low microphone level BMPCC4k

It is well known, that mic sensitivity is 20db ... 30db too low.
Wow, not 3db, but 30db!
Why does BM not fix this?

Does this improve using the XLR input?

What do you mean ?
What did you measure ?

Please see here:

viewtopic.php?f=2&t=90884

I tested microphone sensitivity with two samples of BMPCC4k with internal mics and Sennheiser MKE 400.
Camera gain set to 100%.

You have to boost audio by 30dB in Resolve.

Again my question: will this improve using the XLR input?

Thanks

The XLR inputs don’t have much gain either. Likely they are not meant to be plugged directly into a mic but the output from a preamp / mixer. Camera pre amps with sufficient ( clean) gain to work with XLR pro mics are not cheap so on a budget camera like the pocket this was a necessary cost saving exercise.

The XLR inputs don’t have much gain either. Likely they are not meant to be plugged directly into a mic but the output from a preamp / mixer. Camera pre amps with sufficient ( clean) gain to work with XLR pro mics are not cheap so on a budget camera like the pocket this was a necessary cost saving exercise.

Thanks, however, I do have some more "budget" cameras like Canon XC10 and Nikon D500. They are doing fine, even most smart phones do. Regarding audio, the BMPCC4k performance is simply not acceptable.

Regarding audio, the BMPCC4k performance is simply not acceptable.

Then your options are straightforward. Use a separate preamp, either stand-alone or incorporated into a microphone. Use a separate audio recorder and mike. Or return the camera.

The XLR inputs don’t have much gain either. Likely they are not meant to be plugged directly into a mic but the output from a preamp / mixer. Camera pre amps with sufficient ( clean) gain to work with XLR pro mics are not cheap so on a budget camera like the pocket this was a necessary cost saving exercise.

Thanks, however, I do have some more "budget" cameras like Canon XC10 and Nikon D500. They are doing fine, even most smart phones do. Regarding audio, the BMPCC4k performance is simply not acceptable.

Most cheaper cameras have add on options for XLR inputs with separate gain circuits as the sound quality with the internal camera pre amps is so poor / not good enough for serious use even if they have enough gain.
No one serious about sound quality would plug a mic straight into a BMPCC4k / 6k and expect high quality sound as they are input sockets and not microphone sockets.

even most smart phones do [fine].

Yes and no. Apple makes sound recording with its phones idiot proof. There is a built-in compressor/limiter that prevents clipping. However, it also reduces dynamic range and it can't be turned off.

I discovered this when I used both an iPhone 11 and a Sound Devices MixPre 6ii to record the sound of a NY subway train. The phone recorded the sound with no clipping whatever. However, it also squished the sound into a fairly narrow range of quiet to loud. The MixPre recorded what I actually heard, from low bass rumble to high pitched whines, and gave me the control I wanted over the recording.

This is not a criticism of Apple. I think that what they are doing makes perfect sense for a significant majority of users. However, I sure wouldn't use an iPhone to record a source that has a lot of range between quiet and loud.

I don't know whether some of the budget cameras you mention also have a built-in compressor/limiter, but it's entirely possible.

It's so long ago I don't remember details, but this issue was explored exhaustively when the BMPCC 4K first came out (search for it here). Robert N. performed tests, complete with plotted graphs. As I recall, the 3.5mm has substantially lower gain than the XLR. There was a firmware update early on which improved the 3.5mm slightly, but not by enough.

Unlike others here, I'm not convinced it had to be this way, but that's the situation. For 3.5mm you need either a preamp or a mic with built-in gain boost. For ambient sound, the onboard mics aren't bad (and get much higher levels).

The XLR inputs don’t have much gain either. Likely they are not meant to be plugged directly into a mic but the output from a preamp / mixer. Camera pre amps with sufficient ( clean) gain to work with XLR pro mics are not cheap so on a budget camera like the pocket this was a necessary cost saving exercise.

Thanks, however, I do have some more "budget" cameras like Canon XC10 and Nikon D500. They are doing fine, even most smart phones do. Regarding audio, the BMPCC4k performance is simply not acceptable.

I think you're mixing lots of different things here. Most "budget" cameras and smart phones record 16bit audio and BMPCC4K records 24bit audio so you don't really need to record as "hot" as possible. For example you can find "How much headroom should I leave with 24-bit recording?" article on "Sound On Sound" for more details.

Also in the video there's a comparison between MKE 400 which is a mono shotgun mic and internal microphones of BMPCC4K which are stereo and probably have cardioid polar pattern. Moreover MKE 400 has a high pass filter and audio from the internal mics is unprocessed - you can see it with a spectrum analyzer at around 1:40 in the video. And also I believe most smart phones do some kind of "post processing" steps for audio before recording to make it sounds "good". So all those comparisons are not really fair.

I believe the idea about audio in BMPCC4K is that it always should be post-processed in some way. So to have a somewhat fair comparison it would be good to have the same microphones, the same post-processing steps (like high pass filtering, EQ, compression, etc.) and then normalize audios by loudness. Because otherwise more loud audio (which is not always about the volume level) is always sounds "better".

I posted the tests John Paines mentioned here:
viewtopic.php?p=450455#p450455

My personal solution for nice audio on the PCC4k:
Rode Video Mic NTG at full gain for atmo sound and Sennheiser EW100 G3 receiver at +12dB out for interview situations. Those recordings have passed national broadcast standards and been used for TV commercials all the time.

No one serious about sound quality would plug a mic straight into a BMPCC4k / 6k and expect high quality sound as they are input sockets and not microphone sockets.

Thanks, you are right, they are input sockets and not microphone sockets.
This is why BM writes in the manual:

"3.5mm Microphone Input
Connect a microphone to your camera’s 3.5mm stereo connector. Mic and line level
audio are supported. The microphone level audio is lower than the line level, so if you
are connecting a microphone to the camera and have line level selected, you will find the levels are too low......

Mini XLR Microphone Input
Plug in external balanced analog audio via the mini XLR connector. A standard XLR
microphone can be connected to your camera using an XLR to mini XLR adapter cable.
The mini XLR input provides phantom power for connecting professional microphones
that aren’t internally powered."

I'm so sorry to connect a microphone to these input sockets.

Sorry I for my mistake. So either it's operator error or your microphones are faulty as it clearly states in the manual that the sockets are designed for direct connection with microphones.

Haven’t had an issue when using my old Sennheiser MK416 with phantom power supplied by the UM4.6K but I stopped doing that generally and now plug mics into the Sound Devices MixPre-6 II which improves the sound quality.

Sorry I for my mistake. So either it's operator error or your microphones are faulty as it clearly states in the manual that the sockets are designed for direct connection with microphones.

I don't get it either. How does an "input socket" which conforms to the usual microphone formats differ from a microphone socket and how do you tell the difference? It goes without saying, the discussion above concerns the preamp...

Because "specs" are not that well defined AFAICT.

Just reading https://www.shure.com/en-MEA/performanc ... -mic-level for instance says :

- "Mic level" is anywhere between -60 and -40 dBu .
- "Line level" can be either -10 dBV for consumer equipment or +4 dBu for professional equipment.

No idea what range the BMPCC4k accepts / expects.


Does anyone have the pinout for the mini-xlr input ? I seem to remember the mapping is not 1:1 with the full size XLR.

Haven’t had an issue when using my old Sennheiser MK416 with phantom power supplied by the UM4.6K but I stopped doing that generally and now plug mics into the Sound Devices MixPre-6 II which improves the sound quality.

The UM4.6k is a style of camera where it would be expected to be able to record high quality sound direct from mic's and will therefore have a better audio section built in. For a start it has physical knobs to control the audio levels which mean they are likley real analogue gain controls whereas on the pocket there is likley just a fixed analogue gain and the rest is a digital gain stage.

So I was curious so I quickly connected my BMPCC4k to a signal generator, fed some 1 kHz tone at fixed levels into the 3.5mm jacks and observed the result.

What I came up with is :

* "3.5mm Line": Full scale = 18 dBu (so a 18 dBu signal would read "0 dBFS on the bargraph", in theory because obviously it would be clipping at that point)

So if we go with the line level is - 10 dBV ( ~ -7.5 dBu ), a standard consumer line level would be at -25 dBFS, definitely a bit quiet I think.


* "3.5mm Mic":
- Each 10% of gain is equal to 5 dB
- At 0% gain Full scale = 28 dBu
- So basically the "Mic" setting at 20% gain is the same as the "line" setting ...
- At maximum gain ( 50 dB ), the full scale is -22 dBu. On the shure website I linked above, they gie -60 and -40 dBu as typical mic level which would give us here between -38 dBFS and -18 dBFS.

So if you have a "hot" mic that outputs -40 dBu, at max gain you're right about the right level. But if your mic is lower than that, you'll definitely have to boost it in post.

I've attached the raw measurements I made as a text file in case anyone wants to review / retest or draw other conclusions.

I got an mini XLR connector today to repeat the testing with the XLR input.

* XLR gain is also 5 dB every 10% with the exception of the "0% gain" which is ~4-5 dB lower than you'd expect with that curve. My guess is that at 0% some preamp is completely bypassed.

* XLR Full Scale is 18~19 dBu at "the limit of 0% gain" (see above, so according to the expected linear gain curve but with preamp active), so about 10 dB more sensitive than 3.5mm input.

* XLR line is the same as XLR mic with gain=0%


So to answer the OP original question : yes, I would expect using the XLR input to help with your audio levels. It's not going to get you 30 dB but more like ~ 10 dB.

Many thanks to smunaut for your calculations. This is very helpful.

1.
BMPCC4k Mic (3.5mm) sensitivity at 100% gain for 0dB is -22dBu or 61mV
My Sennheiser MKE400 is specified with booster (!): 20mV / Pa (pascal)
1Pa = 94 dB
Normal voice at 1m distance is about 60 dB or 0.02 Pa or 0.4 mV out of the Sennheiser.
With respect to 61mV for 0dB this will result in -44 dB (!!) audio level in the camera.
To push the camera level up to 0dB at full gain you need a chainsaw in 1m distance !!!

2.
I made a relative measurement with a 1kHz sinus signal.
The BMPCC4k is 25dB less sensitive than a Canon XC10.

3.
I will try with an mini XLR adapter.

4.
This is simply a bug and not a feature.
I will discuss this with BM support.