Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Low level hum when connecting Sound devices Mix Pre D with epic...

T

tomspark

Well-known member
Joined
Apr 21, 2007
Messages
78
Reaction score
0
Points
6
Age
51
Location
Sydney, Australia
I am trying to get my in camera audio sorted for a documentary. I have a Sound Devices Mix Pre D which is feeding audio to camera from a lav or boom mic via the mix pre d outputs set to mic level output. I have researched on the forums the best cable wiring and am using the appropriate cable wiring from the 3 pin XLR to 3.5mm mini-jack as described in this thread. (http://reduser.net/forum/showthread.php?71581-These-Cables-Work-(And-why))
The details:
I have set the input levels inside the epic at 24dB to align visually on the lcd display with the tone feed from the mixer. Audio records fine and is usable but certainly not without issue. In the interests of getting the cleanest possible signal, I can notice a low level hum/digital interference like sound which almost sounds like the digital version of film passing through a gate - it has a regular undulation to it. This sound is recorded to the r3d file because I can hear it on playback in red cine pro. I only get it however once the output of the mixer is plugged into the camera. I do get this sound whether the mixer is turned on or off with the cable connected between the 2. When the mixer output is not plugged into camera there is no hum. I only hear this when monitoring directly from the camera or using the headphone output on camera as a return to the mixer. This sound is not present on the mixer when monitored directly. I can hear this when the mixer input pot levels are dialled down to zero or the mixer is off. Its not too noticeable so its difficult to test with a mic plugged in without the mixer as I cannot turn the mic down. When I do turn the channel audio input down in camera to the minimum 8db, the sound is almost not there (can just hear it when audio output is turned right up. It is definitely introduced however once the mixer and camera are connected and dissappears when they are not. If I use the unbalanced mic out from the Mix pre-D instead of the balanced L and R outputs, (mini xlr to stereo 3.5mm mini jack into one input on red epic) it is actually a little louder.

Anyone else get this? Anyone know how to get rid of it?

Thanks
Tom
 
Try removing the lens. Do you still get the sound? if you do, read this thread.

It seems there have been some instances of cross-talk between the lens data system and audio. I've never experienced this myself, and I use the Mix Pre-D often. But I also seldom use the PL mount.
 
I tried removing the lens and it has not made a difference. I am using a Red canon mount anyways, not the PL. It happens specifically when the audio cable is plugged between the Sound devices Mix pre D output and the epic input. Here is a sample of the sound exported from Red Cine Pro then boosted by 12db in FCP to hear it a bit better.http://snd.sc/HOhvbZ Its subtle but still there.
 
i heard this sound too while monitoring form the camera, However I didn't notice this sound while listening in RedCine X or if its still there is really faint.

Edit: i was using a Rode NT4 with 9volt battery phantom power, didn't use a mixer or pre amp.
 
I have been conducting these tests audio mentioned above with camera plugged into the DMSC AC power adaptor. On powering from battery the hum has been audibly reduced almost to insignificance. The slight clicking artifact is gone. Definitely better when powered by battery which means there is some sort of power contamination from the ac power source.
 
tomspark said:
I have been conducting these tests audio mentioned above with camera plugged into the DMSC AC power adaptor. On powering from battery the hum has been audibly reduced almost to insignificance. The slight clicking artifact is gone. Definitely better when powered by battery which means there is some sort of power contamination from the ac power source.
Click to expand...


I just ran this same test with the Epic Power Adapter and I got the same results (the adapter seems to be responsible for the audio interference). When running the camera off a RedVolt the audio meter is clean.
I'll check as soon as I can with the Wooden camera mount / anton bauer dionic HC.

Is Red aware of this issue ?
 
tomspark said:
I am trying to get my in camera audio sorted for a documentary. I have a Sound Devices Mix Pre D which is feeding audio to camera from a lav or boom mic via the mix pre d outputs set to mic level output. I have researched on the forums the best cable wiring and am using the appropriate cable wiring from the 3 pin XLR to 3.5mm mini-jack as described in this thread. (http://reduser.net/forum/showthread.php?71581-These-Cables-Work-(And-why))
The details:
I have set the input levels inside the epic at 24dB to align visually on the lcd display with the tone feed from the mixer. Audio records fine and is usable but certainly not without issue. In the interests of getting the cleanest possible signal, I can notice a low level hum/digital interference like sound which almost sounds like the digital version of film passing through a gate - it has a regular undulation to it. This sound is recorded to the r3d file because I can hear it on playback in red cine pro. I only get it however once the output of the mixer is plugged into the camera. I do get this sound whether the mixer is turned on or off with the cable connected between the 2. When the mixer output is not plugged into camera there is no hum. I only hear this when monitoring directly from the camera or using the headphone output on camera as a return to the mixer. This sound is not present on the mixer when monitored directly. I can hear this when the mixer input pot levels are dialled down to zero or the mixer is off. Its not too noticeable so its difficult to test with a mic plugged in without the mixer as I cannot turn the mic down. When I do turn the channel audio input down in camera to the minimum 8db, the sound is almost not there (can just hear it when audio output is turned right up. It is definitely introduced however once the mixer and camera are connected and dissappears when they are not. If I use the unbalanced mic out from the Mix pre-D instead of the balanced L and R outputs, (mini xlr to stereo 3.5mm mini jack into one input on red epic) it is actually a little louder.

Anyone else get this? Anyone know how to get rid of it?

Thanks
Tom
Click to expand...

At first I thought I had a defective camera, but both of my Epics do exactly the wame thing:
Any time I raise the input level above the minimum gain, the nose level is obvious. If I plug a mic in directly, and raise the gain about half way to get full level sound, the noise is about the same level as the sound. In other words at mid scale, the noise is near normal recording levels. I have tried it on battery, Ac, no lens, several different cable sonfigurations including open and shorted dummy plugs in the input, balanced and un balanced it's all about the same.
I went and bought an external recorder before I realized that the cam does not put out Tri-level and time code. So that didn't work too well either. (Tried separate TriLevel sync gen and Timecode gen, but since they were not syncable to each other that didn't work either.) What I finally did, using a signal generator, I took the output of the recorder and made a cable with a voltage divider in it to give me the right level with the camera set for 8db gain. This gives Decent sound: (You have to try hard to hear the noise, and the quality is good.) But, with any setting significantly above minimum gain, the sound is totally unusuable on either of my cameras. I am desperately looking forward to the Pro IO adapter, (although I don't know how I will come up with enough money to buy a pair). Since all the mics I normally use require 48V Phantom power, I have to use something external for now anyway.

From an engineering perspective, I sounds like two things are happening: 1. Digital signals are getting into the analog signal. 2. At higher gains, the system breaks into regenerative feedback between analog output and input.

When I view the waveform on the time line with inputs turned up 1/4 or more of the way, there are large pulses at regular intervals as well as a mix of all kinds of frequencies. I can't say that I have had hum though.
Playback was through Adobe PP patched with Lightroom, but the noise is the same in the headphones.

After Red replaced everyones sound cards on the Red Ones, I used to record using all 8 channels (cams synced together), high quality concerts, with no issues.

I just went and tried the camera with input cable connected but recorder turned off. At about 55 db the level meters reach the red. At 24db, the meters show nothing although you can clearly hear noises and occasional signals that go to about 1/4 scale. The noise spikes I mentioned appear to be at one second intervals.

Hope that helps.
 
Dave Neathery said:
After Red replaced everyones sound cards on the Red Ones, I used to record using all 8 channels (cams synced together), high quality concerts, with no issues.
Click to expand...
Why not just use one external audio recorder to record the 8 channels of sound without compromise?

To do so with two cameras, you need two Ambient Clockits or Denecke SB-T tri-level sync generators. Slave these boxes from the timecode coming off the audio recorder. Genlok each box to the camera, and run the cameras from external sync. Now, all three devices (camera 1, camera 2, and sound) are all running off common timecode.

This is commonly done all the time and is not a lot of trouble. The temperature-controlled crystal oscillators (TCXO) in the tri-level sync boxes are stable enough that they hold timecode within a frame for more than 8 hours. And if you're doing a concert, I don't think anything less than a decent multitrack recorder (Zaxcom, Sound Devices, Nagra, Aaton, Tascam -- there's quite a few to choose from) will do.

I agree that having a scratch track for audio syncing on the camera is a good idea, but I wouldn't rely on the camera sound as the only soundtrack for your project. And I would say that about any digital camera, not just Red.
 
Very simple reason: I didn't have anywhere near that kind of money. In order to do the upgrade to Epics I have cashed in my IRA and everything I have saved so far in life and I am retirement age. I also am too much of a soft touch, so I have had feature filmes where they never paid me a dime. (I require cash up front now.)
Also the Red Ones only needed an external Tri-level sync fed to each camera, the time code looped through from one camera to the next to the Deceche timecode slate.
 
You can always hire an experienced sound person to do this for you, which will be much cheaper than actually buying this equipment.

I'm confused about what you're saying with Tri-level sync and timecode: this has not changed between the Red One, the Red Epic, or any other digital camera on the market. With multiple cameras, you always need multiple boxes. Sure, you can loop it, but having a physical wire between two cameras is extremely inconvenient -- even impossible, assuming different camera positions, handheld cameras, etc. Renting two SB-T's is very inexpensive, like $20 a day per box.

The Denecke timecode slates (going all the way back to the TS-1) will jam to incoming timecode, so it only has to be physically connected to the timecode source once every 5-6 hours or so. No permanent connection is needed, assuming it has a Syncbox (which they all have for about 17-18 years or so).
 
For the most part, I agree with you. What you couldn't know is that I am rather experineced in audio, having started out working in a recording studio a bit more than 40 years ago. I later started a small record company where we made a point of producing the best quality records possible. We took over running the pressing plant when our records were being pressed, we brought our studio recorders to the master cutting facility and substuted them for the Ampex machines they were using because we proved to them that our machines were noticably better in speed accuracy, frequency response, S/N etc. The point is that I believe in uncompromizing quality as my patents will also show. But I also have to be pragmatic I don't have the money to do everything like I want, so I have to improvise. I agree that Tri-level sync hasn't changed. I used to bring a gen that had 6 outputs with, and feed each camera with a separate coax. I carried a wooden spool of Canare coax and made up cables as needed. But, I couldn't expect the gen to stay in sync with the slate. Many of our clips are an hour or more. I probably wasted my money on the Deneche since we have never really needed the EL backlight, and the displayed timecode is of no value either with Adobe, as far as I can tell. What did change is that the Epics do not put out Timecode. Feeding the first cam with Tri-level and then taking the timecode out of it to cam B etc, used to produce perfect sync both video and audio. Of course, the last build I had allowed using the USB Master Slave features which were really great.

Here is where I disagree with you: You say that it's virtually necessary to record the sound separately. (In the old days it was.) If you follow that logic, then we should never shoot with only one camera either, we should have a backup running. I see no reason to think that a camera can't also produce reliable, stellar, sound now days. I do feel that for some sounds 48K 24 bit is not as good as one might want. Clearly, Red is having a sound problem with the Epic at the moment, but I have to believe that as always, they will fix it. I would be delighted to have the money to do all that you suggest, but it's also hard to find good sound people. I was involved with a move a while back where they hired a sound guy who absolutely did not know what he was doing. He ran around with a shotgun mic connected to a recorder in a bag, he had no sync to the cameras of any kinid and no ref mic one the cameras either. And he charged FAR more than I did for the 2 Reds. They wound up scrapping all the production and are now trying to refinance and start over. (He wasn't the only problem, of course.) The funniest part is that he got paid and I didn't. I'm not really upset about it, but it did cost me a whole season at a time when Red Ones were still in demand. I'd rather be a nice guy and broke than mean and wealthy.

At any rate, I appreciate your having taken the time to respond. Hope you have great success!
 
Dave Neathery said:
Here is where I disagree with you: You say that it's virtually necessary to record the sound separately. (In the old days it was.) If you follow that logic, then we should never shoot with only one camera either, we should have a backup running. I see no reason to think that a camera can't also produce reliable, stellar, sound now days. I do feel that for some sounds 48K 24 bit is not as good as one might want.
Click to expand...
I don't agree. The preamp sections of almost every camera now on the market is not going to be as good as the preamp stages of pro audio recorders. More importantly, the camera operator can't be responsible for monitoring different tracks and riding levels, both of which are very necessary in filmmaking. Even documentaries these days routinely use upwards of 8 tracks all the time for multiple wireless mikes, plant mikes, and booms. You can't get along with just 1 or 2 channels nowadays for most projects.

It's just a question of maximizing quality. By the same token, I don't think the DP would like to do the work of the camera operator, assistant camera operator, the gaffer, and the grip. Sure, one person can do it, but only with a much greater workload and more compromises.

Clearly, Red is having a sound problem with the Epic at the moment, but I have to believe that as always, they will fix it. I would be delighted to have the money to do all that you suggest, but it's also hard to find good sound people.
Click to expand...
You gotta look in the right places. Try Jeff Wexler's site -- there's mixers from all over the world there, and good sound need not cost you a lot of money.
 
I agree that the sound system in all the cameras I know of is poorer than what is available externally. But I still say that in an ideal world, the camera should have quality equal to what is other wise available expecially when the cost is so low, and have at least 8 tracks as well. Even SDI supported 8 tracks when it was just SD. I am not excluding the sound guys with mixers and operating the boom poles, it just makes more sense, in my opinion, to have the actual recording all together. Then you don't have the problems of sync if someone forgets to slate, or if a sync signal goes missing.

When I said we used to use all 8 tracks on the 2 Red Ones, they were fed as line level from a mix.

Just one less thing to worry about and several less things to go wrong. I don't think for one moment the cam operator should have to ride gains, I agree: he has plenty to worry about.

On a related point, now that the Pro I/O module is out and I have seen the price, I can no longer plan to buy even one, so I will have to budget for a separate 8 channel recorder, especially since that is cheaper ($5000 vrs $7500 for the modules). I did buy one of the new Ti GPS chipsets and found some timecode generator chips, so I am considering making some GPS based boxes to put on cameras and recorders etc to assure sync. Time to do it is a problem though. For now, we're just having to make certain to slate at start and end of each cut and adjust the length of the sound in audition to match the reference track from the camera. If we get projects with bigger budgets, we will be able to deliver the quality we would long to always produce. I do have 2 top nitch sound guys available, I will have to supply the equipment though.

We have strayed from the point of this blog which was to answer how to use the Epic's sound system, I think.

At any rate, given the current audio situation with the Epic, I agree that it is necessary to record separately for anything beyond the simplest situations, and even in the simplest situations external premaps are absolutely necessary.
 
Dave Neathery said:
At any rate, given the current audio situation with the Epic, I agree that it is necessary to record separately for anything beyond the simplest situations, and even in the simplest situations external premaps are absolutely necessary.
Click to expand...
Exactly.
thumbsup.gif


One big issue I could see with trying to do 8 simultaneous channels spread across four cameras or even two cameras: there could be phase issues where one camera is slightly out of sync with the other, unless you're feeding sync reference and have a common clock for the digital audio signals. Without that, all bets are off.

A separate digital audio recorder would definitely be the best place for all 8 tracks, since in one box, they're all going to be in sync, relative to each other. And assuming a decent mixer, mikes, and preamps, there won't be any inherent compromises.

Timecode is not cheap or easy. If you have access to the temperature-controlled crystal oscillators used by Ambient or Denecke, that's half the battle right there, since I think the relative sync error in 12 hours is some infinitesimal number expressed as a part per million, so it's under 1 frame per 12 hours. That's pretty stable. If the box heats up, gets slammed around, or loses power, then there could be issues.
 
My comment was intended to mean that all 8 (or however many are used), channels and the video should all be in one source namely the camera and that the camera could and should record state of the art audio quality. When we used to record on 2 separate cameras, I carefully tested the audio and video sync over a period of at least 1 hour to be sure it was truly locked, with the equipment we used.

One of my very first jobs was designing high stability oscillators for military customers. I also founded and owned a company a few years ago named Pico Time Inc. that made time standards based on WWWVB (before they upped their power), and on GPS sources. They had to remain stable even durning loss of signal. It's a lot cheaper and easier to build stable oscillators now than it was back then. (Like almost everything electronic.) Still, it's enough effort that I probably won't find time to do it. Enjoyed the chat. As we both agree, at this point in time anyway, there are no other practical options with Epics except to record externally.
 
I'm reviving this thread. Please don't tell me to record separate sound. I know I can, I just don't always want to.

Up to now I've worked with rented Epics with either no need for recording sound on camera, or separate sound with no guide track, so not needed to really look at audio on an Epic.

However, having now taken delivery of my own Epic, I really want sound on my camera to be as good as the little camcorders from Sony and Panasonic I've owned or used in the past, and it's not. It's not even as good as the Red One after the earliest audio boards were replaced with redesigned boards.

I have EXACT SAME low level audio interference sound as the above SoundCloud recording from tomspark http://snd.sc/HOhvbZ.

However, in Unbalanced Mode with the Gain In turned up, mine is audible powering the camera from AC power with NOTHING else plugged in. So it's not acoustic mic pickup, and there's no possibility of a ground loop.
There is also a different, very obvious, electrical interference hum from the camera fan, which changes pitch with fan speed - this purely electrical pickup within the camera (no mic attached).
While the fan hum remains, the other hum is not there with Battery power, so it's obviously something to do with the DSMC AC Mains Adapter.
Both hums are recorded on the R3D files too.

An older rental Epic-M with the original (non phantom power) audio board design did not exhibit this hum when powered from AC Mains.

In Balanced mode, the fan hum and AC Adapter hum is predictably much reduced into the noise floor, but the white noise level is nothing to be proud of. Balanced +48V is even noisier. The noise floor is clearly audible on recorded files, when the Gain In is turned up to get decent levels from a Sennheiser 416 plugged direct into either input via Red's own adapter cable.

Yes I could feed the camera a higher level mic feed from an external mixer so I can keep camera Gain In down, but that's another gadget I have to carry and power, and needing mounting.

Anyone found a solution? My Epic is currently at Red Europe, but they have not been able to recreate the AC Adapter interference in the workshop.

See also this thread with the same/similar issue.
http://www.reduser.net/forum/showth...o-meters-even-when-there-is-no-mic-plugged-in
 
Hi Eric,

No answers from me but I too am a fan of having audio recorded in my Camera. I hate syncing (especially when I like looking at all my takes and not simply selecting the one that worked...mostly cause I prefer grabbing different lines from different takes and get the best acting choices).

Broadcasters/ENG News folks use on-camera mic inputs all the time. How many newsreporters run around with a sound guy? Yes, sure there are but the point is not to have or not to have a sound guy, the point is the technology to have good sound. We're not debating the sound guy's job (which is invaluable) or the benefits of a second system. At least I'm not.

I believe video (which use to be the broadcast world) and filmmaking are merging. It will affect both methodology of shooting. It would be nice for the camera which offers the sound capability to, well, offer the "SOUND" capability. We paid for it. They advertise it. How many more years before this get's fixed? Seriously, who likes syncing audio regardless of how easy it is...
 
Wil Wong said:
No answers from me but I too am a fan of having audio recorded in my Camera. I hate syncing (especially when I like looking at all my takes and not simply selecting the one that worked...
Click to expand...
I have synced up thousands and thousands and thousands of takes in my life, just over the last 20 years. Literally, in a nonlinear timeline, it takes about 20 or 30 seconds per take. Assuming 100 takes per day, that's an hour. It's not a big deal, especially if it's done by an assistant or an intern making $10/hour.

Using a separate sound recorder becomes even more important when you have a dozen or more iso channels coming from individual actors, which can potentially save you a lot of time and money in post by helping to avoid ADR.

In documentary or news situations, I can see where recording audio in the camera is a necessary evil. But I think it's still going to ultimately limit the sound quality. For that, I'd say hire a sound mixer to worry about sound, and have them send a wireless feed to camera so at least the camera operator doesn't have to worry about levels.
 
Marc Wielage said:
I have synced up thousands and thousands and thousands of takes in my life, just over the last 20 years. Literally, in a nonlinear timeline, it takes about 20 or 30 seconds per take. Assuming 100 takes per day, that's an hour. It's not a big deal, especially if it's done by an assistant or an intern making $10/hour.

Using a separate sound recorder becomes even more important when you have a dozen or more iso channels coming from individual actors, which can potentially save you a lot of time and money in post by helping to avoid ADR.

In documentary or news situations, I can see where recording audio in the camera is a necessary evil. But I think it's still going to ultimately limit the sound quality. For that, I'd say hire a sound mixer to worry about sound, and have them send a wireless feed to camera so at least the camera operator doesn't have to worry about levels.
Click to expand...

All very true Marc, and I agree, but doesn't address what we are discussing, ie. the quality of Epic audio.
 
You must log in or register to reply here.
Back
Top