GH3 Quick Audio Noise-Floor Test

T

TheDingo

Veteran
...I finally got around to testing the RMS noise-floor of my new GH3 this afternoon.

I tried all the different recording formats and they all have the same noise-floor of -66 dB RMS, which is about 3 dB BETTER than my FlowMotion patched GH2.

I expect the audio sound quality will be better with the non-compressed LPCM format, over the old GH2 AC-3 192 Kbits AVCHD format. ( I won't have a chance to test this until next week )

As far as setting the proper audio level, the new GH3 audio circuitry acts a bit weird when the AGC kicks in at higher level settings. Using a level setting of 10 ( the mid-point ) the recorded audio wasn't much louder, but the noise-floor shoots up quite a bit ( +20 dB ), which is NOT what you want. Doh!

My recommendation for now is to record audio with the GH3 level set to 1, but make sure you are feeding a very strong audio signal to your camera.

I used a Sound Devices MixPre-D, and first tried recording the -10 dB tape-out signal with the GH3 audio recording level set to 1. Recording the MixPre-D's 1kHz test-tone, the recorded audio level shows as -7 dB RMS.

Switching to the MixPre-D's LINE level output, fed through a BeachTek DXA-2s passive audio adapter, I was able to adjust the signal level until the first RED indicator on the GH3 shuts off. Checking this recording the audio level shows as -10 dB RMS.

So far from what I can tell, it looks like the ideal audio signal for the GH3 is a -15 dB signal, with the GH3 recording level set to 1. ( and use the MOV format which records audio in LPCM format )

All in all, I'm quite happy with the audio the GH3 records. The GH2 was already pretty good, and the GH3 is just a little bit better.
 
Can you disable the AGC on the GH3? I'm not really all that familiar with the audio specs on the camera.
 
akivisuals said:
Can you disable the AGC on the GH3?
Click to expand...

Not that I know of, but setting a very low audio recording level on the GH3 ( i.e. set it to level 1 ) pretty much takes the AGC circuit out of the picture. With the GH2 ( and perhaps the GH3 ) the AGC circuit acts more like a limiter, so your audio won't clip at higher settings, but the the noise-floor becomes horrible.

I am going to try to see if I can correspond the actual noise-floor with the GH3 audio recording levels, as I am curious to know if I can use any gain at all and still maintain an acceptable noise-floor. ( for me, acceptable starts at -60 dB, which most people can't hear unless their playback system is really cranked )

...I also want to mention that it's great to finally monitor audio through a headphone jack on the GH3, so I will likely do a lot more in-camera audio recording from now on.
 
Thanks for the test. They really need to add the option of turning off the AGC.

I've checked out reviews of the Juiced Link and that should be another good option of feeding a hot signal to a GH3 that has audio set at 1.
 
TheDingo said:
Not that I know of, but setting a very low audio recording level on the GH3 ( i.e. set it to level 1 ) pretty much takes the AGC circuit out of the picture. With the GH2 ( and perhaps the GH3 ) the AGC circuit acts more like a limiter, so your audio won't clip at higher settings, but the the noise-floor becomes horrible.

I am going to try to see if I can correspond the actual noise-floor with the GH3 audio recording levels, as I am curious to know if I can use any gain at all and still maintain an acceptable noise-floor. ( for me, acceptable starts at -60 dB, which most people can't hear unless their playback system is really cranked )

...I also want to mention that it's great to finally monitor audio through a headphone jack on the GH3, so I will likely do a lot more in-camera audio recording from now on.
Click to expand...

Great to know. I have a JuicedLink DT454 that I'll probably use when recording audio direct. For reference audio I'll probably just set my Rode Videomic Pro to +20 dB and keep the camera audio low.

I like that Panasonic prefers a limiter to AGC. AGC sucks on the Canon.
 
Just I tried a new test...

I recorded the 1 kHz test tone from my Sound Devices MixPre-D using the -10 dB tape-out port for 10 seconds, then shut off the test tone and continued to record for a full 2 minutes.

I tried recording with the GH3 audio recording level set to 1, 10, and 19. For all three recordings, the test tone recorded at -7 dB RMS.

So setting the GH3 recording level to 1 or 19, made no difference to how the test tone recorded, it always showed as -7 dB RMS.

What was different was what happened after I shut off the test tone:

With the GH3 recording level set to 1, when I shut off the test-tone the noise-floor rose from -66 dB to -63. It took about 5 seconds to rise to this level, and then stayed at -63 for the rest of the recording.

With the GH3 recording level set to 10, when I shut off the test-tone the noise-floor rose from -66 dB to -54. It took about 5 seconds to rise to this level, and then stayed at -54 for the rest of the recording.

With the GH3 recording level set to 19, when I shut off the test-tone the noise-floor rose from -66 dB to -48. It took about 12 seconds to rise to this level, and then stayed at -48 for the rest of the recording.

...So it looks like the AGC circuit is still active when set to level 1, but it adds only 3 dB of gain after 5 seconds of silence.

Higher GH3 recording levels just add more gain when things get quiet. ( I don't see any benefit to this )


The optimum setting for GH3 audio recording still appears to be:

- Set the GH3 recording level to 1

- Feed the camera a -15 dB audio signal

- Record in MOV format which uses LPCM encoding for audio

- Monitor with the GH3 headphone port
 
akivisuals said:
I have a JuicedLink DT454 that I'll probably use when recording audio direct.
Click to expand...

JuicedLink units can add up to 30 dB of high quality gain, so they should be a great match with the GH3.

MIC LEVEL at -50 dB ---> JuicedLink adds +30 dB of gain ---> GH3 records -20 dB signal from JuicedLink box


akivisuals said:
For reference audio I'll probably just set my Rode Videomic Pro to +20 dB and keep the camera audio low
Click to expand...

Cool. This also sounds like a good match for the GH3.
 
TheDingo said:
when I shut off the test-tone the noise-floor rose
Click to expand...

Man, it's disappointing that Panny would move away from their normal Limiter to a full blown AGC.

Unless...I wonder if the test tone was too loud, forcing the limiter into action when it was on, and thus being turned off when the signal stopped.

My math isn't the best. You're reporting here in RMS, but most of us just use dBFS for audio levels. So with the test tone active, what's the read in dBFS?
 
Jim Simon said:
Unless...I wonder if the test tone was too loud, forcing the limiter into action when it was on, and thus being turned off when the signal stopped.
Click to expand...

Yesterday, I tried lowering the signal strength and it did not seem to make much difference. This behavior is similar to my GH2 camera, where it was hard to overload the camera, and best results are always obtained with a strong signal and the camera's recording level set as low as possible. It would be nice if the AGC could be shut off completely, but it's effect is minimal when the GH3 recording level is set to 1.

Jim Simon said:
You're reporting here in RMS, but most of us just use dBFS for audio levels. So with the test tone active, what's the read in dBFS?
Click to expand...

I am using the Peak VU meters in Sony Sound Forge/Vegas Pro to measure audio levels which are calibrated to the RMS scale. Not sure how to convert this to dBFS.
 
Sorry to necro this post..

I am looking for some info on which shotgun mic would be best to connect to the GH3's internal recorder. I have narrowed down my choices to the Rode Videomic Pro and the Shure VP83. I heard the Shure sounds better but my concern is that it's supposedly more sensitive, and I don't wan't to have too much incidental noise from cars and what not in my recording.

Anyone try both with the GH3 recording level at 1 & shotgun gain at +20?
 
dgwow123 said:
Sorry to necro this post..

I am looking for some info on which shotgun mic would be best to connect to the GH3's internal recorder. I have narrowed down my choices to the Rode Videomic Pro and the Shure VP83. I heard the Shure sounds better but my concern is that it's supposedly more sensitive, and I don't wan't to have too much incidental noise from cars and what not in my recording.

Anyone try both with the GH3 recording level at 1 & shotgun gain at +20?
Click to expand...

It's really important to recognize that total system sensitivity includes preamp gain and make up gain applied in post. So if you wind up at the same dialog volume level with either mic, you'll have the same ambient noise from either mic, assuming their pickup patterns are similar.

Looking at the polar pattern diagrams

http://wpc.660d.edgecastcdn.net/80660D/downloads/videomicpro_datasheet.pdf

http://cdn.shure.com/specification_sheet/upload/167/VP83_Spec_Sheet.pdf

the Shure appears to have about 3 dB better rejection at 90 degrees off-axis. Both show about -10 dB at 180 degrees. So it appears that the Shure should deliver less ambient noise for the same final signal level.

Fran
 
You must log in or register to reply here.
Back
Top