Jump to content

FL speech - weird delay I can't get rid of


Recommended Posts

when i open up a FL speech thing in the step sequencer, there is for some reason a weird delay (sounds like a 1-note and 4-note delay) on it that isn't there in the window where you type the words/select voice/preview how it sounds. how do I get rid of this?

also, something else that is not so problematic, but I'm just curious about: when the speech i've typed/chosen opens up in the FL slicer, it shows that it is at a different BPM than the one i specified. weird eh? so far it's been like 20 or more BPM higher. not a huge deal, reasonably easy to offset. but it's the delay that I need to get rid of.

Link to comment
Share on other sites

Question 1:

Are you referring to how long it takes a word from onset for the actual "start" of the word? You're gonna have to start speech synth sounds earlier than when you want them to 'sound' pretty much regardless, except for really high rate words/phrases.

Question 2:

I could be wrong, but I don't believe that number refers to a BPM, but rather just some arbitrary 'rate' which refers to its speed. It can't very well be considered a BPM considering there's no info being inputted as far as the rhythm of the words.

Link to comment
Share on other sites

I mean delay in the sense in which it is usually used in the context of music, as in an echoing effect.

I'm sure it is the BPM, as it is displayed that way on the slicer window. Here I have included a JPG of the window, showing that it is actually 86BPM higher than the BPM I selected when creating the speech (which was 200BPM). Also I'm including a short (an very low-quality) MP3 to demonstrate the delay effect. The words are "I am talking."

screencap (56k friendly, 33kb):

http://www.geocities.com/warpspeedwonder/flspeechproblempic.JPG

mp3 sample (also 56k friendly, 7kb):

http://www.geocities.com/warpspeedwonder/flspeechproblemsample.mp3

Incidentally I just found out that by adjusting the number of beats (whatever the hell that means) in the slicer window to 2 or less, I can remove this delay. However that also adjusts the playback rate.

Link to comment
Share on other sites

I'm not sure where the delay effect is coming from. It happens even when there is no slicer involved? It happens when that's the first thing you do, on a blank project? Because that's weird.

I can explain your second question better, maybe. When you created the speech sample and specified a "rate", you were entering a BPM at which you want the speech synth to talk--but it doesn't correspond to one word per beat or anything--ex, when i was testing, "hello" took one beat but "a" was considerably shorter.

The slicer takes the raw waveform output of the speech synth and tries to make sense of it. Given that the slicer was designed with mainly drum loops in mind, plus the fact that the speech sample is irregular, it has no way of determining what the original BPM you wanted was.

What's the solution? Well you're probably going to have to fiddle around with the BPM/beats numbers, as well as the Fit length slider (the one above the "Fill" in "fill gaps"). How you do it depends largely on what you're trying to do in the end.

More importantly, I'd recommend switching the Stretch mode (the four buttons labeled A-D) to the A or B mode, since that seems to be much better adapted to speech samples.

Link to comment
Share on other sites

I'm sure it is the BPM, as it is displayed that way on the slicer window. Here I have included a JPG of the window, showing that it is actually 86BPM higher than the BPM I selected when creating the speech (which was 200BPM). Also I'm including a short (an very low-quality) MP3 to demonstrate the delay effect. The words are "I am talking."

screencap (56k friendly, 33kb):

http://www.geocities.com/warpspeedwonder/flspeechproblempic.JPG

No, I mean the number in the speechsynth window isn't BPM.

But try turning of the stretching, see if that fixes your echo problem.

Link to comment
Share on other sites

I just discovered that it will always open in a slicer when I open the speech in a new channel. If I open up a sampler, then open the speech into that existing channel, it remains a simple sampler and there is no delay. So I suppose if I want to slice the speech up I will need to use a slicer, but I can set it to 2 beats or less and then slice, seeing as how since I am slicing the overall playback rate of the speech won't be affected.

Thanks people, problem solved. It just seems weird that one should have to go to such great lengths to get it to sound normal/default. Oh well, FL isn't the greatest program (and yet, what is?).

Link to comment
Share on other sites

I don't know what you mean by stretching. I can't find a 'stretching' toggle in the speechsynth's slicer window.

Your slicer looks ever so slightly different from mine.. but in FL7 that second fader from the right (next to PS) is labeled TS for time stretch.

Depending on how many words you're going to use and how much tedium you can handle, it's much more rhythmically accurate to do each word individually as a sampler, and time stretch each one (the knob will often snap to the perfect position).

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...