Weather Report Creator/Generator

[shenzhou]

Hi,

Radioboss manual illustration on how to create weather report is so conceptual that visual people find it hard to grab. Had it not from me combing the forum and found someone posting a screenshot showing example "weather New York, US" is to be entered into the Playlist/track filename or command box manually, I'd still be pulling my hair out. Because I thought after selecting "weather" from the dropdown menu, one would logically think "weather New York, US" or " New York, US" need to be entered/typed into maybe the Task name box or in the Text to Speech box. In nowhere in your manual do I find you specifically stated exactly where ""weather New York, US" is to be input! Your manual only said to select weather. Now after what I thought I had finally successfully created the first weather report schedule I then created a second with a different city and then a third city. New problem. The 3 cities differed in climate condition as in heaven and hell but the report announcement on each all stated the same temperature and wind description as the first city????

The way you need the user to set up a weather report involved too much pussyfooting!!!! Can't radioboss design a more automated weather setup module in the program. The less user manual syntax input the better. Give us a restaurant menu style set up. Meal A B C D E!!! Each column input with dropdown menu selection?

Select Weather Server/Select Country/Select State/ Select Province/Select Postcode/Select City/Select Installed Computer Voice/Input prefix text/Input suffix text/

This way, the chances of the user F'kedup is right next to zero. The only screwedup would be then narrowed down to either the weather server and/or the program itself.

Likewise you can implement the same for Time Announcement.

For your consideration.

Thank you.


Regards,
Shenzhou

 

[ormston]

time announcement is no problem cos time is time. but i agree that the weather is a bit of a mess to get the api to feed RB with the required data and then then RB to read it and say it, i worked it out eventually but it could be made much more user friendly

 

[shenzhou]

With time announcement, RB should also design to allow user to easily select an installed voice from Windows 10. Including a Prefix and Suffix optional input text fields/boxes could allow user to spice things up a bit for their broadcast. Time announcement work, just suggestion here how we might make Time announcement more fun/imaginative....

 

[ormston]

With time announcement, RB should also design to allow user to easily select an installed voice from Windows 10. Including a Prefix and Suffix optional input text fields/boxes could allow user to spice things up a bit for their broadcast. Time announcement work, just suggestion here how we might make Time announcement more fun/imaginative....

yes, the time announcements could be better implemented to allow a little more flexibility to allow words to put in before and after the actual time to make it better, or as the previously posted said, "to spice things up a bit"

 

[ormston]

like "the time on xxx radio is", "time" "and have a great day" just like i can do on the weather api thing

 

[djsoft]

Because I thought after selecting "weather" from the dropdown menu, one would logically think "weather New York, US" or " New York, US" need to be entered/typed into maybe the Task name

Gathering weather data is as simple as creating an event with weather City, Country command and set it to repeat e.g. every 2 hours to update the data.

By the way, your second screenshot is wrong, it has a space before the weather command, so it may not work.

but i agree that the weather is a bit of a mess to get the api to feed RB with the required data and then then RB to read it and say it, i worked it out eventually but it could be made much more user friendly

What could be changed about it? Gathering weather data is one event. Announcing it is another event. That's about it.

yes, the time announcements could be better implemented to allow a little more flexibility to allow words to put in before and after the actual time to make it better, or as the previously posted said, "to spice things up a bit"

It's possible with "Multiple actions" event, add the "opener" and then the time announcement.

 

[shenzhou]

Can someone kindly provide me an example(with screenshots if possible) how to set up weather and time announcement using a selected voiced installed in windows 10? Thank you.

 

[shenzhou]

Using the latest version of Radioboss and using the example syntax for weather report.

Temperature today is ?temp degrees Celsius wind speed is ?wind_speed metres per second ?wind_name humidity is ?humidity percent. Tomorrow ?temp_tom degrees Celcius ?wind_speed_tom metres per second ?wind_name_tom humidity is ?humidity_tom percent.

I created two weather reports, one for Singapore, SG and another for Brisbane, AU. I run one after the other. It updated and announced most except humidity it didn't say it so it went "humidity is percent". The weather for Singapore at time of writing is 22°C which it retrieved and was accurate.


However when I run the Brisbane weather event, it is announcing the Singapore data? Why? Brisbane temp is 29°C but it announced 22 which is that of Singapore.

Thank you.

 

[djsoft]

It announces the latest weather data that was retrieved using the weather command. The command is not executed immediately as it needs to gather the data from the weather source, so if you have an event with two actions like

1. weather ...
2. Text-to-speech to annunce

This will not work because it starts "weather" command then, without waiting for it to finish, it will run the announcement, that wil announce the previous weather data.

Currently the system is not designed to work with two or more different cities.

 

[ormston]

this would only work if the "get weather" command for various cities is ran manually (or if a timed event, there is some time elapsed) to allow the command to complete fully and update on the main display, then some time later the text-to-speech part is ran

 

[jharris171185]

Ok, so is it possible to do something like:
The current temperature is ___ degrees, with winds of ___ miles per hour, and ___ percent humidity. The high for today will be ___ degrees, and the overnight low will be ___. Tomorrow, we're expecting a high of ___.
I can't find anything in the documentation to pull that off.

 

[djsoft]

You can find all available weather variables here: https://manual.djsoft.net/radioboss/en/automatic_weather_announcement.htm

 

[jharris171185]

Ok, so sounds like I should just go with my prerecorded files that give current temp, since I'm not seeing more options to do what I want with the tts.

 

[jharris171185]

I am still confused though. I set up my files as stated in the documentation, but I only have temp files, not an intro or humidity, so it makes no sense to have to create a playlist for just that one file. But while inserting a time announcement was super easy, just adding the command saytime=D:\TimeFiles, I haven't figured out how to do the same with the temperature. Could I do:
saytemp_f=D:\TempFiles? If not, you might consider adding this in the next version. It would make things easier, especially for those of us coming from other software, where all you had to do was set the folder, and set a command to give temperature whenever you wanted in the playlist. I'm incredibly impressed with everything Radioboss can do, but if this could be implemented it would be much more simple.

 

[djsoft]

Could I do:
saytemp_f=D:\TempFiles?

No, there's no such thing - and I'm not sure how it'll work, it's possible to use temperature variables in the file names (e.g. in scheduled events like: D:\tempfile\?temp.mp3). For time announcements it's easy because the number of values is small so only 60 + 24 files are needed in total. For weather announcements, there are lots of combinations if you announce temperature and wind.
I suggest doing it using text to speech as described in the help file.

 

[jharris171185]

Hello,
If I use prerecorded files, I would only announce temp, nothing to do with wind. Just the current temperature. That's it. So why couldn't this be made to work?
The problem with doing it via tts is that NONE of the Windows Spanish voices are any good. Plus I already have these prerecorded files that are perfectly fine. They range from -5f to 110f. So long as RB can read the number in the file, what is to stop doing a simple temperature announcement? They are already done with background music and intro expression based on the range of temps. Plus, they mention my city, which being a Spanish station in the United States, I'm pretty sure none of the Spanish Windows voices will ever get right.
Please implement a way to just do a simple temp announcement in the same way we do time announcements.

 

[djsoft]

If I use prerecorded files, I would only announce temp, nothing to do with wind. Just the current temperature. That's it. So why couldn't this be made to work?

In this case, it will work. Use ?temp variable in file names.

Please implement a way to just do a simple temp announcement in the same way we do time announcements.

It's already possible, create a scheduled event with file name like D:\tempfile\?temp.mp3 (change the path and file name to actual).

 

[jharris171185]

Ok, so in your documentation, it says to name the files as tmp01, tmp02, tmp34, tmp100, etc. Am I understanding you that files should be renamed to ?temp_f01 ?temp_f20 ?temp_f50 ?temp_f100, etc? And if so, the actual event should call file path/?temp_f.mp3.
Am I understanding this right so far? Also, all my temp files are .wav. Do they specifically need to be mp3?

 

[djsoft]

Ok, so in your documentation, it says to name the files as tmp01, tmp02, tmp34, tmp100, etc. Am I understanding you that files should be renamed to ?temp_f01 ?temp_f20 ?temp_f50 ?temp_f100, etc? And if so, the actual event should call file path/?temp_f.mp3.
Am I understanding this right so far? Also, all my temp files are .wav. Do they specifically need to be mp3?

Files can be in any format, it doesn't matter. The file names given in the help file are examples, you can name the files differently, it's important that they contain number for the degrees they are "saying" e.g. temperature02.wav must say "It's two degrees" and so on.
The point is that RadioBOSS changes ?temp (and other weather values) to actual number values.

 

[jharris171185]

Okay, but I would assume that in order for the system to know it is announcing farenheit all files should be prefixed with ?temp_f followed by the number, right? Otherwise how does it know to announce farenheit or celsius? Sorry for so many questions, but I really don't find this very clear to set up and I want to make sure I get it right the first time. So that means knowing exactly what my files should look like, and what my command should look like.
So, am I to understand that if I put in a command, calling for ?temp_f02, but the temperature is really 30, RB will automatically know to replace the 02 with the file for 30? Again, it would be so much easier to set up an automatic temp announcement in the same way we do time. With the way the temp works right now, it just seems there are so many things that could go wrong. This is one area where other broadcast software has made it much more straight forward. Falling back on your previous suggestion to just use tts, if we could get access to some higher quality voices that sound a little more human, then it would be easier and worth doing that way. If you could open the software up to use something other than the default windows voices, then it might be easier to figure something out. Also, it would be nice to have more commands to better describe the current weather, to get conditions like sunny, cloudy, currently raining/snowing, etc. Also, in addition to current temp, it would be good to have it pull high and low temps, either for today or tomorrow. Also, a way to insert if there is currently any kind of advisary such as tornado watch/warning, flash flood watch/warning, winter weather, etc. And, each of these should work in whatever language RB is set to use. Of course if there are no special alerts, that part should just be skipped. I know I might be asking a lot, but if we are going to use the tts, it should have as many options as possible and be made to sound as natural as possible.