Issue with play voice and queue voice with translation

[ComputerArt.Club]

Hi all,

I am trying to fix a bug in a translation of a game, it is a label that is called to frequently that plays a voice file.
Unfortunately, when the voiced part is just coded as:

Code: Select all

voice "filename.ogg"

It shows up in the translation file but it doesn't play for either the original or the translated game.

When it is code as

Code: Select all

queue voice "filename.ogg"

It plays the file in the original, but it isn't being included in the generated translation file.

When I code it as:

Code: Select all

play voice "filename.ogg"

I have the same issue. It plays in the original, but it isn't being included in the translation file.

Right before the screen is called there is a sound effect on the sound channel. So that could some how be interfering with the first example, though I am not sure why play voice is successful when voice is not.

As this is a screen, there is no accompanying written "speech" for this line so I don't know how to get the line identifier code to try to direct Renpy to this by myself (it doesn't get exported in the extract dialog operation, as it is not dialog).

All the other voice files are playing just fine, even ones with no text that call up screens. I guess the difference here is the sound file being played right before the voice file.
The sound file is being called like this:

Code: Select all

play sound "sfx.mp3"
call win

win is the label with the relevant voice file.

Any suggestions?
Thanks for your help.

[ComputerArt.Club]

ugly ugly ugly workaround:
Main code:

Code: Select all

label win:
    queue voice "well done.ogg"
    voice "well done.ogg"

The second voice one doesn't get played in the English version for whatever reason started this whole problem (the preceding play sound file command perhaps?), but is translated and passed onto Japanese.

Translation code:

Code: Select all

# game/script.rpy:682
translate japanese win_1817be10:

    # voice "well done.ogg"
    play voice "jwelldone.ogg"

From there, play voice interrupts and stops the English one that was queued to play and so it plays the Japanese instead.
SO HORRIBLE. This feels very very broken and I anticipate problems further down the line.
Anyone got a more elegant solution?

[ComputerArt.Club]

Something that would help address this same issue, there are also a few images that I would ideally change for translations because they are text based. Perhaps if I could get identifier codes for every line I would be able to replace those images/lines of code with:

Code: Select all

translate japanese labelname_IDCODE:

Is there a way to find the ID code for any particular line?

[philat]

Hmm. Might be a bug, honestly, but it seems to work fine if you add _() around the string: e.g., voice _("blahblah.ogg")

[Remix]

Code: Select all

show screen _auto_voice

(recent version of Ren'Py if possible)

Will give you the line identifier to dialogue only. Not sure how you would get one for a line that isn't dialogue as you'd need to hook it into the running code that happens while on that line.

You might also take a look at Ren'Py's auto_voice feature, which would allow you to automate all voice audio and allow you to gear the file names to include a translation folder or component...

Code: Select all

default auto_voice_file_string = "auto_voice/{language}/{identifier}.ogg"

init python:

    def get_auto_voice_file_name(identifier):

        # Amend the dict {language name : folder name} to suit 
        
        vdict = {
            'language' : {
                None : 'en', 
                'Spanish' : 'es',
                'Nihongo' : 'jp',
            }.get(_preferences.language, 'unknown'),
            'identifier' : identifier,
        }

        return auto_voice_file_string.format(**vdict)

    # Tell Ren'py to *try* to voice every line
    # using the above function to locate file
    config.auto_voice = get_auto_voice_file_name

label start:

    show screen _auto_voice

    "..." # This line is voiced if Ren'Py can find a file such as "auto_voice/en/start_a20cefa7.ogg" as shown in the screen

config.auto_voice can be a string, function or None ( https://renpy.org/doc/html/config.html# ... auto_voice )

If you do decide to go for auto_voice, I have a voice artist helper script that might be of use. Send me a PM if you want to try it (it basically reads the entire game and outputs all the lines and related filenames that each character should speak).

[ComputerArt.Club]

Hmm. Might be a bug, honestly, but it seems to work fine if you add _() around the string: e.g., voice _("blahblah.ogg")

Thanks for your time and help and thanks for the tip about the _(), indeed I think I had issues with screens before when I didn't add brackets.
Right now the plain old voice tag is not playing for that instance, tried with brackets, I will try again with underscore soon, but for some reason it seems to want 'play' or 'queue' to play this voice file.

Code: Select all

show screen _auto_voice

(recent version of Ren'Py if possible)

Will give you the line identifier to dialogue only. Not sure how you would get one for a line that isn't dialogue as you'd need to hook it into the running code that happens while on that line.

You might also take a look at Ren'Py's auto_voice feature, which would allow you to automate all voice audio and allow you to gear the file names to include a translation folder or component...

Code: Select all

default auto_voice_file_string = "auto_voice/{language}/{identifier}.ogg"

init python:

    def get_auto_voice_file_name(identifier):

        # Amend the dict {language name : folder name} to suit 
        
        vdict = {
            'language' : {
                None : 'en', 
                'Spanish' : 'es',
                'Nihongo' : 'jp',
            }.get(_preferences.language, 'unknown'),
            'identifier' : identifier,
        }

        return auto_voice_file_string.format(**vdict)

    # Tell Ren'py to *try* to voice every line
    # using the above function to locate file
    config.auto_voice = get_auto_voice_file_name

label start:

    show screen _auto_voice

    "..." # This line is voiced if Ren'Py can find a file such as "auto_voice/en/start_a20cefa7.ogg" as shown in the screen

config.auto_voice can be a string, function or None ( https://renpy.org/doc/html/config.html# ... auto_voice )

If you do decide to go for auto_voice, I have a voice artist helper script that might be of use. Send me a PM if you want to try it (it basically reads the entire game and outputs all the lines and related filenames that each character should speak).

Thanks for your help Remix! This particular game is a remastered version of my first game with some extra langauges added, so I hadn't used auto voicing yet.
I used auto voice for one of my later games and I guess I wouldn't have known what to do when translating auto voice, so I will probably give that a go for a future game. Thanks again!

I'm still interested in finding a way to get the line identifier for any specific line (is it possible or do they only exist for dialog, text and voice?), like I mentioned earlier, sometimes there are images, especially images that contain text, that would be nice to change for translations. I wonder if this method could replace lines of random code like that.

[Matalla]

For "translated" images, you don't need any identifier, just drop them in the tl/whatever_language/original_path/ with the same filename and renpy will show them

So, for example, if you have the image "bg_neon_sign.jpg" in your images folder, and you want to show an alternate one for, say, a spanish translation, put "bg_neon_sign.jpg" with the alternative content in tl/spanish/images/ and you're done

[Remix]

I didn't fully read your initial question...

All your syntax seems to indicate labels, yet you say you are using screens which call labels... Not too sure what you are after, though might advise an approach like:

Code: Select all

screen voice_test():
    text "Here"

    on "show":
        action PlayCharacterVoice( "eileen", _("arg.ogg") )

to add a character voice to a screen (using on "show" and action to avoid possible side effects)
This *should* translate fine