Recording set up

[breezybellows]

I'm as lazy as they come when it comes to recording.

All of us would agree that recording using the smart phone's microphone is a big compromise in sound quality. It's convenient to shoot trim and upload using one device.

I've been slowly doing my toes into improving my audio set up.

This was the set up I've used for the last year or so (when I feel like I want better sound).

Zoom H6 as a audio interface.
Harmonic AC 5001 HQ internal microphones on my accordion.
Accordion to Zoom H6 using instrument cable.
Connect Zoom H6 to Google pixel 5 using USB C cable.
Phone's stock camera detects zoom h6. Record audio and video together.
Copy video file to MacBook (using Android file transfer).
Adjust eq and add some reverb using logic pro.
Copy audio file back to phone.
Add edited audio to video using video editing app (power director)
Upload to YouTube.

That is such a pain. Why does it have to be so hard!

I broke my audio interface!
I dropped it and then it stopped working.
I wanted to buy a new one. Before that I wanted to read a bit more about audio set up. So I asked chatGPT some questions and I watched some YouTube videos.
Some members from this forum have suggested that I need to record the audio directly into the zoom recorder.

These were the changes I made to my sound set up.


1) New Audio Interface Zoom H6Essential
I read about 32 bit float recording format. Its supposed to avoid clipping (when I set the gain too high on the pre amp). I also read that it's not really necessary if you do the gain set correctly. Also the signal to noise ratio on the Harmonik mics were not high enough for 24 bit encoding to be a bottleneck. It looked like 32 bit float can help in some cases and its good to have when you set the gain incorrectly. So I went ahead and bought it.
2) Record as wav file directly to zoom recording device
I recorded the video on my phone and the audio was recorded onto the zoom H6Essential
3) DAW on phone
My pixel 5 died after 5 years. I bough a new pixel 9 pro fold. The bigger screen is convenient enough to do audio editing on the phone. I installed the lighter version of cubase. Now I don't have to copy the files onto the macbook. I copy the audio from the zoom recorder to my phone.

The results:
1) I think they sound significantly better. The sound seems very clear. Not sure how much of this can be attributed to the 32 bit float recording. I'm guessing most of the improvement is because of recording as a wav file directly onto the zoom recorder.
2) Cubase has some interesting presets for EQ and reverb. The process is pretty quick.



Note: I have some AKG C-414 mics that will probably sound better. But I'll need to have a quiet room in order to use that.

 

[jozz]

fine result
maybe up the gain a little so we don't need to touch our volume knobs...

 

[JerryPH]

fine result
maybe up the gain a little so we don't need to touch our volume knobs...

That's actually very astute.

The video was seen as being 70% of what youtube would permit (we want it to be as close to 100% without going over at any point before youtube crushes it). It could be raised a good 14db and be very close to optimal. That said it was CAPTURED at an excellent level so as to not to ever come anywhere near 0 while recording.

When recording I try to peak at around -20db, edit the music and once ready, outputted at a level of -14 LUFS for YouTube.

 

[breezybellows]

That's actually very astute.

The video was seen as being 70% of what youtube would permit (we want it to be as close to 100% without going over at any point before youtube crushes it). It could be raised a good 14db and be very close to optimal. That said it was CAPTURED at an excellent level so as to not to ever come anywhere near 0 while recording.

When recording I try to peak at around -20db, edit the music and once ready, outputted at a level of -14 LUFS for YouTube.

What is that tool?

 

[JerryPH]

What is that tool?

YouTube! Right click any video (yours or someone else's, select "stats for nerds".

Reader's digest version:
- when recording, we capture at around -20db to never come close to 0db because going over 0db is "bad".
- when in post production, we do what we want to the sound (PAN, EQ, reverb, compression, etc...), and then export/render the file to as close to -14db LUFS without going over.
- Youtube has a compression option, we DON'T want to use that (disable in YouTube settings)
- when you upload, anything OVER-14db LUFS gets mangled by YouTube compression during uploading, you cannot stop this
- when you upload, anything UNDER -14db LUFS gets left alone, even if it is too low

So, why doe sYouTube do this? So that their commercials will be at least 14db louder than anything anyone places on YouTube. That's a trick every TV, radio and streaming platform that I know of, does. Annoying as hell.

Television is the worst. I've measured 30db volume increases and that is very annoying. I actually have a compressor/limiter tied in to my sound output from my system that then goes in to my stereo. Since I watch TV for pleasure, EVERYTHING, commercials and all shows come in at the exact same volume... you would not believe how great a difference that makes in your day to day life.

 

[dak]

I read about 32 bit float recording format. Its supposed to avoid clipping (when I set the gain too high on the pre amp).

Nope. It doesn't help against analog clipping. It only helps against clipping in the digital pathways, like when you add effects or use a digital mixer. If you clip because of too high analog gain, recording in float format won't help. The advantages in digital processing are two-fold:

1. If you have slightly too large signals, the end product of converting to a fixed-size format will be clipped with "saturation arithmetic". When the same happens in fixed point arithmetic, you get wraparound instead which is much more audible.
2. If intermediate results exceed the number precision, end results may still be reasonably good.

I'd not assign too much importance here. My expectation would be that 24bit is enough for pretty all purposes, at least when I am recording individual tracks rather than a finished mix.

 

[breezybellows]

Nope. It doesn't help against analog clipping. It only helps against clipping in the digital pathways, like when you add effects or use a digital mixer. If you clip because of too high analog gain, recording in float format won't help. The advantages in digital processing are two-fold:

1. If you have slightly too large signals, the end product of converting to a fixed-size format will be clipped with "saturation arithmetic". When the same happens in fixed point arithmetic, you get wraparound instead which is much more audible.
2. If intermediate results exceed the number precision, end results may still be reasonably good.

I'd not assign too much importance here. My expectation would be that 24bit is enough for pretty all purposes, at least when I am recording individual tracks rather than a finished mix.

I don't know enough about this to understand all of what you said. But I understood enough to know that the usage of the stock camera app was the bottleneck, even If I used the signal from the audio interface as the input (instead of phone's mic)

 

[dak]

I don't know enough about this to understand all of what you said. But I understood enough to know that the usage of the stock camera app was the bottleneck, even If I used the signal from the audio interface as the input (instead of phone's mic)

There can be two reasons for that: one is that the H6 (in the way you used it here either by your own choice or by the options it has available in audio interface mode or by settings or givens of the app recordind mode) just wasn't providing the phone with the full-quality input in audio interface mode. The other is that the format in which the app is recording (again, either by informed or uninformed choice or without choice) does not record in full-quality mode.

Namely, the problem you are experiencing can be introduced before or after the cable connecting the H6 to your phone, and it may be either fixable, or unavoidable without changing the recording workflow and setup. And sometimes one can theoretically fix it, but in practice the options are so obscure or unreliable or awkward that a workaround may be the smarter choice.

 

[breezybellows]

There can be two reasons for that: one is that the H6 (in the way you used it here either by your own choice or by the options it has available in audio interface mode or by settings or givens of the app recordind mode) just wasn't providing the phone with the full-quality input in audio interface mode. The other is that the format in which the app is recording (again, either by informed or uninformed choice or without choice) does not record in full-quality mode.

Namely, the problem you are experiencing can be introduced before or after the cable connecting the H6 to your phone, and it may be either fixable, or unavoidable without changing the recording workflow and setup. And sometimes one can theoretically fix it, but in practice the options are so obscure or unreliable or awkward that a workaround may be the smarter choice.

For now I'm satisfied that it sounds better. Will be cool if there's a smartphone with pro audio options.

 

[JerryPH]

For now I'm satisfied that it sounds better. Will be cool if there's a smartphone with pro audio options.

There have been for a while, it's just that phones have crappy digital/analog converters and tiny crappy mics.

My phone has a "pro video" section that permits better control of the video and the capture of a stereo signal...

Within the "pro video" section I can choose my audio source like all 4 external (tiny) mics, front only, rear only, a USB source (the best!), BlueTooth (I can use my earbuds or BT headset to send vocals) or a mix of BT and other sources.

The phone has fair video capabilities (honestly, not so great, it has a low dynamic range, newer phones are going to be a bit better), but the internal mics are (to be kind), crappy. I did experimentations on getting the BT to receive audio from a 3rd party BT transmitter, that failed saying I need some other 3td party app to capture the audio.

Joseph, your KORG *should* be able to send it's audio to the phone, but I don't know if the BT latency would be a factor or not. Perhaps a cool future experiment?

I don't think we will ever see a phone with prosumer, much less pro level mics and AD/DA converters, hence where external mics and audio interfaces come in.

 

[JerryPH]

I don't know enough about this to understand all of what you said. But I understood enough to know that the usage of the stock camera app was the bottleneck, even If I used the signal from the audio interface as the input (instead of phone's mic)

32-bit floating recordings are a little different. The premise is that you can bring down signals that were captured above 0db and recover 100% of what would be normally distorted. So the answer is yes it can be done... as long as:
1 - you didn't overload the microphone's capacity
2 - you have a software that understands how to work with 32-bit floating files

Under those 2 circumstances, you can record anywhere from some ridiculously low level (-75 db to +75db) and it is cool to see how it works. Normally if you record something low and then raise it in post, it raises the noise with it. Not in 32-bit recordings, you raise everything but the noise. The reverse is true too... you can blast away at ridiculous levels and bring it back down to under 0 without distortion.



The tech is interesting, but it was designed more for voice than music. Recording near 0 and someone starts yelling, you can recover the sound, but tests have shown a small loss in quality. So while this is all well and good, you can save yourself the money and complexity by properly gain staging using 24-bit (96 to 192 kHz) recordings.

Pro studios do NOT use 32-bit recording, its more a tool for someone that needs to capture very quiet sections and save the audio from explosive sound spikes.

 

[dak]

There have been for a while, it's just that phones have crappy digital/analog converters and tiny crappy mics.

Neither should matter at all when using the Zoom H6 as an audio interface (connecting it via USB). However, the expectation of crappy hardware may influence the default (or available) processing options of typical recording apps: if it is hard or impossible to keep them from "improving" the audio or from working the H6 at its best settings, that can make those apps be of comparatively little use. But at least then there is the possibility of other apps providing better options since the limitations will not be the involved hardware.

 

[dak]

Pro studios do NOT use 32-bit recording, its more a tool for someone that needs to capture very quiet sections and save the audio from explosive sound spikes.

32bit make no sense as a recording track format: the best A/D converters put out something like 20 bits. It only makes sense as a mastering/processing format in order to avoid artifacts from sloppy digital gain staging (essentially, it renders gain staging in the digital domain irrelevant).