RadioBOSS Cloud

[djsoft]

The last cloud update notes say playcount and last played have been updated/fixed... where is this information shown in cloud?

It's in the Media section, press the Settings button at the top right, and enable "Show track information" checkbox.

 

[jharris171185]

I appreciate the feature for weather announcements, but I can see there could be a potential problem here. Often, there may be cities with the same name, but in different states. For example, I live in Independence, Kansas, but there are also cities with the same name of Independence in Missouri, Kentucky, and maybe even a few others here in the U.S. The documentation mentions nothing about adding a state. So, in this case, how do we make sure the right city's weather is retrieved? If you haven't thought about that, you could do either city state country, or let us enter a zip code instead.
Also, will this work in other languages than English? My station is in Spanish, and I would really like to be able to use the feature.

 

[jharris171185]

Hello again,
So to test, I entered in the name of a nearby community. I still hope it can be fixed to add in the state to be able to get exact weather for my town, but to at least get up and going, this seems to work. However, there is one big problem, and that is that accented letters like ñ á é í ó ú are removed from the text we enter to have read out. Please fix that ASAP. Lastly, with things the way they are now,, it sounds a bit robotic. While it wouldn't completely solve the problem, it might help if a music bed could be assigned to play along with any tts events, and obviously we would want the music bed to duck during the announcement, but still have a decent enough level. Maybe let the user set the bed and level compared to speech.

 

[djsoft]

I live in Independence, Kansas, but there are also cities with the same name of Independence in Missouri, Kentucky, and maybe even a few others here in the U.S

We'll issue an update soon (probably today) to allow including state name in the query, in your case it'll be like weather Independence, KS, US

there is one big problem, and that is that accented letters like ñ á é í ó ú are removed from the text we enter to have read out

Thank you, we'll check it.

with things the way they are now,, it sounds a bit robotic

It depends on the voice you use. Please see this page: https://www.radioboss.fm/support/radioboss-cloud/text-to-speech/
It has a link to the table that lists all voices. You can probably try "Premium" voices that should sound better.

 

[jharris171185]

We'll issue an update soon (probably today) to allow including state name in the query, in your case it'll be like weather Independence, KS, US

Awesome. Much appreciated.
Another thing I have noticed is that even though the TTS is only scheduled once, it often repeats. So it will read everything out once, and then it reads again. Would appreciate this being addressed.

 

[djsoft]

Awesome. Much appreciated.
Another thing I have noticed is that even though the TTS is only scheduled once, it often repeats. So it will read everything out once, and then it reads again. Would appreciate this being addressed.

A quick test does not confirm this, scheduled TTS item is inserted into the playlist, and after it's played, it's removed and can no longer be played, until the scheduled event runs again.

RadioBOSS Cloud 2.0.1 released with some fixes: https://www.djsoft.net/community/threads/radioboss-cloud-releases.11385/post-50942

 

[jharris171185]

A quick test does not confirm this, scheduled TTS item is inserted into the playlist, and after it's played, it's removed and can no longer be played, until the scheduled event runs again.

RadioBOSS Cloud 2.0.1 released with some fixes: https://www.djsoft.net/community/threads/radioboss-cloud-releases.11385/post-50942

I don't know why it doesn't happen in your test, but I can confirm it has happened multiple times on my station.
This is what I have.
tts=+0::ES-US:2;En este momento, en Independence Kansas, la temperatura es de ?temp_f grados Fahrenheit, y hay vientos de ?wind_speed_m millas por hora.
But like I said, it plays twice. Not good.

 

[jharris171185]

So, just to give more info on my scheduler setup and station config, my settings are play silence when nothing is scheduled. Station ID is played at 59:45 of each hour. Then at 00:00 of each hour, I have a time announcement, not via tts. Then, at 00:05 is the weather announcement, which I can confirm is reading twice each time. The first time it reads most of the way through, but not all the way, and the second it reads all the way through. Then, at 00:15 of each hour, that particular hour's programming starts. That might sound like a big gap, but the idea is to have a bit of space, where the weather report doesn't get cut off by, or mixed with, the ad that plays right after. And since I can't set tts as a filetype where no crossfading will happen, I need to leave that gap. But my guess is, either 1. it's taking the tts a bit to process the weather info and that's why it seems to restart, or 2 it is trying to fill the gap between the event and the next playlist. If we could set tts as a filetype, then we could set it so that it would never play two back to back, and maybe fix the issue.
Here are my logs for the most recent instance. Not sure how much they will help, but will include anyway. I only have this called for once in the scheduler, but you should be able to see that it is playing twice.

Starting event "anuncio de la hora"
2024-06-17 3:00:01 PM
Schedule: [anuncio de la hora] time1500.wav (~files\TimeFiles\time1500.wav)
2024-06-17 3:00:01 PM
time1500
2024-06-17 3:00:01 PM
HTTP Request (#1) successful: 200
2024-06-17 3:00:06 PM
Starting event "Clima para Independence, Ks."
2024-06-17 3:00:06 PM
Schedule: [Clima para Independence, Ks.] tts=+0::ES-US:2;En este momento, en Independence Kansas, la temperatura es de 90 grados Fahrenheit, y hay vientos de 5 millas por hora.
2024-06-17 3:00:07 PM
TextToSpeech: En este momento, en Independence Kansas, la temperatura es de 90 grados Fahrenheit, y hay vientos de 5 millas por hora.
2024-06-17 3:00:13 PM
TextToSpeech: En este momento, en Independence Kansas, la temperatura es de 90 grados Fahrenheit, y hay vientos de 5 millas por hora.
2024-06-17 3:00:15 PM
Starting event "Música Tardes"
2024-06-17 3:00:15 PM
Schedule: [Música Tardes] Música Tardes.m3u8 (~playlists\Música Tardes.m3u8)
2024-06-17 3:00:15 PM
Inserted 38 tracks from Música Tardes.m3u8

 

[jharris171185]

Update: Upon adding the weather as an item in each playlist, and moving it to the top, instead of it being its own timed event, the tts only plays once as expected. Now, if there were just a setting to make it where tts wouldn't crossfade with the next item in the list, this would be good. I say this because, on some hours, I start off with news, and when it mixes with that, it sounds okay, but on others, I start with an Ad, and in those instances, I don't want it mixing, because I broadcast to a shoutcast server with one of my encoders, so I can take advantage of shoutcast monitization, and when ads are triggered, if the tts is getting mixed with the first few seconds of the ad, then the end of the tts would be cut off, since shoutcast replaces your stream audio with relevant ads for the listener. I hope this makes sense.

 

[dnilo]

Wonderful incorporation of the time announcement.
How can I tell the time announcement not to start until the song currently playing is completely over? And that the next song doesn't start until the time announcement is completely over?

 

[djsoft]

Update: Upon adding the weather as an item in each playlist, and moving it to the top, instead of it being its own timed event, the tts only plays once as expected

I'm glad the problem is resolved. We were unable to reproduce it, maybe there was something that we didn't do, so there was no error.

Now, if there were just a setting to make it where tts wouldn't crossfade with the next item in the list, this would be good.

You can add a file type: https://www.radioboss.fm/support/radioboss-cloud/using-file-types/
Set the Identifier to "tts=" (no quotes), this will cover all text to speech items, and you will be able to set the crossfades for them.

 

[djsoft]

Wonderful incorporation of the time announcement.
How can I tell the time announcement not to start until the song currently playing is completely over? And that the next song doesn't start until the time announcement is completely over?

You can schedule an event with TTS item to announce the time, set its option to place it in the queue after the playing track. You can read more about the scheduler here: https://www.radioboss.fm/support/radioboss-cloud/scheduler/

 

[dnilo]

You can schedule an event with TTS item to announce the time, set its option to place it in the queue after the playing track. You can read more about the scheduler here: https://www.radioboss.fm/support/radioboss-cloud/scheduler/

The problem is that the time announcement is broadcast, but it overlaps with the songs.
I think it's because of the crossfader values.
The situation is that when a song is about to end, the time announcement starts, and before the time announcement ends, the next song has already started and the time announcement is not distinguished correctly.
I hope I have explained my problem well.

 

[djsoft]

but it overlaps with the songs.
I think it's because of the crossfader values.

You can use the file types feature (by default there's already a file type exists for TTS): https://manual.djsoft.net/radioboss/en/using_different_types_of_files.htm
Then configure crossfading for those items to prevent track overlap.