Excellent...you helped me a lot. I just purchased a Zoom f3 and run it into my Canon RP which has pretty much the same audio features as the as the R5. Thank you.

There are so many videos out there suggesting to set the gain at two or three above zero, but with no rigorous testing. This is the first video I’ve seen that actually digs into the audio performance of the R5 to figure out what the optimal settings are with different features. This should be more broadly shared, it’s always been a concern for me and I’ve never had a straight answer until now. Thank you.

@kobezhan

9 ай бұрын

Agreed. I noticed that my audio clips too fast below level 16, this video just confirmed my suspicions

Thanks Jason. I find your videos very informative and well-done. I did some of my own audio testing using a different methodology but with somewhat similar results. I used an audio recorder (Sony PCM-M10) to play a fixed level computer generated tone to the EOS R5 (with the input attenuator enabled). I found that a camera record level of about 27 resulted in approximately unity gain (-12db 1kHz tone - LPCM, 48kHZ sampling @16-bits). I extracted the recorded audio from the camera's video recording and used Izotope RX 6 to estimate the signal-to-noise ratio for this setup. I measured an SNR of about 88dB which is surprisingly good since the signal undergoes both a D/A (at the recorder) and an A/D (at the camera). The audio pre-amps in the camera seem to perform well as long the record level isn't pushed too high. One advantage of this test setup is that the audio level and electrical loading should be representative of real usage. BTW, the R5 will not record LPCM (regardless of the Audio Compression setting) if the video format Is not All-I. Typically, when I shoot video, I use Canon IPB-Lite which triggers AAC encoding. The compressed audio can, of course, be decoded to LPCM with little quality loss compared to direct LPCM recording.

@PointsInFocus

11 ай бұрын

I covered the AAC and LPCM stuff in my other video on R5 audio settings (kzread.info/dash/bejne/l65-rNiElMa_lZc.html). AAC is only forced (the camera ignores the compressed audio setting) only when the camera is set to IPB-Lite. The camera will record LPCM in regular IPB mode and up (based on the audio compression custom function setting). As for the test setup. I did something similar when I tested the R5C. Using my Zoom F6 to generate a -6dBFS test tone and routing that to the camera. That said, I'm one of the first to admit that audio stuff isn't my specialty. I know enough about electronics and testing them to be able to muddle through some basic tests, but I'm sure there are either faster or more "standard" ways to go about doing this stuff. What ends up making this kind of stuff really challenging for me is that by the time I get around to doing something audio related again, it'll be far enough in the future that I'll have forgotten everything I figured out now.

@sloan.d

11 ай бұрын

@@PointsInFocus Thanks. So IPB-Lite is the culprit. 🙂 Audio is definitely its own speciality.Wikipedia was some good articles (e.g. see Line Level - en.wikipedia.org/wiki/Line_level). BTW, below are the LINE OUT specs for the PCM-M10.The recorder also has an optional limiter feature to prevent clipping. It seems to work quite well (I recorded a fire engine going by with its siren blaring as a test of the limiter). Sony PCM-M10 LINE OUT - Load impedance: 22 kΩ - Rated output level: 1 Vrms

@kobezhan

9 ай бұрын

​@@sloan.dCan you check what the ideal record level is for the R5 if the attenuator is off? Thank you!

Hi, I found this video hard to understand, not your fault the presentation and pace was excellent. The previous one was easier to understand. I have been recording a friend who is an opera and classical singer in an outdoor situation using my R5 camera In my ignorance I have used the C3 P mode in Auto sound recording and compression. I have not been entirely happy with the sound, I felt the higher ranges of her voice were 'subdued'. Would I be better recording on manual? If so, what should I set the manual figure to? Should I have compression on or off. Would I be able to test this in my house by playing a cd of similar music and trying the different settings? Sorry if this doesn't make sense, I am a stills photographer and trying to get into video recording.

@PointsInFocus

Ай бұрын

Starting with the stills photographer getting into video bit. That's my story too, and audio is one of the things that I still struggle with. Especially being confident that I'm doing things "right", which is odd as that was never something that I really worried about with making an image. With that said, the way I've come to think about audio goes like this. First, sound isn't light. You can't telephoto mic your way to better sound, you need to get the mic closer to the subject. A mediocre quality lav mic clipped to your subject's shirt is almost always going to sound better than a professional shotgun mic on camera 10 feet away. This also helps with SNR (signal to noise ratio) and frequency response. Step one is always to get the most signal with the least noise to the recorder as possible. And keep in mind noise here doesn't just mean like self noise or something like that, but the entire rest of the soundscape where you're recording. Step 2 is optimizign the recording. This is where you'll want to use manual levels, and set them as hot as you can go without clipping. Digital audio, just like digital photography clips when you reach the saturation point. And as you know, when you blow your highlights, there's no data to recover there, the same goes for audio. Generally the rule of thumb is to aim for your levels to bet around 12-18 db below 0dBFs (the clipping point). This gives you some headroom for louder transients. However, this is a rule of thumb, and you may need/want to go lower than that what you're recording has an especially wide dynamic range between the soft and loud parts. Which brings me to compression. In the case of the R5, this controls whether the camera records uncompress linear PCM audio data, or lossy-compressed AAC audio data. Personally, given how small audio is compared to video, I don't see a compelling reason to use the compressed (AAC) audio option. At least not if you're shooting anything serious with the camera. There are further micro optimizations that can be made, but honestly, if you're to the point of trying to micro-optimize the on camera audio, you're probably better off investing in a proper audio recorder and doing 2-system audio with the camera's track just being used for scratch/sync purposes.

@paulsaunders1573

Ай бұрын

@@PointsInFocus Thanks for your reply, I will have to do some experimenting.