@ben @dan I just wanted to let you guys know I’m loving Elements! The Claude Ai integration has made my article development so much better. And it helped me create the code for a Custom Component so my visitors can copy the article by pressing a button. It’s cool.
I added a Listen to Article button this morning, if there is an mp3 recording, the button says Read by Author, if no mp3 is available yet, it says Read by Computer.
Give it a try…
that’s amazing, keep up the good work! 
Really great work - listened to the recording! I like that a lot…
Don’t be afraid to increase the bit rate to 128 and perhaps consider a .m4a format for even better sound
Very very good work!!
It’s really low right now because I ran into a file size limitation trying to upload through Claude or through the CMS Dashboard of 5MB. I know I could do a SFTP upload to get around it, but was trying to keep it in an automated work flow.
Yes - really really consider uploading a better copy - I’m reallly biased, been recording professionally for 40+ years
I - personally- can’t listen long to low quality recordings… very hard to concentrate - for me
Unless it’s a historical recording from the pre 1920’s make the exception…
Not hard to add a link…
All browsers now play .m4a files which are much better than mp3 format
And 128 is really really good…
Just perhaps consider it - it’s just my opinion…
At this point that is what I have decided. Easier to upload using SFTP instead of through the CMS Dashboard which appends file names with a number which is a pain to have to rename to delete the appendage.
I’ve also decided I don’t want to record audio for all these articles either, so I’m going another route. Bantr app which converts the text file into audio with my voice. It’s only $59 lifetime no subscription, unlimited use. It think it does a decent job. Bantr saves in WAV but I had Claude Ai create a script to do the conversion to .m4a format.
I’ve started uploading the new files if you want to hear the quality.
For my podcasts - I use eleven labs, they have a starter level and developer level
I use v2 voices for long format speaking
V3 is nice but only for 30-60 second commercials - uses too many credits
Glad you found a way to use .m4a files - great 
TrumpetStudio.com podcasts… if you’d like to hear…
256 .m4a files
Those sound really good. I saw the Eleven Labs and it does get the highest ratings, but I didn’t want to have to keep paying for credits.
Well done @ercross 
What have you used for the “Read Time” indicator? A custom component?
I used the ReadTime component from RW Pro Space
Thanks, I saw this component on the store but never in action.
Listened…
This is much much better audio - really makes a huge difference in listening!!!
Really good
There is also Eleven Labs - use your own voice…
You give it audio samples.
It will read your entered text…
However:
A lot of music names and terms have to be phonetically spelled out… embouchure into ahm-boo-shure or my name Droste into Drah-stee
Etc…
You might have to do that with biblical names and terms…
But might be something to look into…
You can get Claude to phonetically correct your document for proper speech creation…
Just an update. This has been an amazing discovery trip. I found Bantr was good at creating the narration with my own voice, but it has no command line interface to be scripted to go through all my existing 80+ articles. So it was going to be a one at a time process and I didnt’ want to stay up at night to load the next one. So I asked Claude Ai if it could make a command line app that would convert text to natural voice using my own voice. It said yes with an open source engine from MIT.
It took me the greater part of the day with some trial and error, but I now have a command line app chugging through all my articles. It uses a pronunciation text file for any Biblical names that need to be sounded out phonetically and if there is a particular word that the phonetics can’t get right I can record a audio file of just that word that it substitutes in. I had it compare it’s file results to the same one that I had run through Bantr, and it adjusted it’s parameters to match the output I was getting from Bantr.
It not only generates the file, but it is also uploading them to the web server as it completes them. I really am amazed at the power of Ai.
Very cool…
Wow, that is really cool. Amazing what you can do with little sprinkle of AI 