first of all, as I wrote already, congratulations. This is truly amazing.
That said, there is one important feature that I believe is missing in your CW sender and I am here to point that out to you.
In order to send CW (I use wfview 1.61 on OSX) I currently need to first type the entire message, then press “SEND”. That is fine, but it should also be possible to first press “SEND” and THEN start typing the message. As the message is typed, the radio keeps sending, running after the the text being typed. If the sent part reaches the end of the message, the radio stops sending and waits for more characters to be entered. When “STOP” is pressed, the system switches back to RX (that’s how the fldigi CW sender works, by the way).
This seems a small detail, but it is actually quite big. When making a CW QSO it is practically impossible to preload a message in its entirety, it simply takes too long. I really hope you will consider this, so I can start using wfview also in CW.
Oh boy, you are going to like the next release, which will have what you’re describing only even better. I’ll let @phil describe it when the sun reaches his part of the globe later today. He has improved the CW function immensely, as only an experienced CW op could do.
I’ll just plug a little here, if you sign up on our patreon, you can see part of it in action. We post updates there as features get worked on. It’s our little secret developer’s lab of ideas :-).
I have been working on two features over the last few days, these are:
Immediate send option. This will send the preceding characters to the radio whenever the spacebar is pressed
Local sidetone. As LAN connected radios do not produce CW sidetone, I have added the facility for the client to generate its own local sidetone which should closely match the timing of the radio. This may also be useful for USB connected remote radios due to connection latency.
I did consider sending each character separately, but due to the way that CW is sent over CI-V, this can result in a large volume of messages being sent to the radio, which might overwhelm the connection, so sending ‘per word’ seemed a reasonable compromise.
I have also identified some potential causes of crashes in the CW Sender window, once everything has been more thoroughly tested, we expect to do another release soon.
That is fantastic! Can’t wait to see it working and to finally be able to do CW remotely.
I know it is a complicated mess, but is there any hope to use paddles / straight keys someday? Just curious if this is something you are considering.
these two new features work really great on V1.62. However, I noticed that the sidetone attack waveform seem to be quite harsh creating quite strong clicks on my PC speakers even at lower sidetone volume (that’s on a Windows machine / Notebook).
I wonder if this is just on my machine or a general problem with the audio-generating code?
Would there be a way to create a cleaner tone with perhaps a softer waveform ramp-up / attack times?
(in case this is at all controllable by your code)
Otherwise, besides this tone quality issue, this new local sidetone feature is really great and useful…
(maybe storing last sidetone audio volume setting would also be cool)
I just tried fooling around with CW Sender on a MacBook, I got it to send “QRL?”, but then it wouldn’t send anything after that, even after I quit the program and re-started it.