Working with whispery vocals (Is there such thing as a de-effer?)

Author
Spaceduck
Max Output Level: -50.5 dBFS
  • Total Posts : 2499
  • Joined: 2004/12/29 12:51:03
  • Status: offline
2008/10/19 18:18:08 (permalink)

Working with whispery vocals (Is there such thing as a de-effer?)

Working with a really quiet vocal track, almost whispery. I've got a compressor plugin with a de-eSSer which works ok, but the Fs are still too hot. Is there such thing as a de-eFFer? How about using EQ? Any suggestions?

Spaceduck music [HERE]
Spaceduck videos [HERE]
#1

10 Replies Related Threads

    kwgm
    Max Output Level: -52.5 dBFS
    • Total Posts : 2271
    • Joined: 2006/10/12 00:14:20
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/19 22:02:55 (permalink)
    Yes, use EQ to make your de-effer.

    First setup a parametric EQ to use a narrow band pass boost to exaggerate the fffs. (4 - 8 pole bp filter with high Q and +12db gain at around 4Khz). Then scan the frequency until the efffs get their worst. Once you find the proper frequency, reverse the gain and see what you've got.

    You might also try using a compressor/gate module before the EQ to properly fit your correction above a certain threshold in the attack portion of the envelope.

    Finally, if that's not complicated enough, you may discover that mixing this processed signal with the original signal gives you what you want.

    Have fun with your de-effer and please report back with your results.


    --kwgm
    #2
    mattplaysguitar
    Max Output Level: -55.5 dBFS
    • Total Posts : 1992
    • Joined: 2006/01/02 00:27:42
    • Location: Gold Coast, Australia
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 00:06:12 (permalink)
    A de-esser is essentially a multiband compressor. Use the EQ technique described above to find the centre frequency of the F's, then use a multiband compressor tightly constrained on that area to cut the F's. Better than using just EQ cause this way that frequency is not ALWAYS being cut from your vocal. When the F's sounds get strong, compressor kicks in and cuts it out, otherwise it sits there unnoticed. Use a quick attack and delay. You should be able to work it out yourself but if you need any more clues on it you know what to do!


    In movies they do it a much simpler way - simply blank the F's out with a sine wave at 200Hz... Bleeeeep. Yeah OK that was lame.

    EDIT: To add lame joke..
    post edited by mattplaysguitar - 2008/10/20 01:00:46
    #3
    Spaceduck
    Max Output Level: -50.5 dBFS
    • Total Posts : 2499
    • Joined: 2004/12/29 12:51:03
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 09:54:26 (permalink)

    ORIGINAL: mattplaysguitar

    EDIT: To add lame joke..

    In movies they do it a much simpler way - simply blank the F's out with a sine wave at 200Hz... Bleeeeep. Yeah OK that was lame.


    The lyrics (from an Emily Dickinson poem) start off
    A dew sufficed itself and satisfied a leaf

    With bleeps it would be
    A dew BLEEPED itself and BLEEPED a BLEEP.

    Not bad if I was doing gangsta rap

    Great technique kwgm, I learned something scary. The 'F' sound is not limited to any particular frequency; it runs the full spectrum. But using your technique I managed to isolate the most annoying bands which turn out to be 700-1k. I'd love to chop more, but I don't want to kill the original signal.

    Here's my 1st try using only EQ (-18dB at 700Hz & 1kHz, Q=7)

    First the original harsh F version
    http://soundclick.com/share?songid=6990028

    Now the EQ'd version
    http://soundclick.com/share?songid=6990031

    Mattplays, what plugin do you suggest for multiband compression? I've got the Sonitus Multiband which came with Sonar, but I'm having trouble making heads or tails of it. Do you know of any good ones out there (freebies if any)?

    Spaceduck music [HERE]
    Spaceduck videos [HERE]
    #4
    mattplaysguitar
    Max Output Level: -55.5 dBFS
    • Total Posts : 1992
    • Joined: 2006/01/02 00:27:42
    • Location: Gold Coast, Australia
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 10:01:41 (permalink)
    Sonitus should work fine. Might be a little more difficult that the Fffs come in at different frequencies... I got to go to bed now so won't get into details, but another option is to simply use automation on your EQ, but that's going to take ages..

    If I can get a chance tomorrow (in about 12 hours from now) I will see how I fair with Sonitus Multiband and let you know how if I get it to work. My theory though is that if it don't work with Sonitus, another plug ain't gonna make too much of a difference. There is also Ozone if you have it and want to give it a try while waiting.

    Goodluck!

    Oh and I like the bleep version Sounds funky!
    #5
    bitflipper
    01100010 01101001 01110100 01100110 01101100 01101
    • Total Posts : 26036
    • Joined: 2006/09/17 11:23:23
    • Location: Everett, WA USA
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 17:29:12 (permalink)
    F's aren't any harder to treat than "S"'s. They are similar, both essentially white noise, although the F's energy is shifted downward, with more in the 700-2K region than an S, and less above 8K. In both cases, their broadband nature works to your advantage, since all you have to do is turn 'em down.

    Use V-Vocal. The "F"s and "S"s are easy to identify in the wave. Just go into Dynamics mode, highlight the offending consonant and lower it. Experimenting while auditioning will tell you when you're there, but I find they usually end up being 20-40% lower than the surrounding wave. (No, that percentage is not a joke)

    Now, if the F's are too prominent as a result of the singer being too close to the mic, you may be facing an excess of low-frequency energy, similar to a plosive. EQ can help mitigate that by severely rolling off the low end, but to properly fix it the only way I know is with a spectral editor.


    All else is in doubt, so this is the truth I cling to. 

    My Stuff
    #6
    mattplaysguitar
    Max Output Level: -55.5 dBFS
    • Total Posts : 1992
    • Joined: 2006/01/02 00:27:42
    • Location: Gold Coast, Australia
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 20:06:48 (permalink)
    Ok I gave up on the multiband idea (I think it works better on sibilance). Obviously I am limited as I didn't have the multitrack, but the vocal to music ratio was high so it could have worked.

    Now what I can't work out is why I didn't think of what bitflipper just said.. Volume envelopes. Gosh, I use volume envelopes on all my sibilance if I have it and I never even USE a de-esser! And I still didn't think of it! I think I was too tired. Anyway, volume envelopes are the key. If all else fails, this will work. Always worked well for me, just takes ages.

    You can do it in either v-vocal as bit said, or just with clip automation. I generally use clip automation as v-vocal can sometimes do stuff to your vocals even with NO pitch correction. It makes them sound terrible. It is much easier, but for this reason, I use clip automation. Try, and you should be happy with the results. Just a warning, it takes forever. And it's even harder with F's because the wave does not stand out as prominently as sibilance does.
    #7
    Spaceduck
    Max Output Level: -50.5 dBFS
    • Total Posts : 2499
    • Joined: 2004/12/29 12:51:03
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 20:26:21 (permalink)
    Yup, you guys are right. Volume control is all I need here.

    I was hoping to avoid clip automation because I've already run into that rascally "envelope automation bug" in this project. V-Vocal sounds more promising. Or I could always split the clip at the Fs, overlap the 2 resulting clips, and play with the clip fades to bring down the volume.

    In any case, I think EQ rolloff is a good idea. There's some rumble I'd like to get rid of. Yeah, mic was too close. But unfortunately there was no other way to do it due to ambient noise. Dang garbage trucks flying down my street all day; I really need to soundproof this place ...or not record vocals on Fridays

    Spaceduck music [HERE]
    Spaceduck videos [HERE]
    #8
    mattplaysguitar
    Max Output Level: -55.5 dBFS
    • Total Posts : 1992
    • Joined: 2006/01/02 00:27:42
    • Location: Gold Coast, Australia
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/20 20:58:34 (permalink)
    Yeah splitting slips could work well enough if you're having troubles with a bug (never had that before luckily). I think the close micing does sound good though, but if you had the luxury of a nice vocal booth and soundproofed walls it would still sound close from a foot away. One other option is to really utilise your mic technique if you must mic so close. Try with the mic a little higher, more near your nose (just don't pick up your breathing!!), so your mouth is not in the direct line of sight so it doesn't 'blow' into the diaphragm. Or you could try slightly moving your head of axis when singing 'just' the F sounds. That one may be a little tricky but if you get it timed well it should sound very good. I think the closer you get to the mic, the more important mic technique becomes. So in this case, it's probably very important.
    #9
    Philip
    Max Output Level: -34.5 dBFS
    • Total Posts : 4062
    • Joined: 2007/03/21 13:09:13
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/24 13:51:48 (permalink)
    Excellent thread for me.
    try slightly moving your head of axis when singing 'just' the F sounds


    -- I expect SD may be using a wind shield (whatever its called) on the mic for plosives

    (I love Emily D. poems ... so mysterious, sublime, etc. ... I think she died of Bright's Disease (kidney)) SD, Glad to see you're experimenting with whispers. Its a well documented pearl for vocal fx. I hope to hear your song or sketch soon.

    Offensive *Sibilants* have their Q(s) anywhere between 2K and 5K; I wrestle with these more than anything in my voxes! Drum siblants are oft sweet crackles (to my ears).

    I suppose I should adjust Qs and precise EQs more and listen to all these fine suggestions by Matt, Bit, and Kurt.

    Voxengo Voxformer's side-chain de-ess compressor has a 'center' and 'bandwidth' (which I'll start appreciating). Most of my 'F's I've reduced, seemingly OK, along with the 'S's with a combination of techniques.

    Edit/Correction:
    Offensive *Sibilants* have their Q(s) anywhere from 4000 cycles and up. (probably not as low as 2000 hz)
    post edited by Philip - 2008/10/26 09:13:45

    Philip  
    (Isa 5:12 And the harp, and the viol, the tabret, and pipe, and wine, are in their feasts: but they regard not the work of the LORD)

    Raised-Again 3http://soundclick.com/share.cfm?id=12307501
    #10
    Spaceduck
    Max Output Level: -50.5 dBFS
    • Total Posts : 2499
    • Joined: 2004/12/29 12:51:03
    • Status: offline
    RE: Working with whispery vocals (Is there such thing as a de-effer?) 2008/10/25 13:37:31 (permalink)

    ORIGINAL: Philip

    try slightly moving your head of axis when singing 'just' the F sounds


    -- I expect SD may be using a wind shield (whatever its called) on the mic for plosives


    uh... oopsie. no shield

    The shield I have (a metal one) causes a weird "talking through a fan" effect at whisper levels, so I ditched it. Instead I turned the mic 45 degrees. But with the 'F' sound, the air seems to come out the sides of your mouth--right in the line of fire. Guess I'll just have to back off next time.
    ORIGINAL: Philip
    (I love Emily D. poems ... so mysterious, sublime, etc. ... I think she died of Bright's Disease (kidney)) SD, Glad to see you're experimenting with whispers. Its a well documented pearl for vocal fx. I hope to hear your song or sketch soon.

    Yup, whispers, Emily Dickinson... it's all new territory for me. What a tragic figure she was (wrote 1800 poems, but less than a dozen were published during her lifetime). I stumbled on her poem and it fit the theme of this song as a great prologue. Also that line has a lot of 'F' alliteration which I thought would be a neat idea since I follow it with a line that has some 'S' alliteration.

    Don't ask me why I decided to use the two most dangerous sounds in the vocal world. Seemed like a good idea at the time.

    Thanks P, the song is progressing nicely. I've brought the Fs down with a combination of all the suggestions in this thread. Hope to post in the Songs forum soon!

    Spaceduck music [HERE]
    Spaceduck videos [HERE]
    #11
    Jump to:
    © 2024 APG vNext Commercial Version 5.1