text limit was unable to be accurately determined and seemed to depend on numerous factors, including voice and/or language. In testing I was able to get above 1900 characters with many voices, but certain strings (e.g. a long line of emojis) would result in an error even if the length was well under that. Text of an equivalent length in languages that do not use a Latin character set would also produce an error. Limit is therefore set at 1900 bytes as this should work for most voices.
text limit was unable to be accurately determined and seemed to depend on numerous factors, including voice and/or language. In testing I was able to get above 1900 characters with many voices, but certain strings (e.g. a long line of emojis) would result in an error even if the length was well under that. Text of an equivalent length in languages that do not use a Latin character set would also produce an error. Limit is therefore set at 1900 bytes as this should work for most voices.