SURVEY - Epic-X (DSMC) Audio Quality

[ericyoung]

Having recently taken delivery of my personal Epic-X with current phantom power audio board and Firmware 3.3.15, after initial testing I sent it straight back to Red Europe, as I wasn't happy with the sound quality. It's still under investigation.


PLEASE no comments on the lines of recording separate sound. Obviously the preferred option, but this is about using direct audio on an Epic-X, possibly via an A-Box or similar passive breakout, but not via an external pre-amp/mixer.
Would also prefer that this thread does not degenerate into a bitchfest. Keep it to concise objective findings, so we can see how widespread and common the problems are.

Besides taking the poll (specifically in BALANCED +48V Phantom Power Mode direct plugged mic, with no external PreAmp/Mixer or post noise reduction) - Most helpful would be short comments agreeing or disagreeing with my findings.
If disagreeing, please specify whether your camera is an Epic-X or Epic-M, and whether your audio board supports phantom power, or is the older version that does not, and firmware version.


My personal findings are ( especially with high audio gain in settings ):
- Unbalanced audio has AC power adapter interference when powered from mains, and camera fan interference (electrical, NOT acoustic) is audible powered from mains or battery.
- Balanced audio gets rid of the AC power adapter interference, but the camera fan is still audible at a lower level on Channel 2, but not audible on Channel 1.
- Balanced audio with 48V phantom power is noisier (hissier) than balanced without phantom power.


I initially thought that turning up the audio gain as much as +45dB or maxed out +57dB was unrealistically high. However, plugging in an industry standard Sennheiser MKH416P microphone directly (via Red's XLR adapter cable), +45dB is a perfectly reasonable level to record my normal speaking voice about 1.5 feet away from my mouth without distortion/clipping, and a hushed voice/stage whisper in a drama could easily require the maximum +57dB. I now believe using the higher Audio Gains is not an overly stringent test of the Epic's audio input circuits, especially as Red chose not to include a Line Level option, or to provide a more robust connector than the 3.5mm sockets in the Epic. Besides it should not require feeding higher levels via a separate audio mixer to keep the camera gain levels down. Sadly the Red One with the upgraded board seems a better camera in all audio aspects, and that is hardly the most highly regarded by sound recordists!


I did most of my tests with maxed Audio Gain of +57dB to highlight the interference and noise issues, but even at the lower level of +45dB they are still audible in the recordings, albeit turning the headphones right up to hear it clearly when not listening to an actual microphone recording. Of course actual audio will mask it, but it definitely still audible.

Also had the Fan set high (70% or 100%) - again to highlight the Fan electrical interference.
At 70%, the interference is a lower pitch, but similar level. At lower typical recording Fan speeds, the interference will be at an even lower pitch and arguably less audible, but it will still be there.


Note I didn't have any mics plugged in while testing, so any interference is not acoustic. It's either direct through the camera's circuits or by electrical induction. Turning up gain should only increase hiss (random noise) from a good input mic amplifier circuit.


One extra note:
Looks like the mains audio interference when Audio Inputs are in Unbalanced Mode is at least partly related to my particular DSMC AC Mains Adapter. Using a different one, the "motorboating" and pulsing/ticking sounds are much less, although can still hear electrical interference. Unfortunately, Red Europe have not been able to reproduce this particular finding.
My mains "motorboating" and pulsing/ticking sounds exactly like this SoundCloud file recorded by another Reduser tomspark http://snd.sc/HOhvbZ.


While it's unlikely that I would ever use the Audio Inputs in Unbalanced Mode in a real life recording situation, that mode does reveal how the audio input circuits are poorly isolated from electrical interference and mains noise in Epic, and shows where the interference that still exists even in Balanced Mode is coming from.


Balanced Mode is better, and usable in a pinch, but still not completely clean.
Balanced +48V Phantom Power Mode is noisier/shashier than Balanced Mode.


While I expect some hopefully low amount of white noise with gain all the way up, any hum/non white noise in a balanced audio circuit should not be acceptable, whether battery or mains powered, or the specification is rather poor.


I also compared a Panavision Epic-M with a Panavision Red AC Adapter - while the audio board was the old Epic-M version without phantom power, it's noteworthy that my Epic-X is significantly worse. Ie. the updated board is not an improvement quality wise.


Related threads:
http://www.reduser.net/forum/showthread.php?98779-Scaret-X-amp-Epic-X-audio-with-A-Box
http://www.reduser.net/forum/showthread.php?83691-Yet-another-audio-problem
http://www.reduser.net/forum/showth...-connecting-Sound-devices-Mix-Pre-D-with-epic
http://www.reduser.net/forum/showth...o-meters-even-when-there-is-no-mic-plugged-in

 

[Bob Gundu]

The gain on the Scarlet and presumably on the Epic too, will introduce way too much noise. But going through the the Sound Devices 302, and having zero gain on the camera resulted in excelent sound. I compared the internal RED recored audio to an external recorder and found essentially no difference. So, muy conclusions are that you can definitely get nice audio recorded directly into the RED cameras by not using any gain on the camera.

 

[Mike P.]

Also seem to recall the lens mount causing noise issues back in the earlier days of DSMC... (Gaffer) Taping over the pins/contacts was the work-around at the time. Clearly not a solution, but it'd be worth testing if you're not using the data off your mount... It'd isolate the problem at least.

 

[ericyoung]

Also seem to recall the lens mount causing noise issues back in the earlier days of DSMC... (Gaffer) Taping over the pins/contacts was the work-around at the time. Clearly not a solution, but it'd be worth testing if you're not using the data off your mount... It'd isolate the problem at least.

Thanks Mike - no lens attached.

Bare camera with factory fitted Side SSD Module and Titanium PL Mount with port cap only, to exclude as many variables as possible.

Nothing plugged into mic inputs for most of the tests, except when I wanted to check what actual usable Audio Gain settings would be for a 416 plugged directly into the camera.

 

[ericyoung]

Bump

 

[Peter Stelzhammer]

[x] voted

I'd say "ok-ish" (option number 3, the "middle-way")...
For what we do (doco/live event/image film) it's good enough - for Interviews I always go with 2 mics at least so I have not only a backup but also the choice to pick the cleaner one in post. For "critical" stuff (where clean voice is needed); dedicated sound-guy with proper external recorder -> everything else comes close to suicide (ok, that was a bit overstated now, but you get the idea).

I'm not a sound guy tho -> but I've heard too much sound guys complain - and by that I mean sound guys actually using RED and not only like the ones judging things by just seeing and not trying them...

 

[Wil Wong]

I've used Red's built-in audio for some scenes without any problems. I admit that it might be a hit and miss but I haven't had any major problems. Then again, I haven't done enough testing to hit any. I'd love to pump audio into it for my next feature film and see how well it holds up. I'm sure I'll use it as reference but if it really holds up, I may skip the entire sync process and just go red.

Coming from a broadcast background, I guess I'm used to synced audio on camera Betacam, DV, etc. Now that broadcast and film have begun merging, I want the best of both worlds...ha, ha.

 

[Ulf Krentz]

Unfortunately I just found out something else to report here. There is an interference once the wifi Redmote connection is established. It sounds similar to a cell phone interference when a call is coming in but it is there constantly during the wireless connection. I experience this on Scarlet and EPIC once anything is connected to the audio input, seems like the audio ground is picking it up. It is sensitive to the gain setting (get less loud when gain is lower) so it is probably a good idea to keep the input gain level as low as possible and use an external preamp as already suggested. I will try different cable setups and hope I´ll get rid of the noise, meanwhile I´d like to advice anybody who likes to use the Redmote with wifi connection to thoroughly check the audio prior to shoot.
@Eric, please let me know how it works out for you, I second the same issues here with the power-supply and the fan interference on the EPIC, no probs with the Scarlet.

Cheers, Ulf