I am reading Purcell's dialogue editing and it says to crossfade when I want to put two pieces of dialogue together. For example if I want to match syllables up to the mouths in ADR, or in one case, an actor's scream is too long so I want to shorten and blend with a grunt sound. There are three crossfades I have. Constant gain, constant power, and exponential fade. If I try any one of those, you can hear the transition though, and it doesn't sound natural. The actor's voice fades, and then it fades back, instead of sounding like it goes from one peace of dialogue to other. What am I doing wrong? Thanks.
editing What am I doing wrong in crossfading?
- Thread starter harmonica44
- Start date
The syllables need to have the same pitch, the same tonality, the same emphasis, etc. in order for the edit to work. Changing any of those things via audio suite effects may make the individual pieces sound artificial. So, first, all the pieces need to match. (BTW, you just may not have enough matching pieces to do what you want to do. GIGO - Garbage In, Garbage Out.)
You also need the pieces to overlap, sometimes a lot. Sometimes the fade needs to be short. Sometimes the fade needs to be long. Sometimes an "S" or other type of fade is needed.
This is not something that you are going to get right away. It takes lots of practice and "ears," that indefinable something that audio guys have and others don't. I hate to say it, but you may just not "get" audio. That's okay, I just don't "get" visuals.
You also need the pieces to overlap, sometimes a lot. Sometimes the fade needs to be short. Sometimes the fade needs to be long. Sometimes an "S" or other type of fade is needed.
This is not something that you are going to get right away. It takes lots of practice and "ears," that indefinable something that audio guys have and others don't. I hate to say it, but you may just not "get" audio. That's okay, I just don't "get" visuals.
Okay thanks I'll play around more. So far I got one to work, and it was a short one. I actually think I understand audio better than visuals so far. Where as I have a lot to learn on both, I am getting audio just a little more, at least compared to others who want to work on future projects.
The problem as I see it H44, both in this question and your question on hiring post people is that you are looking for formulas to solve problems and unfortunately, there are no formulas. In other words, you have to take each individual case on it's own and try and find a solution, there is no one way of doing it which always works.
In the case of your blending a scream into grunt, think about what happens to the sound when you yourself try to vocalise a scream changing into a grunt. You'll notice that it's either two different sounds separated by a very short gap or a sort of slide from the scream connecting to the grunt. In either case, a crossfade between the two won't emulate what a human sounds like and will therefore sound wrong. If the slide is what you are after, it might be possible depending on a number of factors but at best it's going to be extremely difficult and time consuming to get right. Far better and easier to re-record what you need in this instance.
G
In the case of your blending a scream into grunt, think about what happens to the sound when you yourself try to vocalise a scream changing into a grunt. You'll notice that it's either two different sounds separated by a very short gap or a sort of slide from the scream connecting to the grunt. In either case, a crossfade between the two won't emulate what a human sounds like and will therefore sound wrong. If the slide is what you are after, it might be possible depending on a number of factors but at best it's going to be extremely difficult and time consuming to get right. Far better and easier to re-record what you need in this instance.
G
I guess so, but two of the actors have moved now and are unavailable. I suppose I could just put the two together and not crossfade it and just leave it as is with a pop in between, since the pop still sounds more convincing than a crossfade so far. It is very tough for the actors to time the screams and grunts just right, especially as they were being thrown around which would change the sound of the grunt, during the process. After failing several times, and me failing to to teach them to do it exactly, they said they would be much better at making the sounds one by one, and for me to edit them all together. Next time I will tell the actors, that that's undo-able I guess.
Last edited: