Many apologies for this one everyone. I re-added my microphone source in OBS and forgot to tick the mono button. So your speakers/headphones aren't faulty. Just me messing up. Ahhh well, it's up there now, but just so you know -:(
lol, you got me there, i was checking my cables and then after some minutes i realized, the intro music was stereo, must have been in the video itself, and after watching i read the comment here, well if we wouldn't like to fiddle arround, we wouldn't be here on this channel, i guess
@@OldTechBloke Bloke, it is little things like this that remind me why I like your channel. You are not some large corporate group pumping out nonsense day after day. Rather, you are one man working by yourself to make well polished, educational videos.
Steve, thanks for the interesting tour through KDE connect and the world of computerized dictation. As always thanks for making such great content and please stay well!
This is a massive time saver when you are creating documents explaining how things work. I'm setting this up tomorrow. I'm already a KDE user with KDE Connect installed. Thank you for this video.
Thanks Steve, you are making my daily work easier with this dictating approach (KDEConnect). I really appreciate it.. I checked it using Kate as text editor..Really works!! Thanks again :)
Nice. I gave this a try a few months ago. I did work as you pointed out but as I am not a KDE person when I installed this it downloaded a ton of KDE stuff... That was a bit of a drag... I had it working on my Arch Cinnamon and it worked fine... It was ok but after a while KDE Connect got under my skin... Lol I have a sister that loves to text every thought in her head, which drives me bonkers! Anyway it did a great job of that. Never did the dictating part. Now If I go to text on my phone I always hit the little microphone... :-) Here's my thinking is, If it works for you, then just use it! I don't really care about the other stuff. If I need it and it works, then there it is! Great Video OTB! Don't work to hard and stay safe! Merry Christmas Happy Holidays! LLAP
As another OTB my thanks! I found this a refreshingly low-key and practical coverage of a complex and drought issue. I, too, am looking for install-and-use v2t apps, not an education in neural networks, and really appreciate your pragmatism. Cheers!
I'm in the same boat as you, but the Office web apps won't handle my massive Excel spreadsheets. A Windows VM works well for me. Since I have a Windows VM, Dragon Naturally Speaking v15 is still my speech to text software.
That looks like a good website at 6:13, Steve. I bet you have a list of them. There's an idea, how about doing a video on your favourite Linux links? I'm off to check out the "best software" on that website now 🙂 That KDE Connect is great. Now I can check my phone when Shotcut has finished a render without having to get up and check the VDU 😀
One problem with speech to text (and text to speech) for the Open Source community is that basically everything is done via neural networks (which is understandable considering how complicated languages are). Training these models take a looong time and even more processing power (which costs a lot of money), something we are generally short of. Especially because we at some point need to do more languages than just English (if you (because of some reason) wonder why, look at the Wikipedia article called English language in Europe), and by far more accents (an example of how different that can be, here a german example: "Eichhörnchenschweif" (standard german) "Oachkatzlschwoaf" (Bavarian, quite different, isn't it?) "squirrel tail" (english)) which are even more problematic if you consider that some people can't speak without their (sometimes quite thick) accent.
Did not know KDE Connect did that much stuff. Looks like there has been some attempts to create a dictation skill for Mycroft, but nothing fully there yet. May be something to keep an eye out for in the future.
This isn't working for me in July 2022. I have KDE on Linux and my KDEConnect screen looks a bit different than in the video, in that the text I input on the phone doe not get automatically sent to the pc as you type/dictate, but there's a "send" button that you need to press when you're done. However, that has no effect for me, the text I dictated on the phone never appears on my linux desktop, and it's making me very sad :-(
Great video, and great tool from kde ..i just wish they leave a 2Fa app which supports dark mode. Authy and Authenticator dont support dark mode. Surprising to see the lack of 2fa apps for linux. Considering its an essential today. Would love to hear you're opinion Thanks Cheers from India❤️
Tried it out with my Huawei Phone (an older P9 lite) - and it doesn't work proberly. I'm running Kubuntu, so i guess it's the phone that is faulty. :-(
@@OldTechBloke oh, ok, I didn't mean that, but it is a quite interesting thing. Scrcpy is the program itself; you can download it from github and use it via its cli.
Good job OTB I’m dictating this on my phone so it does work. I keep forgetting to use the dictation feature and end up doing bad typing instead. Wouldn’t it be great if you could edit documents by voice as well as dictate them?
Does the transcription work offline, or does it rely on processing in the cloud? I'm guessing, like you, that most Linux voice to text leverages the Google API. On the web, I'd want to know which company is handling the data, especially whether they claim the right to retain the text copy Certainly the KDE connect method you demonstrate explicitly sends everything via Google This matters for two reasons: 1) ability to work when you've got a good web connection 2) privacy considerations (see several of videos posted on ua-cam.com/users/BraxMe -- he is very critical of Google on privacy)
I wish you your presentation skipped most of overlay and dove directly into the core findings far quicker. After 10 minutes of friendly banter and unimportant context, I left not knowing, will this guy end in meaningful territory?
Many apologies for this one everyone. I re-added my microphone source in OBS and forgot to tick the mono button. So your speakers/headphones aren't faulty. Just me messing up. Ahhh well, it's up there now, but just so you know -:(
Well bugger! That was the thing I noticed... :-)
Its ok this time ... LOL
On with the Video! More comments to come... :-)
LLAP
Cheers Bruce and apologies
I was really checking my expensive USB headset for 10 minutes before reading the comment :)
@@mylinuxgr5050 lmao i do same thing haha
lol, you got me there, i was checking my cables and then after some minutes i realized, the intro music was stereo, must have been in the video itself, and after watching i read the comment here, well if we wouldn't like to fiddle arround, we wouldn't be here on this channel, i guess
Please don't retire your channel. We need is blokes like us.
You scare me, man! For a second I was worried about my headphone or headphone-cable. Mono-sound (left). :p
Damn, just realised I’ve been re adding sources in obs and failed to Tick the mono box. Ahhhh! Too late now
@@OldTechBloke Bloke, it is little things like this that remind me why I like your channel. You are not some large corporate group pumping out nonsense day after day. Rather, you are one man working by yourself to make well polished, educational videos.
Steve, thanks for the interesting tour through KDE connect and the world of computerized dictation. As always thanks for making such great content and please stay well!
Thanks Deech
This is a massive time saver when you are creating documents explaining how things work. I'm setting this up tomorrow. I'm already a KDE user with KDE Connect installed. Thank you for this video.
Glad it was helpful!
Thanks Steve, you are making my daily work easier with this dictating approach (KDEConnect). I really appreciate it.. I checked it using Kate as text editor..Really works!! Thanks again :)
You’re welcome mate
KDEConnect is the killer app that got me hooked on KDE Plasma
Nice. I gave this a try a few months ago. I did work as you pointed out but as I am not a KDE person when I installed this it downloaded a ton of KDE stuff... That was a bit of a drag... I had it working on my Arch Cinnamon and it worked fine...
It was ok but after a while KDE Connect got under my skin... Lol I have a sister that loves to text every thought in her head, which drives me bonkers! Anyway it did a great job of that. Never did the dictating part. Now If I go to text on my phone I always hit the little microphone... :-)
Here's my thinking is, If it works for you, then just use it! I don't really care about the other stuff. If I need it and it works, then there it is!
Great Video OTB! Don't work to hard and stay safe!
Merry Christmas Happy Holidays!
LLAP
Cheers Bruce, hope you and yours have a great Christmas as well
Audio issue aside.... Great video Steve, really enjoyed the content, I learned some new things, which is always welcome! :-)
Sorry about that Tim :-(
As another OTB my thanks! I found this a refreshingly low-key and practical coverage of a complex and drought issue. I, too, am looking for install-and-use v2t apps, not an education in neural networks, and really appreciate your pragmatism. Cheers!
Cheers Steve I've been looking for something like this
You can open you terminal from the phone and interact with it using the remote from kde connect, like typing any command.
Thanks for this. I have a disability and rely on voice to text.
let's talk after the intro - with me thinking, no, no need for an intro, don't waste your time making one.... just, pls continue.
I'm in the same boat as you, but the Office web apps won't handle my massive Excel spreadsheets. A Windows VM works well for me. Since I have a Windows VM, Dragon Naturally Speaking v15 is still my speech to text software.
That looks like a good website at 6:13, Steve. I bet you have a list of them. There's an idea, how about doing a video on your favourite Linux links? I'm off to check out the "best software" on that website now 🙂
That KDE Connect is great. Now I can check my phone when Shotcut has finished a render without having to get up and check the VDU 😀
Not a bad idea Stephen
18:00 i tried doing this exactly . I don't know why it doesn't work on my setup . I would just dictate 'hello' and i wouldn't even move a cursor . 😢
I have my pc directed connect to my network router and my smartphone in wireless, kde connect works!
KDEconnect has been very great to me as well in Awesome
One problem with speech to text (and text to speech) for the Open Source community is that basically everything is done via neural networks (which is understandable considering how complicated languages are).
Training these models take a looong time and even more processing power (which costs a lot of money), something we are generally short of. Especially because we at some point need to do more languages than just English (if you (because of some reason) wonder why, look at the Wikipedia article called English language in Europe), and by far more accents (an example of how different that can be, here a german example: "Eichhörnchenschweif" (standard german) "Oachkatzlschwoaf" (Bavarian, quite different, isn't it?) "squirrel tail" (english)) which are even more problematic if you consider that some people can't speak without their (sometimes quite thick) accent.
Did not know KDE Connect did that much stuff.
Looks like there has been some attempts to create a dictation skill for Mycroft, but nothing fully there yet. May be something to keep an eye out for in the future.
Mycroft looks promising but we're not there yet unfortunately
why is is this audio only on the left side of Stereo
This isn't working for me in July 2022. I have KDE on Linux and my KDEConnect screen looks a bit different than in the video, in that the text I input on the phone doe not get automatically sent to the pc as you type/dictate, but there's a "send" button that you need to press when you're done. However, that has no effect for me, the text I dictated on the phone never appears on my linux desktop, and it's making me very sad :-(
Great video, and great tool from kde ..i just wish they leave a 2Fa app which supports dark mode.
Authy and Authenticator dont support dark mode.
Surprising to see the lack of 2fa apps for linux. Considering its an essential today.
Would love to hear you're opinion
Thanks
Cheers from India❤️
what android version? gboard installed but voice typing does not work. it doesn't transfer the voice typed text. android 8 obviousely too old
I just need a tool to dictate (after a training session) and in my word processor the text should appear...
Tried it out with my Huawei Phone (an older P9 lite) - and it doesn't work proberly. I'm running Kubuntu, so i guess it's the phone that is faulty. :-(
That's a shame
Merci !!!!!!
This is more about connecting your phone to your pc, not really about voice to text dictation. Wish it didn't take 19 mins to figure that out
Hello, how would you describe the screen visualising on pc? Was it a perfect latency (~
I take it you mean Vysor? It's not bad actually. I use it to train people remotely how to use mobile apps we use at work
@@OldTechBloke oh, ok, I didn't mean that, but it is a quite interesting thing. Scrcpy is the program itself; you can download it from github and use it via its cli.
Good job OTB I’m dictating this on my phone so it does work. I keep forgetting to use the dictation feature and end up doing bad typing instead. Wouldn’t it be great if you could edit documents by voice as well as dictate them?
Oh yes!
I cut my laptop in half to find out why my RH speakers not working :-)
Oh dear :-( I did pin a comment. Sorry
can anyone explain why my top transfer speed is capped at 2mbps on kdeconnect
Could this be your WiFi connection speed? Remember everything going between the phone and the computer travels over your WiFi network
Not a clue, can't say I noticed any issues
@@trueriver1950 but the components connected are being my phone and laptop
Does the transcription work offline, or does it rely on processing in the cloud?
I'm guessing, like you, that most Linux voice to text leverages the Google API.
On the web, I'd want to know which company is handling the data, especially whether they claim the right to retain the text copy
Certainly the KDE connect method you demonstrate explicitly sends everything via Google
This matters for two reasons:
1) ability to work when you've got a good web connection
2) privacy considerations (see several of videos posted on
ua-cam.com/users/BraxMe
-- he is very critical of Google on privacy)
Unfortunately they all leverage the google api. Not ideal at all, but it's what we have at the moment
I wish you your presentation skipped most of overlay and dove directly into the core findings far quicker. After 10 minutes of friendly banter and unimportant context, I left not knowing, will this guy end in meaningful territory?
16 minutes in and you haven't gotten to the point. Bye.