Experimenting with FFT EQ Filter in Samplitude

Discussion in 'Tracking / Mixing / Editing' started by DonnyThompson, Oct 12, 2014.

  1. DonnyThompson

    DonnyThompson Distinguished Member

    LOL... okay, so let me explain.

    I've recently been working with, and digging into, the FFT EQ Filter in Samplitude.

    One of the very cool things about it, is the ability for FFT to "learn" EQ curves... so, if you are working with a vocalist on one session, and they want to come in on another day and "patch" another vocal part to that same track - like a phrase or something - FFT has the ability to "learn" the EQ of the first performance and allow you to apply the results of that analysis to the new part, making it much less noticeable.

    This got me to thinking...

    What would happen if I fed the FFT a solo voice track of another singer?

    It would analyze that singer's voice, or, maybe even analyze the tonal characteristics of a certain mic or pre...

    So, say that I really liked the tone/EQ or textures of a solo Peter Gabriel vocal... what would happen if I then applied that analysis to my own voice? Well, of course, I won't be instantly turned into Peter - LOL - but, could his tonal characteristics - EQ, Timbre, Mic character, etc., be applied to another voice with interesting results? Maybe even favorable results?

    Here's the experiment, if you want to play along...

    Send me a 5 to 10 second .wav file of you either singing, or just even speaking... using a nice mic, and a nice pre. Use your own natural voice. Don't try to mimic anyone else.

    I'll then feed the FFT your vocal/tonal characteristics, and then, using the FFT "save as FX insert" function, apply the settings of the FFT analysis to my own singing voice, and I'll place the audio results here on RO.

    And, if you don't want to be a part of the experiment, that's totally cool. I'm just presenting this as something new and interesting to discuss and critique.

    It may all turn out to be foolish. But who knows?

    Send your MONO 44.1/24 five to ten second wave files to:


    Please give me your RO screen name for reference.

    Notate the mic, pre, DAW platform you used.

    Please don't add any processing - no EQ, no GR, no reverb... Nothing. Just raw and flat!


  2. DonnyThompson

    DonnyThompson Distinguished Member

    For those who are interested, here's a very cool video by Tim Dolbear explaining the use of FFT for "take matching" vocals recorded at different times:

    View: https://www.youtube.com/watch?v=J1WamQ1j48U

    at around 1:49.

    pcrecord and bigtree like this.
  3. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    Donny, it would be more effective if you would choose a know song and a specific part we could sing for you.
    The reason is, if I sing a hi pitch part and you apply it to a lower part, I bet it won't translate well because it will cut too much low end.
    (just me thinking ;) )
  4. DonnyThompson

    DonnyThompson Distinguished Member

    It doesn't even need to be a singing part, Marco... just your normal speaking voice will do for the initial experiment. Something as simple as "Hi Donny this is Marco Speaking through a_____mic into a ____ preamp" would be enough at this point to feed the FFT.

    During the second stage of the experiment, we could get into a singing part. The experiment is to see the result of applying the normal timbre of your voice. ;)
  5. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    If I understand correctly, the FFT is an EQ matching tool right ?
    I'm not sure if it will act on the character of the voice but I'm willing to help you !

    When I get the chance, I'll send you a file or two ;)

    AH ! I got an Idea, why don't you send a file of your voice, I'll send it to the LA610 and the 4-710, you could analyse how the signal is transformed...
    Comparing the same signal through different preamps or other units could be of an interest too. ;)
  6. DonnyThompson

    DonnyThompson Distinguished Member

    You got a deal. ;)

    Can you PM me your email address again... I'm sure I have it somewhere but I recently upgraded Mozilla and I can't guarantee it kept it in the address book...
  7. Reverend Lucas

    Reverend Lucas Active Member


    Your experiment is an interesting one. I'm curious about the algorithm this plugin uses to determine the EQ curve it applies. It looks like a potentially powerful tool, and I'm wondering how it accounts for different types of inputs. What happens if you apply a sine wave, for example? Does it just surgically notch that frequency? I could see having all sorts of fun applying different signals and seeing what happens.

    Anyway, my gear is modest, but I'd be happy to lay down another blurb if you're still looking for another wav file. I noticed the video suggested using at least 20 seconds' worth. Have you had good luck with shorter samples?

  8. DonnyThompson

    DonnyThompson Distinguished Member

    The length of the file doesn't have to be 20 seconds, Rev. I have "sampled" clips as short as 5 seconds. I think that 20 seconds is just the default setting.

    Again gang, this is just an experiment, hence the title. It may not prove to be worth anything... I just thought I'd see what happens.

    Please do send a file! The more the merrier! :)
  9. audiokid

    audiokid Staff

    I've hardly used this.
    I'm up for this as well, Donny! My days for singing are behind me for now but I'll do a speech through my Soundelux e250 and M-2b. It will be a while before I have time though, so slammed but please continue in the mean time.
  10. DonnyThompson

    DonnyThompson Distinguished Member

    That would be great Chris. :) Rev Lucas has sent me a file, I'm gonna work on that one today, and Marco and I have a related side experiment going on that might be interesting as well.
  11. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    I sent you a file as well Donny, (used the email address of our last exchange)
    4 different speaches 2 preamps 2 mics. Hope it's ok ;)
  12. DonnyThompson

    DonnyThompson Distinguished Member

    awesome! I'm gonna work on these today once I get the little woman off to work. ;)
  13. DonnyThompson

    DonnyThompson Distinguished Member

    just a quick update guys... I kinda got swamped with work the last 5 days or so, seemed to hit out of nowhere. I have NOT forgotten about this and I'll get to it just as soon as I get a few of these client's projects done, probably within next few days.

  14. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    Having a lot of work is a good thing... ;)
  15. DonnyThompson

    DonnyThompson Distinguished Member

    Yeah, it's a lot of editing...sigh... trying to fix rough performances. I didn't track the project so I'm left at the mercy of the person who did. I say "person" because I'm hesitant to refer to this sloppy type of recording as being done by a real engineer. LOL

    For as convenient as I find digital to be, sometimes I miss the days of tape - because you could only fix things to a certain extent; and you could actually and rightfully blame a bad performance on the performer, as opposed to today's digital, where it's becoming all too popular to expect a mixing engineer to take crap and make it perfect.

    Typical conversation:

    Client: I know I played some wrong notes on the bass part, and I'm a little behind the beat too..

    Donny: I would call playing a full 16th note off as being more than just a little behind the beat...

    Client: I think it's referred to a playing in "the pocket".

    Donny: Yeah? Who's pocket are ya playing in? Maybe the band next door? Actually, I think it's called "playing out of time".

    Client: Well yeah, maybe a little, I guess. But you can fix it can't you?

    Donny: (Sigh)....To a point... I don't suppose that you would perhaps ever consider recording a performance that was good to begin with...

    Client: Total Silence. Looks at me like I'm from Venus.

    Donny: Okay, I'll see what I can do. However, I don't know that I can do much about making your lead singer sound any less like a singing worm.

    Client: (stoned laugh) Hehe. Singing worm.

    Donny: Alright, I'll see what I can do.

    Client: Great. Can I pick this up in an hour?

  16. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    The sad part is, if you play the instrument yourself, it's sometime faster to play the song rather than fix the bad performance. But at this point they would need to put your name on the record ;)
  17. kmetal

    kmetal Kyle P. Gushue Well-Known Member

    Then they ask why it doesn't sound like their favorite band, and also ask why it took so long to "mix". It's so worth it tho to be disciplined and fix all the technical stuff before I start messing around w sonics. I hate editing, but once you get going, especially on mediocre performances, it becomes hard to not edit everything thru and thru. It's like if you put one part the way it "should" be, the whole rest of the song suddenly looks worse.

    I can't wait till touchscreen editing is the norm because, just what I can do in garage band on an iPad is awsome, but quite limited. The tedium of lining up things w a mouse and in my case squinting to see the I beam tool. Ditto for midi editing/programming, great stuff, but I ignore it due to the workflow, when its touch screen, it'll be so much more right brained. It's gonna take a while I. Sure tho...
  18. DonnyThompson

    DonnyThompson Distinguished Member

    Yup. That's the thing. Because as you fix certain things, those improvements can actually make the other unimproved parts sound even worse... as if that's even possible sometimes. LOL
  19. kmetal

    kmetal Kyle P. Gushue Well-Known Member

    Dude it's bad enough alone, how bout when the client catches wind of what you can do then then starts wi the pointing at the screen and the " little bit more that ways" or worse when your lining something up, and every time you hit stop and play to quickly check audition it its like " nope" "nope" , it's like dude, do you really think I think I'm getting it each time? I'm seeing how far off I still am, wait till I say "check this out, or ok, listen."

    I swear the gui s for eqs are terrible for that. Hey move that red dot over there, what's it sound like over there!?. Lol I don't really mind, it, it's cool they're interested. It just takes longer sometimes.

    This is a big thing w people not familiar w the process, and why I don't charge for setup time, or I'll take extra time that I don't charge for to make it right as I can. Cuz they don't see the value in that. If they physically see you moving stuff it looks like ur doing something important. When your micing mics around and hitting chords, it just looks like your messing around, which you are but professionally. But honesltly the editing to death seems necessary about half the time for me, and boy is it tediues.
  20. pcrecord

    pcrecord Don't you want the best recording like I do ? Well-Known Member

    I had a signer/guitar player who once came to the studio asking to make the recording of him and add all the other instruments (as with a band)
    after 15min of hearing him strugle with the time of the first song.. I said, wait, if you want I can play and you'll save money. I recorded 2 full album for him overtime.. He would give me the chord progression and I'd do everything except I would sing at the end (The vocal made my eye bleeds but he was paying...)
    kmetal likes this.

Share This Page