Insert pause or silence

[lacorrep]

Is there any way to add pauses between sounds when writing them to "fp"?..

Here is my ugly trick until then:

pause = gTTS("¤", lang='fr') # character that isn't pronounced, thus resulting in a silence

with open('french_japanese.mp3', 'wb') as f:
	for i, w in enumerate(words):

		gTTS(w[0], lang='fr').write_to_fp(f)

		pause.write_to_fp(f)
		pause.write_to_fp(f)
		pause.write_to_fp(f)
		pause.write_to_fp(f)

		gTTS(w[1], lang='ja').write_to_fp(f)

		pause.write_to_fp(f)
		pause.write_to_fp(f)
		pause.write_to_fp(f)
		pause.write_to_fp(f)
		pause.write_to_fp(f)

This is very slow.

[lacorrep]

Since I have a very long list of words to record, I had to improve performance to the detriment of beauty... I hope someone finds this helpful.

from gtts import gTTS
import pandas as pd

# --------------------------------

## Create binary code for a short silence
# Ugly trick: utter a character that gTTS doesn't pronounce, thus resulting in a silence
# This is better than doing "pause.write_to_fp(f)" (it's too slow)
pause = gTTS("¤", lang='fr')
# Write the pause in a file
with open('test.mp3', 'wb') as f:
	pause.write_to_fp(f)
# Save the binary corresponding to a short pause
with open('test.mp3', 'rb') as f:
	pause_binary = f.read()

# --------------------------------

# Load words from external file
df = pd.read_csv('liste_vocab.tsv',sep="\t",names=['fr','ja'])

# --------------------------------

# Utter and create MP3
with open('french_japanese.mp3', 'wb') as f:
	for i, w in enumerate(df.values):

		gTTS(w[0], lang='fr').write_to_fp(f)#.save(f'fr_{i}.mp3')

		# apply 5 short pauses
		f.write(pause_binary*5)

		gTTS(w[1], lang='ja').write_to_fp(f)#.save(f'ja_{i}.mp3')

		f.write(pause_binary*6)

[omega3]

Can you tell me how to apply silence to text file that looks like a normal text, not csv? For example a book, or article where there is sentence after sentence, only separated by "." Like:

Financial anxiety can also get worse around times of financial crises, such as job loss or unexpected bills. It can even lead to other dysfunctional behavior patterns that further entrench difficulties with money. He says some people turn to gambling, for instance, to find another income source.

Should I prepare csv first dividing text into rows?
What if I don't have pair fr-jp, but only single language: english?

That silence between sentences would be useful for language learning method called shadowing.

[lacorrep]

gTTS will automatically make small pauses between sentences. If this is not enough for you, you can divide your text using:

from gtts import gTTS
my_text = """Financial anxiety can also get worse around times of financial crises, such as job loss or unexpected bills. It can even lead to other dysfunctional behavior patterns that further entrench difficulties with money. He says some people turn to gambling, for instance, to find another income source."""
my_sentences = my_text.split('. ') # creates an array ['Financial ... bills', 'It can... money', 

# ##############################################################
# Option 1) loop over this array to export 1 file per sentence
for i, sentence in enumerate(my_sentences):
    gTTS(sentence , lang='en').save(f'my_text_{i:0>4}.mp3')

# ##############################################################
# Option 2) loop over this array to export a single file with pauses between sentences

# --------------------------------

## Create binary code for a short silence
# Ugly trick: utter a character that gTTS doesn't pronounce, thus resulting in a silence
# This is better than doing "pause.write_to_fp(f)" (it's too slow)
pause = gTTS("¤", lang='fr')
# Write the pause in a file
with open('test.mp3', 'wb') as f:
	pause.write_to_fp(f)
# Save the binary corresponding to a short pause
with open('test.mp3', 'rb') as f:
	pause_binary = f.read()

# --------------------------------

with open('my_long_text.mp3', 'wb') as f:
	for sentence in my_sentences:
		gTTS(sentence , lang='en').write_to_fp(f)

		# write 5 short pauses
		f.write(pause_binary*5)

Language selection: you should be able to do this by yourself. Note the use of lang='en'. This is what I have used in my answer to your other question in #432.

Since you said that you "don't have any knowledge of programming at all", I suggest that you learn the basics of Python programming of get help from AI to tackle this small project.

[omega3]

Thank you very much.

[splash58]

I ran into this problem and found another approach - generate a set of silent audio frames of the desired length. there you can get the code - https://github.com/splash58/silence