The voice does not sound like a native speaker. I don't mean it sounds like a non-human character, I mean something is subtly off. Timing or something. Is that intentional? Maybe pick a different TTS solution?
This looks very cool, but I find it very hard to read the text against the moving background. The lights in the "windows" of the voxel building do not provide good contrast.
I don't know why, but that was an extreme load on my computer/browser. The picture changed every ten-fifteen seconds, I don't know if that's intentional or not, but it wasn't something I wanted to watch. The computer fan ran at maximum and it was hard to get enough CPU to close the browser tab.
For me it showed 400 fps in the top left corner and my laptop fans spun up immediately as well. Seems to render frames continuously rather than waiting for the screen to refresh. Would probably be much less load when limited to 60 fps.
If you right click on the page, you can 'open image in new tab' and get an image of the current screen. I assume that means it is rendering an entirely new image each frame... which is a tad wild.
It's wild if it's running at 400 fps because nobody has a screen that refreshes at 400Hz. Every frame rendered past the screen refresh rate is wasted compute. Easily solved by limiting the frame rate.
It's not running at 400 fps, the UI has the current global listener count next to a button that says "fps". The fps is only shown once you click that button. Weird design, I know.
Fun. I studied Japanese for two years, let it slide and now every Kanji is like "Hmmm, I've seen that before but..." I can still read kana though, which is nice to know.
This site caused my iPhone to start playing music and wouldn’t stop until I rebooted. I’m on the latest security release of iOS. Something off about that.
I’m on an 11 Pro. The site works fine. The issue is the audio. It somehow continues playing after the page is closed and after the application I’m using is closed. It took a phone reboot to stop playing.
I’m not claiming that there’s malicious activity on the part of the site. I just wonder if there is something anomalous about how it plays audio.
Tokyo has very pleasant vibes. It's an amazing city. My brother lives there since 30 years and I've been many times (it's so far that when I go, I go for 3 months).
My daughter loves it and, motivated by having japanese family, she's currently trying to learn japanese.
I think she'll love that site (I'll show it to her as soon as she comes back home).
Meanwhile I am sitting in an InterCity"Express" train in Germany, which already started 14 minutes late, and somehow managed to drive so slowly, that at every stop another ~2 minutes were lost. Now just standing around at a stop, because of some medical emergency at the next station, and announced a delay of fucking 90 minutes. How long can it take to remove whoever causes the medical emergency from the tracks or wherever they are and continue driving? It is all so nuts here, all so idiotic. The last 10 times or so, that I have been traveling, every fucking time they have stupid issues.
And they don't even manage to have functioning displays inside the trains. Every now and then the displays are turned off. Why am I not allowed to see the up-to-date status?
Yep, but then what takes 1h? They must have had that multiple times already. Either get them off the tracks in whatever state they are in, or get them out of the train. I don't find it to be a good reason to punish the collective amount of all train passengers in any train in the vicinity by making everyone an hour late. It seems very unprofessional.
The train has probably already hit them and they need to be cleaned off it because you don't want to give your train employees PTSD or you run out of employees.
Medical assistance can be received outside of the train, or once medical professionals are on board, while the train moves to the next station at the very least, if it isn't there already.
you listen to the audio and try to determine if you can either read or understand the audio. it's repeated twice, once formally and once casually, so good for listening practice, but also good for adding new words/phrases into your vocabulary since some of them might be familiar but maybe reconfigured in a way that you don't normally see in your practice.
this is probably only useful if you've started learning a very small amount of grammar, know hiragana well enough to make furigana useful, and have started memorizing enough kanji/vocab to make 'overheard train chatter' useful. probably, generously, something maybe 3-6 months into your japanese language journey, so not good for bootstrapping.
That makes sense. It seems a bit too slow to be really useful, as in, I feel like you'd want either to replay the audio as much as you want, or you can kind of listen passively/actively to something simple and slow.
I watched for a couple of minutes, and while it kept saying Japanese phrases, it just showed me floating by a large city landscape. Certainly nothing like being in a train.
Do you just need to wait a while to be in the train? Or for it to start?
the experience for attempting to record my voice during practices is atrocious. the background music continues to play, and the obnoxious AI voice interrupted me at least twice, then it switched to "transcribing..." wth?
I get the same feeling. There’s a lack of polish and intentionality to it. It’s a cool idea and a nice demo, and there’s certainly a lot of buttons and features.
But it’s been laggy on my device, it’s visually distracting, and the UX doesn’t seem particularly well-suited for practicing Japanese.
I think it's just one of the standard JP voice synthesisers available. I know I've heard almost exactly the same one on other Youtube videos using text-to-speech for Japanese.
So.. I know many people can read hiragana, but it is a very annoying habit of people who know a bit of Japanese to post un-transliterated Japanese text on an English language forum. For someone who doesn’t know Japanese your post reads ‘IIRC it’s a popular TTS character called ??NOT?FOR??YOU??’ - it communicates no information.
Writing ‘a popular TTS character called ずんだもん (zundamon)’ takes you very little time and gives readers a little more to work with, and which they can use to Google English language resources on the subject if they are interested.
Try to be more generous with your change requests - not everyone has enough time to fix every bug in their comment.
The strongest plausible interpretation is that when keywords are in another language, it is best to give the original language keyword rather than the anglecised/romanised version (which is so often incorrect e.g. Huawei's Tau[1]). It is also plausible that English is their second language.
Copy-pasting "ずんだもん" into Google gives you everything you want with the sidebar info, copy pasting "zundamon" into Google gives the same Wikipedia link on the sidebar. "Popular TTS character called" is enough to imply that what follows is the name, that you can then search.
As an addendum: Japanese isn't as all-in on pitch having semantics as Chinese, but pitch is very important. Words have "pitch accent" rather than "stress" the way English has, and overall pitch is strongly politeness-coded.
Speaking higher is politer, and noticeably dropping your pitch is threatening. You probably would pitch-up when speaking to your boss, and people working in shops or restaurants nearly sing welcome/thank you for coming type things.
Comparing our culture to others, women in the U.S. are raised to do much the same. Ask anyone woman constantly misgendered for not having a high-pitched woman’s speaking tone (e.g. pacific islanders) and they’ll confirm. Best not throw stones from glass houses.
The voice does not sound like a native speaker. I don't mean it sounds like a non-human character, I mean something is subtly off. Timing or something. Is that intentional? Maybe pick a different TTS solution?
i was expecting this lady https://www.youtube.com/watch?v=qUTD9r3G8W0 :D
I opened it and the TTS mispronounced the first sentence on screen. (ie it ignored the furigana and misread the first word)
This looks very cool, but I find it very hard to read the text against the moving background. The lights in the "windows" of the voxel building do not provide good contrast.
I don't know why, but that was an extreme load on my computer/browser. The picture changed every ten-fifteen seconds, I don't know if that's intentional or not, but it wasn't something I wanted to watch. The computer fan ran at maximum and it was hard to get enough CPU to close the browser tab.
For me it showed 400 fps in the top left corner and my laptop fans spun up immediately as well. Seems to render frames continuously rather than waiting for the screen to refresh. Would probably be much less load when limited to 60 fps.
You need to click the "fps" text for it to actually show the FPS value (it'll be green). The 400 you saw was something else.
If you right click on the page, you can 'open image in new tab' and get an image of the current screen. I assume that means it is rendering an entirely new image each frame... which is a tad wild.
> I assume that means it is rendering an entirely new image each frame... which is a tad wild.
What's wild about that? That's exactly what it's doing and what most uses of canvas/WebGL do.
It's wild if it's running at 400 fps because nobody has a screen that refreshes at 400Hz. Every frame rendered past the screen refresh rate is wasted compute. Easily solved by limiting the frame rate.
It's not running at 400 fps, the UI has the current global listener count next to a button that says "fps". The fps is only shown once you click that button. Weird design, I know.
Fun. I studied Japanese for two years, let it slide and now every Kanji is like "Hmmm, I've seen that before but..." I can still read kana though, which is nice to know.
Furigana is a lifesaver
The evening mode has vibes of Ghost in the Shell's "chase UI", glimpses of which is in this scene: https://www.youtube.com/watch?v=amKt8ttja1E
This site caused my iPhone to start playing music and wouldn’t stop until I rebooted. I’m on the latest security release of iOS. Something off about that.
Works fine for me on a 17e.
I’m on an 11 Pro. The site works fine. The issue is the audio. It somehow continues playing after the page is closed and after the application I’m using is closed. It took a phone reboot to stop playing.
I’m not claiming that there’s malicious activity on the part of the site. I just wonder if there is something anomalous about how it plays audio.
It's a bug in WebKit.
I am going to reach out to you via email. I like this project and I am building something similar. Maybe we can collaborate or keep in contact.
Voxel Tokyo has very pleasant vibes.
Tokyo has very pleasant vibes. It's an amazing city. My brother lives there since 30 years and I've been many times (it's so far that when I go, I go for 3 months).
My daughter loves it and, motivated by having japanese family, she's currently trying to learn japanese.
I think she'll love that site (I'll show it to her as soon as she comes back home).
Meanwhile I am sitting in an InterCity"Express" train in Germany, which already started 14 minutes late, and somehow managed to drive so slowly, that at every stop another ~2 minutes were lost. Now just standing around at a stop, because of some medical emergency at the next station, and announced a delay of fucking 90 minutes. How long can it take to remove whoever causes the medical emergency from the tracks or wherever they are and continue driving? It is all so nuts here, all so idiotic. The last 10 times or so, that I have been traveling, every fucking time they have stupid issues.
And they don't even manage to have functioning displays inside the trains. Every now and then the displays are turned off. Why am I not allowed to see the up-to-date status?
Medical emergency might well be somebody committing suicide (or trying to).
Yep, but then what takes 1h? They must have had that multiple times already. Either get them off the tracks in whatever state they are in, or get them out of the train. I don't find it to be a good reason to punish the collective amount of all train passengers in any train in the vicinity by making everyone an hour late. It seems very unprofessional.
The train has probably already hit them and they need to be cleaned off it because you don't want to give your train employees PTSD or you run out of employees.
Sounds like you have a solution for the trolley problem but modified so that the many are mildly inconvenienced instead of killed.
One day you might be that person who requires medical assistance. Think about that for a bit, considering what you just said.
Medical assistance can be received outside of the train, or once medical professionals are on board, while the train moves to the next station at the very least, if it isn't there already.
This looks good and I like the idea but I don't get what the "practice" is here?
Yeah I loaded it up and had no idea what to do.
you listen to the audio and try to determine if you can either read or understand the audio. it's repeated twice, once formally and once casually, so good for listening practice, but also good for adding new words/phrases into your vocabulary since some of them might be familiar but maybe reconfigured in a way that you don't normally see in your practice.
this is probably only useful if you've started learning a very small amount of grammar, know hiragana well enough to make furigana useful, and have started memorizing enough kanji/vocab to make 'overheard train chatter' useful. probably, generously, something maybe 3-6 months into your japanese language journey, so not good for bootstrapping.
That makes sense. It seems a bit too slow to be really useful, as in, I feel like you'd want either to replay the audio as much as you want, or you can kind of listen passively/actively to something simple and slow.
I watched for a couple of minutes, and while it kept saying Japanese phrases, it just showed me floating by a large city landscape. Certainly nothing like being in a train.
Do you just need to wait a while to be in the train? Or for it to start?
very cool and runs pretty nicely on my 2019 mac mini
This reminds me of Sim Copter from when I was a kid
What a cool idea! Need something similar for Korean - perhaps the EverLine light rail line in Yongin.
any volume control? voice is a bit loud and i couldn't find anything to turn it down.
the experience for attempting to record my voice during practices is atrocious. the background music continues to play, and the obnoxious AI voice interrupted me at least twice, then it switched to "transcribing..." wth?
cute concept, please fix that flow..
Does this scale beyond N5/N4?
This could be a really useful tool.
Epic!
love the tunes
lovely
Slop
I get the same feeling. There’s a lack of polish and intentionality to it. It’s a cool idea and a nice demo, and there’s certainly a lot of buttons and features.
But it’s been laggy on my device, it’s visually distracting, and the UX doesn’t seem particularly well-suited for practicing Japanese.
Looks nice, but what's the deal with Mickey Mouse voice? Is it sponsored by Disney or something?
I think it's just one of the standard JP voice synthesisers available. I know I've heard almost exactly the same one on other Youtube videos using text-to-speech for Japanese.
IIRC it's a popular TTS character called ずんだもん.
So.. I know many people can read hiragana, but it is a very annoying habit of people who know a bit of Japanese to post un-transliterated Japanese text on an English language forum. For someone who doesn’t know Japanese your post reads ‘IIRC it’s a popular TTS character called ??NOT?FOR??YOU??’ - it communicates no information.
Writing ‘a popular TTS character called ずんだもん (zundamon)’ takes you very little time and gives readers a little more to work with, and which they can use to Google English language resources on the subject if they are interested.
I did but I removed the English part, not because I assume people can read hiragana, but because I assume people will need to search it up anyway.
For me all the search results for https://www.google.com/search?q=%E3%81%9A%E3%82%93%E3%81%A0%... are in Japanese.
https://www.google.com/search?q=zundamon produces English language resources.
fwiw i recognize "zundamon" (and many other romanized japanese names/terms) without needing to look them up, but can not read the japanese
Try to be more generous with your change requests - not everyone has enough time to fix every bug in their comment.
The strongest plausible interpretation is that when keywords are in another language, it is best to give the original language keyword rather than the anglecised/romanised version (which is so often incorrect e.g. Huawei's Tau[1]). It is also plausible that English is their second language.
[1] https://news.ycombinator.com/item?id=48816034
Copy-pasting "ずんだもん" into Google gives you everything you want with the sidebar info, copy pasting "zundamon" into Google gives the same Wikipedia link on the sidebar. "Popular TTS character called" is enough to imply that what follows is the name, that you can then search.
Doesn't really sound like ずんだもん tbh. Can't recall which voice it is however.
Japan likes their cutesy little-girl voices.
that's just how their women sound to us. cutesy anime girls are voiced by aunties over 30.
As an addendum: Japanese isn't as all-in on pitch having semantics as Chinese, but pitch is very important. Words have "pitch accent" rather than "stress" the way English has, and overall pitch is strongly politeness-coded.
Speaking higher is politer, and noticeably dropping your pitch is threatening. You probably would pitch-up when speaking to your boss, and people working in shops or restaurants nearly sing welcome/thank you for coming type things.
How would you pitch-up?
Does the entire Japanese population learn these things while growing up?
I think it's called learning to speak.
So are cartoon boys in the west.
How infantile then
Comparing our culture to others, women in the U.S. are raised to do much the same. Ask anyone woman constantly misgendered for not having a high-pitched woman’s speaking tone (e.g. pacific islanders) and they’ll confirm. Best not throw stones from glass houses.
I don't know, I don't live or am from US.
Well, rephrased then in more literal speak: being less dismissively-judgmental and more observational will go a long way in this community.
That is an anime voice.