How to equalize two different recordings to sound similar?

[zee944]

How to equalize two different recordings to sound similar?

What I really have to do is to mix two movie audio tracks into one: I have the original audio and a dubbed as well (it's a foreign movie). I have to put in parts from the original track to the dubbed track (due to missing scenes in the latter), and sometimes vice versa - from the dubbed into the original (I have no room here to explain why). The dub sounds very different from the original; muffled, the high pitched sounds are hollow, bass sometimes thin etc. despite the fact they came from the same source when the movie was new.

An example: a piece of music was missing from the dubbing. I took that piece from the original and pasted in (of course with some fade in-fade out trick on the edges), but sounded way off, incongruous! Then I took the equalizer and played with it around a LOT, half an hour for an 5 second long part! In the end, it became almost perfect - sounded very much like the dub. Wasn't annoying anymore.

Is there an audio software which can do this automatically if I ask for it? I've checked, tried, searched a lot... nothing yet. Please don't say I have to do it all by myself. I don't want to... There must be a solution - an function which compares the EQ graps and tries to adjust one of them to have the same amount of bass, high pitches etc.

I'd appreciate any help. Thank you!

[jagabo]

Did you keep the settings for the short part you equalized successfully? The same settings will probably work for the rest of the film.

[zee944]

It doesn't work for the whole movie. It's very rare when the same settings are fine, in a new scene almost always have to adjust again. Besides, I have more than one movie to do.

[Soopafresh]

http://www.elevayta.net/product3.htm

And

Voxengo CurveEQ implements SpectruMatch spectrum matching technology that allows you to perfectly transfer a spectrum's slope from one recording to another. This allows you to `copy' a frequency balance of existing mixes so that other mixes that are still in the works will sound better. This technology also greatly helps in music disc mastering, since using it allows you to easily lessen any dramatic differences in the area of frequency balance between various tracks.

http://www.voxengo.com/product/curveeq/

[zee944]

Hey, these seem real good stuff! If something, these should do the job.

Thanks man!

[zee944]

A lot of time passed (had to do work).

Finally I've tried out Voxengo CurveEQ. It is indeed capable of match tones of different recording. It works in his own way quite well, but something important is missing. Elevayta FreEqBoy seem to work like the same.
Getting back to my example in my original post:

An example: a piece of music was missing from the dubbing. I took that piece from the original and pasted in (of course with some fade in-fade out trick on the edges), but sounded way off, incongruous! Then I took the equalizer and played with it around a LOT, half an hour for an 5 second long part! In the end, it became almost perfect - sounded very much like the dub. Wasn't annoying anymore.

In this case above, this can't be properly done with Voxengo CurveEq or Elevayta FreEqBoy. They can save spectrums and match another record's spectrum to it, but when you don't have the very same piece of the sound in both recordings it becames a guessing game. You can save a spectrum of a similar piece and use that, but what's the solution when you don't really know what to choose as a reference for a missing part? I think the solution would be to calculate the difference between two spectrums (how one should be equalized to sound similar to the other). After I have the difference saved into a file, I can use it to adjust the tone any time. Is this possible?

[Soopafresh]

You can save a spectrum of a similar piece and use that

Exactly. The pieces must be similar in timbre, or tone quality and in energy.

That's the biggest weakness of frequency matching.


Here's a pretty good article on it. Read about convolution and impulse responses, where you can capture and apply the character of a sound. But you can't completely mask the character of the file you're applying the impulse response to.

http://www.soundonsound.com/sos/apr05/articles/impulse.htm

And plugin http://www.knufinke.de/sir/sir1.html


You can also ask this question on KVRaudio's website, which specializes in audio plugin questions.

http://www.kvraudio.com/forum/

[zee944]

Thanks for the article, I don't understand the half of it (my English is not that good), but understanding the problem in details wouldn't solve my problem anyway.

On the KVRaudio forum they couldn't help me.

What about a Spectrum viewer and an Equalizer built in one? So I could see how would the Equalizer sliders modify my spectrum curve. Is there a plugin or audio software where I can see real-time how my spectrum will look if I apply the Equalizer settings?