Jump to content

Epic Crosstalk


Rachel Cameron

Recommended Posts

Hey folks,

Passing through a public building, I saw a camera operator with a Red Epic (no talent there yet). I had to stop to pester him. It turns out that he was having audio problems, so I was glad I did.

The issue (seemed to be) crosstalk from channel 2 to channel 1 (during peaks from the receiver, but only when the G3 was really peaking hard, as in the Epic meters at full red overload), so I quickly looked over his gear patch. Dunno what the Epic gain levels were set at, but the metering seemed fine on channel 2 unless a hard SPL peak was introduced. 

Otherwise, (barring a listen on the cans) I saw several issues. 

1). He was using a Red, which seems to be about the most audio "unfriendly" camera discussed here.

2). No mixer of any kind (also no soundie: the underlying issue).

3). No limiters in the signal path (didn't check the Epic's 'soft limiters' settings)

4). Overloading (poorly designed?) camera audio (the Epic manual calls them high quality) pre-amps.

5). Dividing a balanced line level with a split XLR cable. I never do this, and I've always wondered about the efficacy of splitting any XLR signal, mic or line level.

There were no crosstalk problems as long as the receiver was not in full overload, so I'm guessing this was the issue ~ just too much signal for the Epic's pre-amps to handle.

(Marc, you're right about the Red forums. Appalling.)

 

 

Link to comment
Share on other sites

The question is: why did the receiver overload? Any analog circuit will crosstalk at some point and to a certain degree. 

I'm betting he didn't have the pilot tone engaged or squelch set properly and was getting heavy rf hits. there also a limiter in the G3, i think

Edited by Constantin
Link to comment
Share on other sites

 

Well, my question is more like: Why did the Epic crosstalk? Apologies for the confusion.

EG: to demonstrate the crosstalk problem on the Epic, the camera op had to overload the G3 quite a bit (he should've set the limiter in the G3 to avoid this, and probably didn't), but I didn't expect to see such a hot signal crossing from Epic Ch 2, over to Epic Ch 1. He had to sort of 'thump' the lav head to illustrate the Epic's crosstalk, and yes, when he did, it took a full red meter on the Epic Ch 2, to show a fair amount of crosstalk on Epic Ch 1.

But in essence, Constantin, you answered my question. 

Any analog circuit will crosstalk at some point and to a certain degree. 

Copious thanks!

Link to comment
Share on other sites

+1 to Simon's response. I've had problems feeding RED cams a scratch track from a G3/G2 regarding levels before. Meters looked sort of OK, but I realized my output level from the G3 was screaming hot (my operator error). I turned down the receiver output, set input to mic, and things were better. note: don't remember if this was an Epic or other model. Fortunately this was in prep and not on the job. Maybe a limiter in the red's preamp circuit was engagedso the meters looked OK even though it was clipping big time? REDs scare me.

Link to comment
Share on other sites

Yep, you gotta pad the Red down by at least 20-25dB because of the Mic level input. It's possible to see levels that technically are hitting the right meter levels, but they're still overloading the input stage. I generally took the level on a Lectro SRb down by at least -25dB to help the camera assistant knock tone down to -20dB without distortion. If the G3s have a Mic Level output mode, that would solve the problem.

Note that the Epic input is also an unusual 3.5mm TRS balanced connector, so it's kind of goofy and non-standard.

Link to comment
Share on other sites

Thanks so much for the information, everyone. I took his number, so I'll get in touch and relay the suggestions. It'll also give me a chance to put my mixer on it, and goof with the Epic settings myself, some. I'm dying to see what it was, but with a 411/SMV...and then his G3. Acknowledging the crosstalk issue, I'm now itching to know if the output of the G3 was really that mismatched with the Epic's first gain stage. I guess I was looking at a giant impedance mismatch (with heinous crosstalk), that is, if the Epic is looking for a balanced mic level signal, and the G3 sends a balanced line.

Link to comment
Share on other sites

I'm pretty certain that this is a problem with the Red's audio circuitry. I've seen this happen many times with Red Epics, Scarlets and Dragons. I recently AC'd a feature on two Epic Dragons and a Scarlet and the mixer was using R1a's as hops and those quite regularly made the signal crosstalk from channel 1 to 2 on the Red so we had to be very careful about our levels to the camera at all stages. As it was stated, keeping the level low is key. 

Link to comment
Share on other sites

I've fed the Epic and Scarlet mic level with an unbalanced TS plug (and a TRS w/ the ring/shield tied together) from a G2/3 quite a few times, the Red set to unbalanced   It sounded ..well.. as good as it can. (but only meant for sync anyway). Furthermore,  the 3.5 to 3.5mm connector cable that comes bungled with the G2/3 EK100 portable receivers won't work well at all, the recorder end is a dual mono TRS with the tip and ring terminals tied together.

 ( I'm not aware of an actual limiter in any of the G2/3 systems,  certainly not an adjustable one or even on/off switchable.)

Link to comment
Share on other sites

Thanks so much for the information, everyone. I took his number, so I'll get in touch and relay the suggestions. It'll also give me a chance to put my mixer on it, and goof with the Epic settings myself, some. I'm dying to see what it was, but with a 411/SMV...and then his G3. Acknowledging the crosstalk issue, I'm now itching to know if the output of the G3 was really that mismatched with the Epic's first gain stage. I guess I was looking at a giant impedance mismatch (with heinous crosstalk), that is, if the Epic is looking for a balanced mic level signal, and the G3 sends a balanced line.

Please note that impedance matching, balanced circuits, and correct gain staging are three entirely different things. 

-- Impedance matching has not been an issue with the types of circuits we're talking about for 40+ years -- ever since the advent of solid state circuitry when bridging inputs became the norm. 

-- Balanced circuits are for higher CMMR (Common Mode Rejection Ratio), with each leg having the same potential to ground, so that extraneous noise induced equally on both legs of the balanced line are nulled. 

-- Proper gain staging is necessary to ensure that not only each device but each stage within each device is operating at the ideal area between max signal handling (distortion) and minimum signal handling (noise) while maintaining sufficient headroom for peak signals.

 

Link to comment
Share on other sites

Thanks so much John, for taking the time to hammer these points out. That's great information. All of this is very difficult voodoo for someone with no worthy mathematics or electronics background. Words, definitions, and simple wiring schema, though?...I can handle.

I'm pretty sure I have a grip on balanced/unbalanced, regarding the +/- polarity conductors, and phase canceling of noise, balanced long cable runs vs. short unbalanced runs, for noise immunity, etc.

I'm pretty sure I have a grip on proper gain staging (through my experience with mixing consoles). 

I've obviously confused line/mic matching (voltages, and a step up/step down transformer) circuitry, with impedance (resistance to voltage) matching circuitry, EG: bridging. Thanks for that. Big difference.

In my original post, (of course, [cam op] had no cans available to hear what Simon, Chris and Marc all arrived at), the cam op said it 'sounded fine' which I would have to judge for myself, until he (purposefully, for demonstration purposes) overloaded the input (the Epic displayed a full, solid red meter all the way across, before the Epic would bleed over some, to the other channel), so I deduced..crosstalk issues, as there was sooo much signal. I always thought some of the better constructed analog audio circuits were immune to crosstalk, though. Wrong again.

Of course, I couldn't tell whether the G3 was sending a line or a mic level, or what the Epic was expecting. It probably sounded 'sort of okay', just as Chris alluded. Thanks everyone. I'll RTFM on that one. 

And I guess I need to carry around a pair of cans. :mellow:

Link to comment
Share on other sites

Here's a reprint of a post I made a few months ago that might be a bit useful in understanding some of this:

"Forgive the following geek rambling, but maybe it'll help some emerge from the darkness of misleading hearsay and assumptions that are often bandied about.

This topic has to do with voltage transfer vs power transfer and bridging circuits vs impedance matching circuits.

In our application, a bridging circuit typically means that an input impedance is ten or more times higher than the output impedance feeding it. Below the ten times ratio you could be loading down the output, causing some reduced performance.

The output impedance of a professional mixer is typically in the range of fifty ohms or so. The Input impedance of most modern gear (meaning the vast majority of solid state equipment manufactured in the last half century) is often in the range of five to ten thousand ohms, or more. That is obviously well over a factor of ten times.

In the audio world, Impedance matching was critical back in the 1950s, 1960s, and before, with telephone systems, and also with many common tube circuits that deployed impedance matching transformer designs that made power transfer important.

For the last fifty to sixty years, about the only place impedance matching has been important for audio circuits is with a power amplifier to speaker connection, as, in that application, power transfer is still important.

However, most of the typical audio circuits we deal with today are voltage circuits rather than power circuits. So, if your output voltage at a given time is two volts, and you feed five recorders bridging the same output, they each see the same two volts (bridging circuits are connected in parallel to an output).  If each of these recorders has an input impedance of ten thousand ohms, then with all five bridged across the mixer's output, they present a combined load of two thousand ohms, which is still way more than ten times the roughly fifty ohm output of the mixer.

The main caveat when feeding several devices from one output, is if your output is balanced, each of the devices you're feeding needs also to be balanced. Unbalance one of the connections and you've unbalanced them all. Also, if the devices are located at some distance, the combined capacitance of the cables comes into consideration. For cable runs not in the hundreds of feet category, cable capacitance is seldom an issue if the circuit is fed from a low impedance output."

Link to comment
Share on other sites

Hey folks,

Passing through a public building, I saw a camera operator with a Red Epic (no talent there yet). I had to stop to pester him. It turns out that he was having audio problems, so I was glad I did.

The issue (seemed to be) crosstalk from channel 2 to channel 1 (during peaks from the receiver, but only when the G3 was really peaking hard, as in the Epic meters at full red overload), so I quickly looked over his gear patch. Dunno what the Epic gain levels were set at, but the metering seemed fine on channel 2 unless a hard SPL peak was introduced. 

Otherwise, (barring a listen on the cans) I saw several issues. 

1). He was using a Red, which seems to be about the most audio "unfriendly" camera discussed here.

2). No mixer of any kind (also no soundie: the underlying issue).

3). No limiters in the signal path (didn't check the Epic's 'soft limiters' settings)

4). Overloading (poorly designed?) camera audio (the Epic manual calls them high quality) pre-amps.

5). Dividing a balanced line level with a split XLR cable. I never do this, and I've always wondered about the efficacy of splitting any XLR signal, mic or line level.

There were no crosstalk problems as long as the receiver was not in full overload, so I'm guessing this was the issue ~ just too much signal for the Epic's pre-amps to handle.

(Marc, you're right about the Red forums. Appalling.)

 

 

Hi Rachel,

With the possible exception of #4 and #2, the items listed above are not the problem.

1) When all things are correct (which can be done with the G3 system), the Red Epic will record audio cleanly.

2) A mixer and someone to operate it could prevent the Epic's audio channels from overloading, but not necessarily required to prevent overload.

3) Actually, there were limiters in the signal path, if only the limiters on the G3 transmitters. In an unmonitored uncontrolled scenario (no sound person), the gain structure can be set so that the transmitter limiter prevents everything in the record chain from going above maximum. Of course, it probably takes a skilled sound person to set the gain structure in such a way, but it is not the fault of the equipment.

4) Overloaded analog input stages generally increase cross talk, often to an easily noticeable level. My guess is that this was the problem.

5) Splitting a line output with a Y cable will not degrade the audio. In fact, the standard low impedance outputs and hi impedance of inputs became the standard, in part, for this very application. It's true that consideration should be given with regards to adapting a balanced to an unbalanced circuit (or vise versa), and while this could cause output distortion, it would not increase the cross talk of the input device.

Sticking with simple basics would probably solve the problem:

First, a cable is needed so that a tone can be sent to the transmitter from an external source. Turn recorder limiters off. Send a tone into the G3 transmitter until the limiter kicks in (the point that an increase in input causes little or no increase in output), then reduce the tone by about 15dB. This tone is now your 0VU reference level. Set the receiver output level to a range suitable for the recorder (in this case the Red Epic), and adjust the recorder's input so that the tone meters 20dB below maximum (sometimes its a bit of a guess with low rez camera meters, but with the reference tone set 15dB before limiting, you have about 5dB to play with). Turn the limiter on in the recorder at this time if desired, but there should be a specific reason to have them on. The gain structure is now pretty well aligned so that there will not be input overload, which was probably causing the crosstalk. All that has to be done now is set the transmitter input sensitivity for the actor's speaking volume (occasional limiter action, usually indicated by a red flicker), then press the red button.

gt

 

Link to comment
Share on other sites

Thanks to you both, for those explanations. Hugely helpful.

John: It took me a few days of mulling, and about five reads of what you wrote (studying schematics, and then reading it again)..but I think I got it. Impedance matching (these days) would refer to matching impedances across passive speakers/powered amps..to the impedances of 2, 4, and 8 ohm loads (and that's all), in order to not overwork a power amplifier to a thermal shutdown or just a plain ol' meltdown, such as line arrays and live concert gear, which has little to do with what we do. I previously thought a D.I. box was doing an impedance match. Wrong again. The line/mic voltage issue is handled with the bridged circuitry you described, which seems to be just inside the input jack on a board.

Glen: That's what I feared. Since I wasn't there when he set it up, it seems cam op didn't follow proper gain staging of the unfettered signal, before setting limiters. From his responses to my questions, he probably wasn't even aware that the G3 had limiters. It's nice to see the method confirmed so cleanly. That's what I normally do with my old 442 and 744T..or when I'm wiring to a camera's recorder: Move my way through the signal path making sure no part of the path is overloaded or underloaded? underfed? undermodulated?....ain't enough, before moving to the next gain stage. So it's highly likely that he didn't structure the signal path properly, and was paying that price. And thanks for clearing up the split XLR question. I still don't feel quite comfy splitting audio signals by soldering a Y-cable. It's creepy voodoo to me. I never know what to expect, not knowing all about voltages and electronics, like some of the brains here do. 

Constantin: Your comment about crosstalk and analog circuits was an eye opener for me.

In my OP, I called the Red cameras about 'the most audio unfriendly camera' discussed in this forum. Wrong again. Sorry, RED. : (  

DSLR's generally wear that badge these days. But what a great excuse to go double system. 

Incidentally, years back (OT alert), I got to send audio to what the camera owner called 'the very first RED ONE in the field'. They called it (him?) Elvis. It was RED serial #0006, because, as the story goes, the owner of RED kept the first five RED ONE's for himself. Did he? Maybe someone here can confirm that tale for me. The camera owner was from NYC, and just upstairs (or downstairs?) from RED, in the same edifice. That's my beat up old breakaway cable sullying the frame. 

Gracious thanks, everyone.

Rachel

DSCN1524.thumb.JPG.70558d381f4eeb6b87c80

 

Link to comment
Share on other sites

Rachel,

You don't really have it all correct, and some of your summations have me scratching my head and saying, "what," so be careful not to assume you have it all figured out.  Many of us have lived with, and dealt with, these things for years, so you have nothing to feel bad about, and your quest for knowledge will serve you well.

One point to get some of it aimed in the right direction:  impedance matching is typically about efficient power transfer rather than the protection that you implied.  Power transfer is not usually an issue with modern circuitry that typically operates at much lower impedances than its former counterparts did -- unless it's an output stage that needs power to drive something such as a speaker, where proper matching offers maximum power transfer.  A DI box is a different case that typically does impedance matching because the output impedance of a guitar pickup is sometimes quite high (can be over a million ohms), so it needs to prevent the loading effect of going into a much lower impedance input, often a mic input that may be as low as one thousand ohms.

Keep in mind that we're just brushing the surface here with these explanations.

 

Link to comment
Share on other sites

Ahh..okay. So a D.I. box actually does do an impedance match? I'm glad I had that part right. At those tiny voltages, I guess mismatches still would cause load up, even though there's no real thermal issues. It's still doing an impedance match, as with the (hi-z) instrument pick up to a (low-z) input? So..okay, it's more about efficiency in the power transfer, not really protection from a meltdown, even though with those large P.A. systems, any inefficiency (or impedance mismatching) happens to cause thermal issues as a side effect of the poor efficiency? I guess I'm seeking the common denominator that impedance has through all these examples....which would be said efficient transfer of power?

And I realize that with words..this discussion could only go so far. Scratching the surface is an understatement, I know. No way could I ever have this all figured out.

I'll never have a strong grip on mathematics and its formulas. Though I stare and stare at them (same with wiring diagrams), I guess the best I could ever seek is a basic working, practical knowledge here, which is *very far* from having electronics figured out. I'm sure this is like quicksand, as the more I writhe in the math, the more I'll sink. It's probably hell on a math mind to try to get this into words, when the formulas say it all so succinctly. I know that's why they exist. And thanks for such patience and kindness. I feel badly that I dragged you into this. They're my own limitations.

I so wish I didn't have such anxiety with math. I got as far as basic algebra/trig. Electronics seems to go way beyond that. I followed my brainy older brother (very adept at math) through school, and really frustrated the hell out of both him, and my math teachers. They all thought I'd be like he was. However, my literature, stage, and English teachers were okay with me.

In the end..I have my ears..which is what I use when my feet are on the ground (and great agility).  That, and some good meters.

Link to comment
Share on other sites

Rachel,

Nothing whatsoever to feel bad about.

As for the math involved in electronics, it's often much simpler than you're allowing yourself to believe.  Two simple formulas will take you further than many have ventured.

Look up "Ohm's Law" -- only three parts to learn there.  Then learn the primary correlary -- and you only have to learn one new element for that.

I=E/R

&

P=IxE

(I=Current, E=Voltage, R=Resistance, and P=Power)

I'll bet Wikipedia will be your best buddy in the process.  UPDATE:  I just looked at the Wikipedia entry for Ohm's Law and now don't recommend it as a starting point.  You want to find a resource that begins with the simplest applications of the formula and builds knowledge from there.

One step at a time... Have fun...It'll be less daunting than you ever imagined -- now that it actually means something in connection with your work, the pieces will fall right into place.

 

 

Link to comment
Share on other sites

Rachel,

Nothing whatsoever to feel bad about.

As for the math involved in electronics, it's often much simpler than you're allowing yourself to believe.  Two simple formulas will take you further than many have ventured.

Look up "Ohm's Law" -- only three parts to learn there.  Then learn the primary correlary -- and you only have to learn one new element for that.

I=E/R

&

P=IxE

(I=Current, E=Voltage, R=Resistance, and P=Power)

I'll bet Wikipedia will be your best buddy in the process.  UPDATE:  I just looked at the Wikipedia entry for Ohm's Law and now don't recommend it as a starting point.  You want to find a resource that begins with the simplest applications of the formula and builds knowledge from there.

One step at a time... Have fun...It'll be less daunting than you ever imagined -- now that it actually means something in connection with your work, the pieces will fall right into place.

 

 

PM sent.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...