digital delay problems.

[martin]

I'm trying to use a digital delay to double a lead vocal (ie make it sound like the singer is doubling himself) and I'm having problems making it sound natural. What is the typical way to do this?
I have my delay set at 3 milliseconds, with equal parts original and delayed signal. This sounds right when he's speaking, but when he starts singing it gets a really robotic quality. It seems like I need to add some EQ to get back the clarity in the sibilants of the voice - I've been boosting at about 8K and this helps, but something else is making it sound really artificial and cold. What am I not doing?

[William Choi]

Well, If you haven't already fiddled with the processor to get it the way you want it.
3milliseconds is WAY TOO fast for the double vocal effect(although, 3milliseconds gets you a totally awesome guitar/sitar effect if you set the decay to non stop:Take note that from about 1 to 9 milliseconds gets you different harmony effects altho can also give you flat notes),
anyway. You should set it about 20 more or less milliseconds. and with 0 or 1 decay factor(just as long as you just get 1 echo), and set the efffect level to about the same level as you got the dry signal. And there you go. maybe 20-30 or so millisec.
Hope this helps.

[martin]

sorry - i think it must have already been at 30ms. My alesis says "003" in the display and i guess i interpreted it wrong, now that i think about it. The time is right. It just seems to be losing clarity because the sibilants are getting doubled and lost, and somewhere lower (around 1K or so) is giving me this really robotic sound when the fundamental pitch is being doubled. I know professional studios often pan the dry signal left and the delayed right, etc., but I don't have the capabilities to do that.
I'll fiddle with the EQ a bit more, but i was just wondering if there's something i need to do to make it sound natural.

[William Choi]

Sibilants is a bitch. I guess your singer has too much of the "SsS" coming out of the mouth. My suggestion is to use a pop filter for your mic if you don't already have one or just try your best to get rid of the highs frequencies(which isn't good at all for the sound but if you're on a tight budget...)
So I got rid of puhs and sehs with a just a cotton cloth way back when.
reading your 1st post it seems as you took care of the sibilances.Now you got the robo voice thing. Well, thats the effect buddy.Thats what it sounds like.
If you want natural, your best bet is to double it manually. Or try at different lengths of delay(normal acoustical lengths).Maybe that will work better.
Just what is the effect that you want?
Im thinking of Billy Corgan's(smashing pumpkins) hefty use of doubling vocals.

[This message has been edited by William Choi (edited 07-27-1999).]

[martin]

i think the sound i want is probably used on smashing pumpkins records, although i don't have any around to tell you. i want the subtle effect of two people singing, without the human imperfections of actually recording two different tracks. as well, i don't have two free tracks to record on anyway. if you listen to weezer's "buddy holly" in headphones or almost anything by sloan you can hear what i'm aiming for - a very crisp, short delay. it's used in pop music everywhere.

one guy told me to add about 15% "depth" (i don't have the machine with me to see if i have that effect) which randomizes the pitch and time a bit. does this help? what do you mean by "normal acoustical lengths"?

[William Choi]

normal acoustic lengths(a word i made up), i mean like typical presets like a warm,dry upside down sideways whathaveyou auditorium or showroom thing. a normal Environment. Although,you explained what you wanted with the example of Sloan. Im Canadian you know,eh. Fiddle with a chorus effect. with the delay unless you've already tried this.
One thing you can try, if you have aux-sends. Is pan your vocals to either Left or RIght and send a signal to your FX process and use a panning delay and set the delay accordingly. This might work and might work even better with a mult-tracker that has its returns that are L and R and jacks for each. Because you can't send only the wet signals to the returns, you might get some of the orignal signal coming in at mono so you don't get the Left/Right delaying effect. But give it a shot.

[Dragon]

Smashing Pumpkins...is that like Gallagher?



(I think that joke was mostly for drstawl)

[bsutton]

I am getting something of the same thing using a Digitech S100 for the doubling delay. I'm actually just playing with it and using the built-in doubling delay setting which I think is about 20 ms.

Anyway, I get a very tinny high end that sounds like what you are describing. I have been playing with EQing the upper freq range down substantially on the effect return but leaving it normal on the dry. This has the effect of doubling the mids and lows but seems to get rid of that annoying (to me, anyway) fake sound. I wish I could be more specific on the range I'm playing with but I'm at work and can't remember. It's an Alesis Studio 32 board and I'm pulling the upper freq EQ back to about 9:00 - 10:00 ...

Being a newbie myself, this info is probably worth about what you're paying for it

Bill