musicbycsharp
Stephane Clavel wrote :
musicbycsharp wrote :
This is the way it should be, if you have stemsRealTime set to "on demand". If the song was already transcribed by another virtualdj user, we don't need the stems2, so you will get the lyrics instantly. But if nobody transcribed it before, we need the stems2, so if you are set to on-demand, it will not transcribed automatically unless you demand it by opening the lyrics editor (or the lyrics or karaoke video plugins)I really like the AI Lyrics in 2026.
I just want to verify that this is proper behavior. I have had some songs that load with the lyrics and others where they don't load. To force the loading, I right-click and choose the lyrics editor that invokes the lyric extraction and this works. Is this the way we should be using it? Or is there another way to force the lyric extractions without going into the editor?
I just want to verify that this is proper behavior. I have had some songs that load with the lyrics and others where they don't load. To force the loading, I right-click and choose the lyrics editor that invokes the lyric extraction and this works. Is this the way we should be using it? Or is there another way to force the lyric extractions without going into the editor?
excellent. Thank you for that.
It would be nice to be able to extract the lyrics on demand in the background. The use-case is that during a set, I get a request from a student at a prom. The request is a hip-hop song and I cannot locate a clean version. But if I can extract the lyrics in the background while I play and not have the editor on my screen then I can play normally without the window. Then later I can view the lyrics and make use of the AI censoring.
发表时间 7 days ago @ 7:57 pm
djkrysr
Stephane Clavel wrote :
That's the problem with AI, it cannot be 100% accurate.
But usually, right-click -> Lyrics Editor -> Re-Analyze, will solve the problem.
But usually, right-click -> Lyrics Editor -> Re-Analyze, will solve the problem.
As I have already stated, occasionally Re-Analyze does solve the problem, however more often than not I am still left with this 'filler' chinese text instead.
And of course AI isn't perfect, but you won't know how imperfect unless we report it.
发表时间 7 days ago @ 8:39 pm
francky108
Hello, I just retested the latest V2026 B8978. First off, congratulations, because from my point of view, there's a very clear improvement.
So, congratulations, because it must have been a lot of work to achieve this. As I've already mentioned, and others on the forum have already asked, would it be possible to import or display the lyrics as "organized text"?
After a period, we could have a line break, or even better, allow us to type our own alignment and syntax for a smoother display, and also within certain clips.
Because of the sound, some words in the lyrics need to be highlighted individually.
There are still too many blocks of lyrics (despite the clear improvement). These blocks still exist, making it difficult to read.
And in the karaoke effects settings editor, you should be able to adjust the display length and height and manage the font size for better display and therefore better readability.
I should point out that the tests were performed by reanalyzing the system after each update.
And I'm still waiting for feedback on the handling of clips with two singers (one female singer and one male singer), which means different colored lyrics, and also on cases where the same duo doesn't have the same lyrics.
This isn't a minor detail, but a real challenge in managing professional karaoke.
I wish the entire team and all the users who have read this a very Merry Christmas.
So, congratulations, because it must have been a lot of work to achieve this. As I've already mentioned, and others on the forum have already asked, would it be possible to import or display the lyrics as "organized text"?
After a period, we could have a line break, or even better, allow us to type our own alignment and syntax for a smoother display, and also within certain clips.
Because of the sound, some words in the lyrics need to be highlighted individually.
There are still too many blocks of lyrics (despite the clear improvement). These blocks still exist, making it difficult to read.
And in the karaoke effects settings editor, you should be able to adjust the display length and height and manage the font size for better display and therefore better readability.
I should point out that the tests were performed by reanalyzing the system after each update.
And I'm still waiting for feedback on the handling of clips with two singers (one female singer and one male singer), which means different colored lyrics, and also on cases where the same duo doesn't have the same lyrics.
This isn't a minor detail, but a real challenge in managing professional karaoke.
I wish the entire team and all the users who have read this a very Merry Christmas.
发表时间 6 days ago @ 5:59 am