Recording Spoken Word - Help required please.

[Eddsmusic]

Hi everyone,

I have been a musician for many years and play keyboards & synths. I am relatively new to ableton live and music production.
I have googled for advice on this topic but there doesn't seem to be much out there. Hoping someone on the forum may be able to
offer advice.

I am trying to record spoken word to create meditation tracks. I am having trouble capturing a noise free vocal. I have been trying to use gates and EQ8, Spectrum to clean up the noise. I thought I had done a relatively good job (listening through headphones via audio interface jack) but then exported the project as a MP3. Listening back on laptop & mobile phone the hiss is terrible!

I have a few questions (some which will highlight my newbie status, ha ha) any suggestions gratefully received.

1) rather than using a gate to clean up a recording, is it possible to place the gate 'on the front end' so that it cleans up the noise before it is recorded?
2) if so, what is the signal route?

I am using an SE2000 large diaphragm condenser with pop shield. I also have access to Sennheiser e845S super cardioid, which I am going to borrow to try out. Via a zoom U24 audio interface in to live 10.

3) any tips on how best to go about capturing the best audio (least noise) would be much appreciated
4) tips on production & mastering for this type of vocal track - should I be using compression as well?

Many thanks
Edd

[jlgrimes]

Hi everyone,

I have been a musician for many years and play keyboards & synths. I am relatively new to ableton live and music production.
I have googled for advice on this topic but there doesn't seem to be much out there. Hoping someone on the forum may be able to
offer advice.

I am trying to record spoken word to create meditation tracks. I am having trouble capturing a noise free vocal. I have been trying to use gates and EQ8, Spectrum to clean up the noise. I thought I had done a relatively good job (listening through headphones via audio interface jack) but then exported the project as a MP3. Listening back on laptop & mobile phone the hiss is terrible!

I have a few questions (some which will highlight my newbie status, ha ha) any suggestions gratefully received.

1) rather than using a gate to clean up a recording, is it possible to place the gate 'on the front end' so that it cleans up the noise before it is recorded?
2) if so, what is the signal route?

I am using an SE2000 large diaphragm condenser with pop shield. I also have access to Sennheiser e845S super cardioid, which I am going to borrow to try out. Via a zoom U24 audio interface in to live 10.

3) any tips on how best to go about capturing the best audio (least noise) would be much appreciated
4) tips on production & mastering for this type of vocal track - should I be using compression as well?

Many thanks
Edd

Some compressors have built-in noise gates but you are probably better off post processing.

Basically though what is the source of the noise?

1. Acoustics (traffic/AC/Motors, fans etc)
2. Microphone
3. Mic Pre.
4. Audio interface/cabling, improper gain staging.



Depending on how quiet the noise is, it might not be as bad as perceived.

There are noise reduction algorithms you can also try.

Reaper has a plugin that does it (forget the name).

[Eddsmusic]

Thanks jlgrimes

It's general ambient hiss. so could be the room, mic, cables, gain staging.

I am playing around with things at the moment. When I export the individual spoken word track, my processing has worked.

I am just about to import that track back into live and see if that sorts the problem?

If that doesn't work I will try a process of elimination to try and find the source of the noise.

Cheers Edd

[dougdi]

Basically first use a hi pass filter to reduce any rumble, then if you gate use it before any compression because that would raise the noise floor. Of course avoiding pumping or hearing the gate work. Apply processing lightly and you could use more than one serially to help lessen any artifacts.

I use Izotope RX 7's spectral filter, which is pricey, there are other plugs that do it for less though I’m not up on what’s good or best.

Recording in a closet could solve your problems too, or setting up blankets around the mic position.

I hope that helps, ask if you have more questions, I’ve recorded thousands of hours at least throughout my career and am happy to help.

[fishmonkey]

gain staging is really important to maximise the signal-to-noise ratio. if you record at too low a level, then when you raise the gain (either with a fader or compressor) the noise will become more apparent.

if you are in an acoustically noisy space, then a dynamic microphone can be a better choice, as they are less sensitive than condensers.

also check out vocal recording techniques, there are lots of guides out there. where and how you position the mic in your room is also important.

[Da hand]

The best approach is to tackle as much of the noise problems at the pre-recording stage, rather than post-recording.

1) Making sure your environment is as noise-free as possible. Depending on the place, it can be computer fans, ventilation, outside noise, noisy fridges, etc. Try to eliminate as much of this noise as possible.

2) Type of microphone you choose will make a difference - a condenser is nice for voice, but if it captures a wide area, then all the noise you didn't eliminate totally in point 1 will still be captured to maybe a large enough degree that it remains problematic. Dynamic mics can be a solution, but I find them not as nice sounding for typical spoken word - in my experience. In a less than perfect environment, I find that using a condenser shotgun mic (the type of "Boom" mic used on film sets) works very well as it combines the nuances of a condenser mic with a very narrow area of sound capture. Just make sure you are not placing the voice actor in the same audio capture line as the noise you want to eliminate.

3) Pre-amps may add some hiss when pushed past the 3/4 volume mark and cheaper pre-amps will definitely add a lot of hiss if you push their volume too high. These are often the hiss culprits so watch out for this. You don't want your recording volume to be too low, as already mentioned, but if you are recording in 24 or 32 bit then you don't need to be recording close to 0dBFS either, you can be around -10/12 dbFS with no problems of loosing sound quality and then you don't have to drive your pre-amps too much.

Test all this on your headphones before recording by listening to the mic through the pre-amp/sound card to find the best settings. Once you have the best sound possible captured, then you can move onto post-recording treatments mentioned.

As for Gates at the recording stage, well they are not the best solution in my opinion. They will eliminate the noise between words or phrases, but once the gate is open to let the voice through, the noise is passed along with it. And then you will have parts of the recording with no noise at all, and parts with full noise, often making a big contrast between these parts - all automated with a bit of guesswork. I rather have the recording with the noise as it is (after steps 1-3) for the duration of the recording and then clean it up in post - depending on what is needed and when.

[fishmonkey]

100% agree. concentrate on capturing the best raw recording possible. this applies whatever you are recording.

avoid turd polishing!

[Shift Gorden]

Yup, addressing the source is the way to go!

That being said, if you still have persistent hiss - and as you know - a gate won't remove it. Dougdi mentioned RX7. I use it all the time for client spoken-word stuff. It's expensive, but it's also magical. Hiss, hum, pops and clicks, breath noises, it's fab.

[Eddsmusic]

Hello everyone

Thanks to you all for some quality advice.

I have been trying out a few of the suggestions and seen much improvement.

Much appreciated
Craig

[TLW]

Audacity is a free audio editing/recording application that has pretty good noise reduction/removal abilities. So does Adobe Audition if you already have a Creative Cloud account.

Gating is often the worst way of removing hiss and background noise from recording, because it doesn’t remove background hiss and noise from the bits you want people to be able to hear so the hiss keeps switching on and off as the gate opens and closes. It’s tedious to do, but a manually created volume automation curve which smoothly fades the noise out of the gaps in the audio can often sound a lot better.

Gates have another problem known as “chattering” - this is when the gate rapidly alternates between closed and open when the audio is close to the gate’s threshold setting and it sounds weird and unnatural. Getting threshold, attack and release set up to be as unobtrusive as possible can often be tricky.

Preventing the noise being recorded in the first place is obviously the best thing to do. A hardware gate immediately after the pre-amp can help but is still subject to chattering and the “noise cutting in and out” problems. If the speaker doesn’t maintain a sufficiently loud volume to keep the gate above its threshold then the gate will kick in when it’s not wanted.

For live sound engineering gates are often not so noticeable in a band setting but for a single recorded speaking voice it’s not what I’d consider the ideal solution to noise issues. As well as directional mics the shields that mount on mic stands behind the mic can often add quite a lot of isolation from noise sources behind the shield. Another approach that can be combined with shielding and general noise reduction is to use a mic with a figure 8 pickup pattern and place it so the noise source is off to the side of the pickup pattern - figure 8 mics often have better noise rejection than cardoid or hyper-cardoid ones. They also suffer much less from proximity effect which can lead to bass frequencies being boosted by cardoid mics if the sound source or anything reflecting sound into them is too close to the mic. Some singers are very good at using the proximity effect for vocal tonal changes and effects but for spoken word or instruments it can be a liability.

Finally, avoid using compression if at all possible - it will raise the noise floor and make it much more noticable.

[Eddsmusic]

Thanks TLW for taking the time to construct such an in depth response. Yes I have found that with compression. Many people on youtube recommend using compression for spoken word, I thought I was using it incorrectly. Probably a bit of both! haha.

[TLW]

Compression can be used unobtrusively, but only if you get rid of the noise first. You might find Waves Vocal Rider useful, it basically maintains the vocal volume within limits you set but without being a compressor and is quite good at ignoring very low volumes such as low volume noise.