Rode mics

[newzhack]

I found this entertaining, unfortunately. 

 

 

 

Edited September 6, 2023 by newzhack
wrong name

[Ty Ford]

It's not Rhode, it's Rode.

[newzhack]

1 hour ago, Ty Ford said:

It's not Rhode, it's Rode.

thanks Ty

 

[Allen Rowand]

Did someone from Rode hurt this guy when he was a child?

[Michael Scrip]

1 hour ago, Allen Rowand said:

Did someone from Rode hurt this guy when he was a child?

Seems like it.  🤣

 

I stumbled onto his channel a week ago and he's made a few videos on this topic.  

 

It's funny that we've had recorders with 32-bit float for a few years... yet he's the only guy shouting "It's a scam!"

 

In one of his videos his thesis was basically "They say you can't clip with 32-bit float... but I made it clip!"  

 

Well... there were a number of things wrong with his testing methodology... from poor mic placement to erroneous editing in his DAW.

 

I now watch his videos for comic relief.

 

😋

[Rick Reineke]

you can't clip with 32-bit float.

You can clip the audio prior to the converter though and it will still be clipped regardless of FP32.

[borjam]

I understand his frustration with overly enthusiastic marketing departments, but he doesn't seem to understand very well how digital audio works.

 

Dynamic range (let's say, macrodynamics) is different from resolution.  When he says his voice can be recorded with 8 bits I think he is talking about envelopes, range of amplitudes, not failthful audio recording. So much for 8 bit audio, even good old digital telephone audio used 12 bits! (Yes, non linear, A or µ law but it's still 12).

 

Resolution is related to S/N ratio because low resolution means quantization noise. So when you record you aim to make a good use of the A/D dynamic range leaving some headroom to prevent clipping. 

 

Of course everyone here knows that there is a chain of elements with different dynamic ranges involved in audio recording.

 

Air: Yes, it can become non linear for loud enough levels! (*)

Microphone 

Preamp

ADC.

 

And turns out many beginners (for whom many Rode products are aimed!) tend to fail setting up recording levels and clip. So, what's wrong with making their lives easier? 

 

As for controlling all of the circumstances or not, well, it depends. For people shooting guerrilla style, or documentaries, or even doing nature recordings, all kinds of unpredictable stuff can happen! Imagine you are recording a distant bird (with lots of preamp gain!) and suddenly an interesting bird comes close and sings. 

 

So yes, it won't capture 32 bit audio. Fine. But it will be much more lenient with recording levels. So what's wrong?

 

 

Youtube videos. UGH!

 

 

[Michael Scrip]

57 minutes ago, Rick Reineke said:

You can clip the audio prior to the converter though and it will still be clipped regardless of FP32.

That's correct.  But assuming you had everyone right... except you had the levels on your recorder set a little too high or there was an unexpected loud sound... there's still a good chance that you will be able to recover those clipped peaks in editing.  That's the big benefit of recording to a 32-bit float file.

 

In one of his "tests" he yelled into a tiny lav mic that was right in front of his mouth... so of course he overloaded the mic and it sounded distorted.  Obviously there's no fixing that in post.

 

But I downloaded some of his other sample files and I was able to fix most of what he said was unfixable.  It seems he doesn't understand how to properly edit in a DAW.  He knew enough to lower the levels in Adobe Audition... but he still had it peaking above 0db so it sounded clipped and distorted.  

 

And his conclusion was "See?  It's still clipped!"

 

Meanwhile... I took that same file... normalized each section... and made it sound great.

 

¯\_(ツ)_/¯

[Jim Feeley]

3 hours ago, Rick Reineke said:

you can't clip with 32-bit float.

You can clip the audio prior to the converter though and it will still be clipped regardless of FP32.

 

You Shure can. 

[glenn]

32 bit recording and it’s promotion as a better way to go over 24 bit recording is plain and simple horse shit. While some of the you tube videos message is confused by talking about 8 bit recording, the basic message that 24 bit recording with its 144 db dynamic range and 16.7 million levels is more than capable of capturing the output of any A-D converter without any loss of dynamic range.  As others have pointed out if your mic input is clipped or not adjusted correctly to provide enough signal to the A-D, 32 bit floating recording will only do a great job of getting bad sound from point A to point B. Adjusting the mic preamp to the correct level is mandatory to get the best recording. Any mic input that can not be adjusted provides sub par audio that can never be improved further down the signal chain by changing the gain in a mixer or recorder. Not even 64 bits will fix that. Maybe that’s the next big thing in snake oil advertising and promotion. 

[Michael Scrip]

Yes you should always try to set your levels.  But sometimes accidents happen.

 

Do people use limiters?  Or safety tracks?  Why would anyone need limiters or safety tracks if they always set their levels correctly every time?  🤔

 

Answer:  these are all safety features.

 

I think it's cool that you have the ability to lower the level of a 32-bit float file afterwards if the recording gain was set too high.  Or raise the level if the recording gain was set too low.  Ideally you shouldn't set your levels wrong.  But sometimes accidents happen.

 

If you make a mistake and clip a 24-bit file... it's ruined.  But with a 32-bit float file... there's a good chance you can recover it like in the example below:

 

 

The 24-bit file is clipped and there's nothing you can do about it.  But the 32-bit float file can be adjusted afterwards.  That doesn't seem like snake oil to me.  But I'm just an enthusiast.  

 

Professionals don't have 32-bit float in their expensive recorders.  That's because they never make mistakes.

 

But manufacturers (Sound Devices, Zoom, Tascam, Deity, etc) seem to think 32-bit float is a good safety feature to have in their less-expensive consumer-oriented recorders.

 

It's just another tool in the toolbox.  

 

😊

 

 

 

 

[joepfeil]

A good safety feature is to use appropriate levels through any signal flow; not to use a "black box" (32bit float magic), to relinquish the duty of maintaining appropriate levels through the signal chain.

 

My 2 cents

[Ridley]

On 9/7/2023 at 7:23 AM, joepfeil said:

A good safety feature is to use appropriate levels through any signal flow; not to use a "black box" (32bit float magic), to relinquish the duty of maintaining appropriate levels through the signal chain.

 

My 2 cents

 

Agreed. There is no substitute for a properly staged signal chain.