Advice on Audio Quality

Mark Nicholson

Well-Known Member
Joined
Nov 11, 2020
Messages
61
Karma
29
From
San Diego, CA
Gear owned
Model 12
Hello everyone,

I'm a noob at sound processing, and I would be very grateful for any advice you can give me on how I'm getting on. I have recorded some audio for my YouTube channel on optical design. It's a yawnfest if you're not into it, but my goal is to have a really clear, professional voice track for spoken voice. This sample file:

https://soundcloud.com/user-6836741...683674132-115654328/sets/tascam-model12-tests

is the audio as it comes off the Tascam. I have my microphone going into channel 1, Gain at about 1 o'clock and compression at about 2 o'clock. Everything else is flat. I record on the Tascam, and just transfer the file to PC over USB when finished. I then put this track through Audacity and process it to give

https://soundcloud.com/user-6836741...683674132-115654328/sets/tascam-model12-tests

Within Audacity, I use compression, limiting, a bit of bass boost EQ, a bit of top boost EQ and a bit of de-essing with a clip around the 6 kHz region. I also do noise reduction and normalization.

I'm looking for feedback on how well I've done this, and any steps you think I should take to get a better final result. Thanks in advance!

- Mark
 
Hi Mark, here's my take on it. There is a clear difference between the two clips and the processed one certainly sounds clearer and more intelligible. However, I do have some remarks:

1. You can make quite an improvement IMO with starting at the beginning: the mic and use of it. You don't mention it, but what type is used and how close to your mouth is it? I hear quite some ambience in the track (the sound of the 'room') but ofcourse you are apparently doing video so you may be limited in what you can do here.

2. I think your top end EQ boost is defeating the de-esser function - to me the sibilance did not get better, maybe worse (but I must say I'm listening on laptop speakers).

3. If you use compression and limiting there should be no need to also normalize. Without crushing everything to a brick I think you can use more compression and use the limiter as a final safeguard against clipping and incidental peaks. Set it to a maximum of at least -1 dB (because of further conversions down the line in the video editor) and you should be good to go.
Hope this helps!
 
  • Like
Reactions: Mark Nicholson
Thanks Arjan! Yes the room is very hard. I'm using a Comica wireless Lavalier microphone. I agree, I need to combine the hi-boost and de-ess into a single operation. I'll try your suggestions on the compression. Thanks!
 
  • Like
Reactions: Arjan P
@Mark Nicholson high end EQ boost often causes sibilance issues. See how little boost you can get away with. I could tell you stories....
 
OK, so I re-edited the original in Audacity. I used their basic Bass Boost filter with a notch for de-essing:

hd3DonY.png


Here's the result on SoundCloud:

https://soundcloud.com/user-683674132-115654328/reexport-mainbody

What say you?
 
The audio is clean but it sounds very unnatural. There is a sharp, shrill quality to it that brings out those S's. The curve above bothers me because the bottom end is really boosted (and there is no reason to boost anything below 40Hz with dialog, maybe higher depending) and also you've dipped 6.3 and 8k yet it sounds shrill. The sibilance is mostly likely below that part of the range that you've dipped though.

Please send me the file to info @ phoenixmediaforge dot com and I will analyze this with RX8 Advanced and find where the sibilance is. I'll process the file and send it back to you. Hopefully we can find an EQ curve that works with this dialog. RX8 has tools specifically for dialog cleanup. I'll share what I find and you can use that going forward.
 
  • Like
Reactions: Mark Nicholson
If it's the one I think, then the problem is quite probably the mic.

You seem to be using a wireless system which costs something like $120 US total, and the MICROPHONE part is sold for something like $5, : https://www.adorama.com/iacvmmo2.html

That suggests your quality issues may well be hardware based. Lavalier mics are always going to be harder to work with in general because of where they are located w/r/t your voice. (Of course they get used a lot by pros, but, well, those are pros. I'm not a pro and I assume you aren't either.) The combination of super low end $ mic AND lower-end wireless AND lavalier location AND not-a-pro may make this very hard.

If it were me, I would spend a bit of money on a different mic. Even the same $5 capsule on a headset will be much better than what you have now, because the microphone will be in front of your mouth instead of on your chest. And if you can spend closer to $40 there are many more solid options.

The headset units come with all sorts of connector options so you should be able to find one which will work with your existing wireless transmitter. Honestly anything is likely to be an improvement.

If you're handy and can skip buying lattes for a few days, you can always get a Shure PGA31 and rewire it to an appropriate terminator for your system (double check to make sure everything plays nicely w/ phantom power, if needed.) That's a super popular and well regarded basic headset mic which costs only $39, and if you can solder then adding a connector is not hard.

Or you can search Amazon or Adorama etc. and find something cheap which already has the right plug.

If you're investing all the capital into the recording you may want to give that a shot; that's what I would do. I have used a headset mic in a band as a keyboard player and frankly it was pretty great.
 
On the microphone front, I'm considering the Shure MV7. For you guys who are sound engineers, is that a good choice or is there something you'd recommend instead?
 
What do you want it to pick up? What DON'T you want it to pick up?

Anyway:

I have never heard of an MV7 but having just looked it up:

A big part of the MV7 appears to be that it is an all-inclusive solution. If you own an MV7, all you need is a laptop. To quote the materials, it will basically save you the cost of things like A/D conversion, EQ, phantom power if needed, etc.

Moreover, the MV7 is presumably easy to use. It is probably tuned reasonably well i.e. the preamps inside it are already calibrated to produce a workable signal. And there's supposedly some included software for EQ and such.

That is because Shure assumes you don't own anything as complex as a mixing board. So the MV7 will be simpler to use if the only thing you want to do is plug and play.

But if you have a T12/24/32 then you ALREADY own quite a bit of underlying equipment. If you're interested in maximizing that equipment you might be happier with the same money on a different mic. Or not. try both...?
 
This looks like the classic radio mic by Shure, the SM-7B. Considering the price of the MV7 (and the added functionality) it will not be the same mic, which doesn't mean it's bad. But Hammarlund is right: What is your primary use?

I was under the impression that you needed to be mobile as you were moving around, showing people whatever it is that your videos are about - and that is quite a restriction (to lavalier or headset mics). If that's not the case, you're open to a lot of mics!
 
Thanks guys. I'm not doing out and about stuff.My focus is on good speech quality from a sitting down or standing up presenter. No huge need to be mobile.

Despite owning the Tascam, I have no problems with something that is simple and easy to use :)
 
"good speech quality from a sitting down or standing up presenter"

It doesn't matter if you're sitting or standing, what matters is whether or not you're staying still in front of the microphone.

Simply put, it's hard to mic people who move around, unless you want to mic the whole room. Going from sitting to standing may count as "moving around," depending on what you mean. Though of course, you can always just grab your mic boom and adjust it every time you change from standing to sitting.

Anyway, if the question here is "will that mic work?" the answer is yes: It will shurely (pun intended) be better than a lot of mics (though probably not as good as a properly-set-up-and-mixed SM7) and it is virtually certain to sound a lot better than what you have right now.

The real questions are the competitors.

1) Other similarly designed large dynamic mics in same $250 price range, like the Rode Procaster (XLR only) or Rode Podcaster (USB.)

2) other styles of mics in $250 range, like a higher end headset mic or even a headphones-and-mic broadcast set

3) Cheaper USB mics like Blue Yeti.

4) Cheaper other styles.
 
Please send me the file to info @ phoenixmediaforge dot com and I will analyze this with RX8 Advanced and find where the sibilance is. I'll process the file and send it back to you.

Hey @MJK, I see that RX Elements is available for $29 right now, and I guess that at $29 you can't go wrong. How useful are the tools in that version for doing this kind of thing?
 
@Mark Nicholson no use at all. I have Advanced. Check the version differences and you'll see that none of the dialog tools you need are available in the lowest version.

FWIW, in the little movie/TV work I've done, they always use a wireless lav.
 
I have a related question. I took a voice recording into Cakewalk and used the MAnalyser plugin to get a spectrum of my voice, averaged over ~3 seconds:

BLCvzYm.png


What is this telling me, and what would a 'good radio voice' look like? For example, should it be nice and flat, like a plateau? I can see the effect on the spectrum of the various EQ options, but I'm not really sure what my goal is? Flatten the spectrum? Flat with a tilt?
 
Hi Mark. There is inherent danger in engineering by looking at a curve. Looking at the graphic above doesn't tell me how it sounds. There is a small peak above 5k (not enough resolution to say exactly where it is). This peak may, or may not be an issue. There is a roll off at around 80 Hz. If there is rumble on the audio, that may be a good thing. Or, it may not be a good thing depending upon the overall sound. When we roll off the bottom, under compression it brings up the high end. I cannot tell you what a "good radio voice" looks like but I can tell you what a good radio voice sounds like. The bottom line is: Train Your Ears.
 
  • Like
Reactions: Arjan P

New posts

New threads

Members online