line level output. XLR>TRS is the correct adapter to get the receiver output into the line input on the onyx mixer.

don't forget to adjust the "SENSIT" (input sensitivity) on the G3 transmitter. with the mic in place you want your normal mic level to hit at least 4 or 5 segments on the transmitter audio meter or the mic will be low and noisy (hissy), but NOT loud enough to light the red peak LED (distorted)

FWIW, i'd stick to at least the 1220i mixer, and skip the 820i due to it's lack of faders-much easier to ride levels on faders, even the short 60mm ones on the mackies.

the Sony wireless (UWP-V1) units have a switch on the bodypack for mic & line for input...
Is this the same as the G3?
So, I can take the male XLR from the receiver and connect it to a female XLR to TRS cable, right?

There is no mic/line switch on the Sennheiser bodypack transmitter. An input cable can be made for -10dB using the ring connector for 'hot' instead of the 'tip' on a 3.5mm TRS input plug.
On the camera mount receiver I use -18db AF output setting which gives a strong mic level w/o distorting any mic level input I've come across.
(This is with the transmitter's sensitivity input setting on -10 or -20, with a Tram for instance)