> For the full page index, fetch https://echonova.ai/llms.txt

# Filters, previews and favourites

Narrow the voice library by search, dialect and voice traits, use the preview player, and keep a list of saved voices that follows you into every picker.

The library gets long. By the end of this page you will be able to cut it to
the voices that fit, hear each one through the player, and save a
shortlist that shows up again wherever you pick a voice.

## What you need

- The voice library open: **Voices** in the sidebar. See
  [Browse the voice library](/docs/playground/voice-library) for the page
  layout.

## Steps

### Step 1: Search by name or character

Type in the box that reads **Search a name, a tone or a use**. The rows
narrow as you type. A search matches a voice's name, its description, its
dialect, gender, age, tone and pace, and its uses, in any case.

On **My voices** the box reads **Search my voices** and searches only the
voices you have added. Each tab keeps its own search.

### Step 2: Filter by dialect

Under the search box is a row of chips, one per dialect, each with its voice
count. Dialects that have no voices yet follow with the word **soon** and
cannot be pressed.

Press a chip to keep only that dialect. Press it again to clear it. One
dialect is active at a time.

When the use buttons appear under the dialect chips, press one, such as
**Narration and audiobooks**, to keep only the voices recorded for that use.
Press it again to clear it.

<!-- screenshot: filters-previews-favourites/1 the dialect chip row with one chip active -->

### Step 3: Filter by gender, age, tone or pace

Press **Filters**. A drawer opens with a row for each of **Gender**, **Age**,
**Tone** and **Pace**. A row lists only the values that at least one voice
has, for example **Female** and **Male** under **Gender**, or **Warm**,
**Calm** and **Serious** under **Tone**. Press a value to keep only the voices
with it, and press it again to clear it. **Clear all** removes every filter.

**Sort** next to it orders the list by **Newest**, **Name** or **Most used**.

The voice picker that opens from **Text to Speech** has its own **Type** menu
with **All types**, **Premade** and **Cloned**, plus the dialect chips.

### Step 4: Play a preview

Press the orb or the play control on a row or on a **Most used** card. The
player opens along the bottom of the page and the clip plays.

| Control                | What it does                                                                                 |
| ---------------------- | -------------------------------------------------------------------------------------------- |
| The orb on the row     | Pauses and resumes this voice                                                                |
| The seek slider        | Move through the clip. The time beside it reads elapsed and total, for example `0:03 / 0:07` |
| **Play** and **Pause** | The play buttons on the player                                                               |
| **Stop**               | Stops the clip                                                                               |
| **Download**           | Saves the preview clip                                                                       |
| **Close**              | Closes the player and stops the clip                                                         |

One clip plays at a time; starting another voice replaces it. Leaving the page
stops it.

<!-- screenshot: filters-previews-favourites/2 the player with a clip half way through -->

### Step 5: Save a voice

Press the **+** button on a row or a card, or **Add to My voices** in the
voice's details panel. The button changes to **In My voices**. Press it again,
or **Remove from My voices**, to take the voice off your list.

On **My voices**, removing a voice takes its row out of the list.

### Step 6: Use your saved voices

Open **My voices**. The voices you added are listed under **Added from the
library**, with the same row actions as **Library**. Voices you cloned are
listed below them.

When the list is empty the page reads **Nothing here yet.** and **Voices you
keep from the library land here.**

Your saved voices also appear under the **My Saved Voices** tab of the voice
picker, both on **Text to Speech** and when you press **+ Add voice** on an
agent's **Voice** tab.

Saved voices belong to you, not to the workspace.

<!-- screenshot: filters-previews-favourites/3 the My voices tab with a few saved rows -->

## Get it right

Save the voices you have compared, not every voice you have played. The saved
list is the shortlist behind the picker's **My Saved Voices** tab, so it works
when it is short.

**Recommended**

- Play the Egyptian voices, save the two that fit the brand, then pick between
  them in the Text to Speech picker.

**Less effective**

- Save every voice in the library so that **My voices** becomes a second
  copy of **Library**.

## Worked example

You need a Saudi voice for a product video. Press the **Saudi** chip, and the
rows show only Saudi voices. Press each orb and listen through the player,
dragging the seek slider to hear the end of each clip.

Two are close. On each row, press the **+** button. Open **My voices**: both
are listed. Open **Text to Speech**, press **All voices**, and the picker's **My Saved Voices** tab
holds just those two. Generate the same paragraph with each and keep the one
that reads better.

## What happens next

A saved voice stays on your list until you remove it, or until the voice is
withdrawn from the library. A voice that is withdrawn drops off **My voices**
and out of every picker.

Filters and search apply to the page you are on and are cleared when you
switch tabs. Nothing you filter, play or save is charged.

## Limits

| Limit                        | Value                   | Applies to                    |
| ---------------------------- | ----------------------- | ----------------------------- |
| Saved voices per person      | No limit                | **My voices**                 |
| Dialect chips active at once | 1                       | **Library**                   |
| Voices a search covers       | The 100 the page loaded | **Library** and **My voices** |

## Troubleshooting

### No voice matches.

Your search and the active filters leave nothing. Press **Clear the
filters**, or clear the search box and release the active chips.

### The orb does nothing

That voice has no preview clip. Choose another voice, or press **Use in Text
to Speech** and generate a sentence on **Text to Speech** to hear it on your
own text.

### We couldn't load the voices.

The list did not load. Press **Try again**.

## FAQ

### Can I filter by gender or language?

You can filter by gender, age, tone, pace and use, and by dialect; see
[Step 3](#step-3-filter-by-gender-age-tone-or-pace). Each voice in the Developer
API's voices call also carries a `gender` and a `language` field you can filter
on in your own code; see
[Text to Speech voices and options](/docs/developers/tts-voices-and-options).

### Are saved voices shared?

No. A saved list belongs to you, not to the workspace.

### Why did a voice vanish from My voices?

The voice was withdrawn from the library. Only voices that are in the
library today are listed, so it will not appear again unless it returns.

### Is the preview the same quality I get when I generate speech?

The preview is a fixed sample clip. To hear a voice on your own text, press
**Use in Text to Speech** and generate; that is charged in credits.

## Related

- [Browse the voice library](/docs/playground/voice-library): the page layout, the row actions and adding a voice to an agent
- [Dialects](/docs/playground/dialects): what the dialect chips mean and how to pick one for a region
- [Choosing a voice](/docs/playground/choosing-a-voice): what to listen for in a preview
- [Voices and language](/docs/agents/voice-and-language): the agent's voice picker and per-voice tuning
- [Text to Speech voices and options](/docs/developers/tts-voices-and-options): the voice fields the API returns