Change the voice of several speech clips at once

This is the screen before any change. The four speech clips sit on the voice track named ava_1の音声.

1. 1Check that there are four green speech clips. The first one is selected in this screenshot.
2. 2The voice button at the bottom reads 明星レイカ. It shows the voice for lines you add from now on.
3. 3Check that the avatar clip shows Preview generated.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that there are four green speech clips. The first one is selected in this screenshot.
2. 2
   Check The voice button at the bottom reads 明星レイカ. It shows the voice for lines you add from now on.
3. 3
   Check Check that the avatar clip shows Preview generated.

Check The bulk change applies to speech clips created from text (voice type TTS). All four clips in this example qualify.

First, change only the last two clips. Put the playhead (the red vertical line) just before the third clip, then use the menu of the voice track header to select every clip to its right.

1. 1Before opening the menu, drag the playhead (the red vertical line) along the time ruler to a point inside the second clip (8.72 seconds in this example). The current time is shown at the lower left.
2. 2Right-click the ava_1の音声 header at the left. In this screenshot the opened menu hides part of the header.
3. 3Click the last item, Select clips to the right of the playhead on this track only. It shows “2 本” (two clips).

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Action Before opening the menu, drag the playhead (the red vertical line) along the time ruler to a point inside the second clip (8.72 seconds in this example). The current time is shown at the lower left.
2. 2
   Action Right-click the ava_1の音声 header at the left. In this screenshot the opened menu hides part of the header.
3. 3
   Action Click the last item, Select clips to the right of the playhead on this track only. It shows “2 本” (two clips).

Check The same menu has Select all clips on this track (Ctrl+A). Adding clips one by one with Ctrl+click is covered in the track-editing guide.

The third and fourth clips now have a selection outline.

1. 1Check that “選択中 2 本” (2 selected) appears at the top.
2. 2Only the third and fourth clips are outlined. The first two are not selected.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that “選択中 2 本” (2 selected) appears at the top.
2. 2
   Check Only the third and fourth clips are outlined. The first two are not selected.

Check To start over, press Esc to clear the selection, then select again. That Esc clears the selection was confirmed in the demonstration for the Keyboard shortcuts guide (it was not pressed in this project).

Operations for the selected clips are gathered in the Selection menu at the top.

1. 1Click 選択部処理 (Selection) at the top.
2. 2Click 声を変更…（2 件） (Change voice). The number in brackets is how many speech clips will change.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Action Click 選択部処理 (Selection) at the top.
2. 2
   Action Click 声を変更…（2 件） (Change voice). The number in brackets is how many speech clips will change.

Check As the menu says, changing the voice regenerates the audio and uses credits. Regenerate audio, just above it, is a different operation that keeps the voice and only redoes the reading.

The voice picker opens. Opened from here it lists the voices, so filtering by name makes the one you want easier to find. In this practice we choose “花純ナギサ”.

1. 1Type 花純 in the Filter by voice name field.
2. 2Click the remaining 花純ナギサ row to select it. The 試聴 (Listen) button at the right end of the row is for hearing a sample; it was not pressed in this demonstration.
3. 3Click この声にする (Use this voice) at the lower right.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Enter Type 花純 in the Filter by voice name field.
2. 2
   Action Click the remaining 花純ナギサ row to select it. The 試聴 (Listen) button at the right end of the row is for hearing a sample; it was not pressed in this demonstration.
3. 3
   Action Click この声にする (Use this voice) at the lower right.

Check Speaking style at the lower left stays at ナチュラル（ナレーション調B） in this example. Pressing the button does not change anything yet: a confirmation dialog comes next.

This is the confirmation dialog titled 声の変更. Nothing is regenerated until you run it here.

1. 1Check which clips change to which voice, and how many clips and seconds of audio will be regenerated. This example shows 2 clips, about 4.6 seconds.
2. 2Clip positions is set to 置き直す (Re-place), which shifts the following clips to match the new audio length. Leave it as it is here.
3. 3Click 変更して作り直す (Change and regenerate) at the lower right.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check which clips change to which voice, and how many clips and seconds of audio will be regenerated. This example shows 2 clips, about 4.6 seconds.
2. 2
   Check Clip positions is set to 置き直す (Re-place), which shifts the following clips to match the new audio length. Leave it as it is here.
3. 3
   Action Click 変更して作り直す (Change and regenerate) at the lower right.

Check The dialog notes that the audio length depends on the voice and that Undo (Ctrl+Z) restores the previous state. The middle section estimates the avatar clips whose video you will regenerate afterwards. Choose やめる to cancel.

When you run it, the audio is regenerated one clip at a time.

1. 1The display shows which clip is being regenerated and its line. Here it is the first of two.
2. 2A 中止 (Stop) control is provided in case you need to stop. It was not used in this demonstration.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check The display shows which clip is being regenerated and its line. Here it is the first of two.
2. 2
   Check A 中止 (Stop) control is provided in case you need to stop. It was not used in this demonstration.

Check In this demonstration the two clips finished while we watched the screen (the time was not measured). The wait depends on the number of clips, their length and how busy the service is.

This is the screen after the regeneration. Only the two changed clips have new audio.

1. 1The third and fourth clips changed length slightly: in this example the third went from 2.02 to 2.00 seconds of audio and the fourth from 2.59 to 2.74 seconds.
2. 2The avatar clip now says to generate the preview with Generate pending. Because the voice changed, the lip-sync video has to be regenerated.
3. 3Generate pending in the toolbar shows 1. Generate at the end, once the voice is settled.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check The third and fourth clips changed length slightly: in this example the third went from 2.02 to 2.00 seconds of audio and the fourth from 2.59 to 2.74 seconds.
2. 2
   Check The avatar clip now says to generate the preview with Generate pending. Because the voice changed, the lip-sync video has to be regenerated.
3. 3
   Check Generate pending in the toolbar shows 1. Generate at the end, once the voice is settled.

Check Because Re-place was chosen, the fourth clip moved to follow the new end of the third, and the avatar clip and the background were extended automatically from 19.00 to 19.11 seconds. The first two clips did not change.

To confirm a change, select a speech clip and look at the panel on the right.

1. 1Click the third speech clip to select it.
2. 2Check that Voice on the right reads 花純ナギサ. If you select the first or second clip, it still reads 明星レイカ.
3. 3The voice button at the bottom still reads 明星レイカ. It is not the voice of existing clips but of the next line you add.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Action Click the third speech clip to select it.
2. 2
   Check Check that Voice on the right reads 花純ナギサ. If you select the first or second clip, it still reads 明星レイカ.
3. 3
   Check The voice button at the bottom still reads 明星レイカ. It is not the voice of existing clips but of the next line you add.

Check Listen where you can play sound to compare the voices. The Play button under Audio on the right is for playing the audio of the selected line; it was not pressed in this demonstration.

However many clips you changed together, one Undo restores them. Try it here.

1. 1Click 編集 (Edit) at the top.
2. 2Click 元に戻す (Undo). Ctrl+Z does the same.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Action Click 編集 (Edit) at the top.
2. 2
   Action Click 元に戻す (Undo). Ctrl+Z does the same.

Check If you change your mind after undoing, Redo (Ctrl+Y) is just below. Redo was not used in this demonstration.

This is the screen after one Undo.

1. 1The third and fourth clips are back to their original length and position.
2. 2The avatar clip shows Preview generated again. No regeneration is needed.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check The third and fourth clips are back to their original length and position.
2. 2
   Check The avatar clip shows Preview generated again. No regeneration is needed.

Check In this demonstration a single Undo restored the voice, the previous audio, the clip positions and the generated state of the avatar clip together.

This time, skip the selection and change all four clips on the voice track.

1. 1Right-click the ava_1の音声 header at the left.
2. 2Click このトラックの声を変える…（4 件） (Change the voice of this track).

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Action Right-click the ava_1の音声 header at the left.
2. 2
   Action Click このトラックの声を変える…（4 件） (Change the voice of this track).

Check In the voice picker that follows, the row of the current voice (明星レイカ) is selected. As before, filter by 花純, choose 花純ナギサ and press Use this voice.

The confirmation dialog now refers to the four clips on ava_1の音声. This time try the other choice under Clip positions.

1. 1Check that four clips are targeted and that 4 clips, about 13 seconds of audio will be regenerated.
2. 2Click 動かさない (Keep) to select it. Clips keep their start positions and only the audio is replaced.
3. 3Click 変更して作り直す (Change and regenerate).

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that four clips are targeted and that 4 clips, about 13 seconds of audio will be regenerated.
2. 2
   Action Click 動かさない (Keep) to select it. Clips keep their start positions and only the audio is replaced.
3. 3
   Action Click 変更して作り直す (Change and regenerate).

Check When it finishes, a notice at the lower right says the four clips were changed and regenerated. In this demonstration all four kept their start positions and only their lengths changed. Because the new audio was slightly longer, the end of a clip passed the start of the next by up to about 0.08 seconds (an overlap of silent padding that is barely visible). Keep suits timing that must stay aligned with captions or visuals; Re-place suits lines that should follow one another closely.

This example is finished with 明星レイカ, the first recommended voice. Use the same steps to return the four clips to it: right-click the header and choose Change the voice of this track to open this screen.

1. 1Type 明星 in the filter field.
2. 2Click the 明星レイカ row to select it.
3. 3Click この声にする (Use this voice).

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Enter Type 明星 in the filter field.
2. 2
   Action Click the 明星レイカ row to select it.
3. 3
   Action Click この声にする (Use this voice).

Check In the confirmation dialog that follows, Clip positions is back to Re-place: the Keep choice from last time is not remembered. In this demonstration we ran it with Re-place, and a notice reported that five following clips were re-placed to match the new audio. Even with the same voice and text, the regenerated audio differed slightly in length from the first version (the fourth clip went from 2.59 to 2.74 seconds).

The small overlaps created by Keep remained after returning the voice with Re-place. Line the speech clips up again in one step. With the voice track selected, press Ctrl+A to select all four.

1. 1Check that “選択中 4 本” (4 selected) is shown.
2. 2Click 選択部処理 (Selection).
3. 3Click タイミング: 間を詰めて並べる (Timing: pack). It lines the clips up in start-time order with no gaps and no overlaps.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that “選択中 4 本” (4 selected) is shown.
2. 2
   Action Click 選択部処理 (Selection).
3. 3
   Action Click タイミング: 間を詰めて並べる (Timing: pack). It lines the clips up in start-time order with no gaps and no overlaps.

Check Timing: arrange with a set gap, just below, lines clips up with a fixed number of seconds between them. It is not used in this guide.

The four clips now follow one another without a break. With the voice settled, regenerate the avatar video.

1. 1Check that the speech clips run from 0.40 to 18.77 seconds with no gaps and no overlaps.
2. 2Click Generate pending in the toolbar. The avatar video is regenerated to match the new audio.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that the speech clips run from 0.40 to 18.77 seconds with no gaps and no overlaps.
2. 2
   Action Click Generate pending in the toolbar. The avatar video is regenerated to match the new audio.

Check Generation uses credits. If you change the voice several times, generate once after the voice is settled rather than after every change.

When generation finishes, play from the beginning. The screenshot shows the playback in progress.

1. 1Check that the avatar clip shows Preview generated.
2. 2Click the Play button at the left. It turns into the Pause mark during playback.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that the avatar clip shows Preview generated.
2. 2
   Action Click the Play button at the left. It turns into the Pause mark during playback.

Check Check that the lips match the lines and that the pauses between lines feel natural. Check how the voice sounds where you can play sound. In this demonstration the 19-second project played to the end.

Finally, reopen the saved project and check it. The screenshot shows the fourth clip selected after reopening.

1. 1Check that Voice on the right reads 明星レイカ. Select the other three clips in turn and check them too.
2. 2Look at the voice button at the bottom as well. If it shows a voice you did not intend, press it and choose again.
3. 3Check that Saved (auto-save) is shown at the lower right.

↘Red arrows and white labels are guide annotations ＋ Enlarge with callouts

1. 1
   Check Check that Voice on the right reads 明星レイカ. Select the other three clips in turn and check them too.
2. 2
   Check Look at the voice button at the bottom as well. If it shows a voice you did not intend, press it and choose again.
3. 3
   Check Check that Saved (auto-save) is shown at the lower right.

Check In this demonstration the bottom voice button showed 花純ナギサ, the voice used on the way, right after reopening (the four speech clips still used 明星レイカ). This screenshot was taken after choosing 明星レイカ, the first entry under Recommended for Sachiyo, from that button again. Glance at the bottom voice before adding more lines.