How do you use reference tracks or spectral analysis to EQ match ADR to production sound?

When you record additional dialogue replacement (ADR) for a film or TV show, you want it to sound as natural and consistent as possible with the original production sound. But how do you achieve that? In this article, we'll explore two common techniques for EQ matching ADR to production sound: using reference tracks and spectral analysis.

1 Reference tracks

A reference track is a segment of the production sound that you use as a model for the ADR. You can use a reference track to compare the frequency response, tonal balance, and dynamics of the ADR with the production sound. To do this, you need a software tool that can measure and display the frequency spectrum of both tracks, such as a spectrum analyzer or an EQ plugin with a spectrum view. You can then adjust the EQ of the ADR track until it matches the reference track as closely as possible. You may also need to adjust the volume, compression, and reverb of the ADR track to match the reference track.

Add your perspective

Jeremy Lee - CTS

Senior Technology Consultant at Electrosonic
Report contribution
There’s no AI writing algorithm that will really understand what it takes to do this work. It’s just regurgitating press releases and generalities.

Like
Martin Rieger

Make a statement with 3D audio - before everyone else does | Brand and Product Experiences, immersive audio experiences with storytelling
Report contribution
Match-EQs helfen schon mal ein großes Stück. Kombiniert mit dem Ambience match kommt man schon recht nah ran. Wenn dann noch der Reverb angepasst wird, ist man auf einem guten Weg. Oder man kauft sich von iZotope Dialogue Match, was genau diese Faktoren beinhaltet und Zeit sparen kann. Dennoch hatte ich noch nie einen Fall wo es sofort genau funktioniert hat, man muss immer etwas feintunen und sich auf ein gutes Gehört verlassen

Translated

Like

2 Spectral analysis

Spectral analysis is a technique that uses a mathematical algorithm to analyze the frequency spectrum of a sound and generate an EQ curve that matches it. You can use spectral analysis to EQ match ADR to production sound without using a reference track. To do this, you need a software tool that can perform spectral analysis and apply the resulting EQ curve to another track, such as iZotope RX or FabFilter Pro-Q. You can then select a segment of the production sound that contains the dialogue you want to match, run the spectral analysis on it, and apply the EQ curve to the ADR track. You may also need to fine-tune the EQ curve and adjust the volume, compression, and reverb of the ADR track to match the production sound.

Add your perspective

David Abrahamsen, CAS

Re-Recording Mixer SSI Advanced Post
Report contribution
I do not agree that we should be handcuffed to algorithms only. Part of the art of good ADR is knowing what mics you have and their own sonic character, what color to paint with. Last it’s all feel, to the ear and the eye. I trust my ears and feel, does a computer feel? Uh, no.

Like
Martin Rieger

Make a statement with 3D audio - before everyone else does | Brand and Product Experiences, immersive audio experiences with storytelling
Report contribution
Bevor man sich in der Post-Produktion wund EQt ist es wichtiger schon vor dem ADR Prozess einfach das passende Mikrofon zu nutzen. Am besten eben genau das, was auch beim Originalton verwendet wurde. Noch besser ist, wenn man sogar zur originalen Location reisen kann. Aber meistens passiert die ADR natürlich im Studio und die Hintergrundgeräusche und Reverb müssen später hinzugefügt werden. Damit ist man zwar nicht so schnell, aber auch flexibler

Translated

Like

3 Benefits and limitations

EQ matching techniques for ADR can help you achieve a more seamless and realistic dialogue mix for your audio post production project. By using reference tracks or spectral analysis, you can reduce the sonic differences between the ADR and the production sound, and create a more uniform and coherent sound field. However, EQ matching techniques also have some limitations and challenges. For example, you may not always have a suitable reference track for the ADR, especially if the production sound is noisy, distorted, or has a lot of background sounds. Also, spectral analysis may not capture all the nuances and variations of the human voice, and may introduce unwanted artifacts or resonances to the ADR.

Add your perspective

Martin Rieger

Make a statement with 3D audio - before everyone else does | Brand and Product Experiences, immersive audio experiences with storytelling
Report contribution
Man darf sich nicht ganz auf den EQ verlassen. Wie erwähnt spielen die Ambience und Reverb auch eine große Rolle. Außerdem misst der EQ nur das Spektrum über eine gewisse Zeit. Die Dynamik kann aber eine andere sein, oder v.a. spielt die Performance auch eine große Rolle der Glaubwürdigkeit. Oft lohnt es sich nach dem Match EQ noch einen weiteren EQ zu nutzen mit eher breiten Einstellungen um noch einmal manuelle Anpassungen zu machen

Translated

Like

4 Best practices and tips

To get the best results from EQ matching techniques for ADR, there are some best practices and tips to keep in mind. Make sure to use high-quality recordings for both the ADR and the production sound, free of clipping, distortion, or noise. Choose reference tracks that match the scene, location, and mood of the ADR, avoiding those with different speakers, accents, or emotions. Utilize spectral analysis on short and clear segments of the production sound that only contain the dialogue you want to match. When applying EQ to the ADR, use a gentle and transparent curve that avoids boosting or cutting too much. Compare the ADR and production sound in mono and stereo, as well as in different speakers and headphones for any phase issues, frequency clashes, or volume differences. Ultimately, use your ears and your judgment when EQ matching; don't rely solely on visual feedback or numerical values.

Add your perspective

Martin Rieger

Make a statement with 3D audio - before everyone else does | Brand and Product Experiences, immersive audio experiences with storytelling
Report contribution
In Zukunft kann das AI Voice-Replacement interessant werden. Ich hatte schon ADR aufnehmen, weil ein Wort falsch ausgesprochen wurde, oder vom Sinn angepasst werden musste. Ich sehe hier große Möglichkeiten das mit einer KI - natürlich in Absprache mit den Beteiligten - zu reparieren. Auch sind Sprachfassungen für KI gefundenes fressen. Für eher monotone Audioinhalte klappt es schon recht gut. Für emotionale, fiktionale Spielfilminhalte wird aber ADR noch eine gute Weile relevant sein

Translated

Like

How do you use reference tracks or spectral analysis to EQ match ADR to production sound?

1

2

3

4

1 Reference tracks

2 Spectral analysis

3 Benefits and limitations

4 Best practices and tips

Audio Post Production

Rate this article

Thanks for your feedback

More articles on Audio Post Production

More relevant reading