Jump to content


Kawai James

Member
  • Posts

    1,385
  • Joined

  • Last visited

1 Follower

About Kawai James

  • Birthday 02/20/1981

Converted

  • occupation
    Digital piano marketing and materials producer at Kawai Japan
  • hobbies
    Cycling, yoga, nutrition.
  • Location
    Hamamatsu, Japan

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. RIP Peter. My understanding is that he was immensely respected within the industry.
  2. Does Moises perform the stem separation on the computer itself, or is the audio uploaded and manipulated within the cloud?
  3. I've noticed that I get different results when uploading discrete left and right channels of a track compared to when uploading the stereo original. Sometimes the results are worse, other times, such as one of the tracks I'm working on, "Harlem Clavinet", the keys part is more cleanly separated. It seems this is not an exact science, and a fair bit of experimentation is required, but is great fun (re)listening to some of your favourite tunes in this way!
  4. Woah, that takes me back to my Amiga days, 8-bit samplers, and creating music in OctaMED. James x
  5. That's pretty cool. I used Audiostretch on iOS for quite a while to slow down and loop difficult parts when learning new songs, so the ability to do something similar within a tool that can also extract the different parts would be very handy. As it happens, the other day one of my bandmates would like to play the Bobby Womack tune "Across 110th Street", so I ran it through Gaudio Studio and sent him a link. However, he said he would also like to play it a half-step lower so that it can segue into another tune. I noticed that in one of the promo videos for Gaudio Studio, there was no "Piano" option - just a "Coming Soon" button. It seems that the developers are continuing to improve their algorithms, so perhaps we will see improved separation for guitar and keyboard parts in the future? Cheers, James x
  6. I've also been using the Gaudio Studio site - it works really well, and the ability to share the "separated" tracks via a URL with bandmates is really great. There was one tune I tried which struggled to extract the bass part compared to another tool I have been playing with, but other than that, the results have been very impressive. I think it will also be a great tool for pratising, especially for tunes that have a strong keyboard/piano part, as you can essentially mute that part and play-along to the original accompaniment. I would share an example, however I'm unsure about the legalities of this kind of thing... Cheers, James x
  7. Thanks very much for the replies folks! I'll definitely have to check out some of these other options - the MusicRadar article is interesting too! The songs that I play with my bandmates are typically mid/late-60s, however sometimes the sax player will recommend a 50's R&B number. Those source recordings are typically mono, and the reference I receive is often a YouTube link. The other day I spent some time search for cleaner/remastered versions of a tune, and managed to purchase and download the track as an uncompressed WAV, as like dbran I believe that's the best source to use for this kind of processing (rather than a lossy MP3 that has already had frequencies "thrown away"). It's really great fun to run some of your favourite tunes through these algorithms, and listen to the individual instrument parts (or combinations of parts) in isolation. On some occasions, I've heard new things that I had not noticed previously, and enhances my appreciation for the players' musicality. Thanks again! Cheers, James x
  8. Hello folks, Tools that use AI to isolate tracks/stems from songs are nothing new, and I had experimented with them a few times in the past. However, I recently started getting back into this technology, and using it as a way to learn/practise songs with my bandmates. The "Spliteroo" app is quite nice, as it runs on both iOS and macOS, and allows stems to be extracted from uploaded audio files pretty quickly. The other day I sent the parts to a song that our drummer and bass player had been struggling with, and I believe it really helped their understanding of the tune. I'm only using the ad-supported free version which is limited to 4 parts (vocal, drums, bass, others), but I'm a little curious to see what the paid version can achieve with the more advanced AI models. I also stumbled upon Fadr.com which offers a similar service, and can even export a transcribed MIDI version of a separated track. I'm curious how other folks are using these tools in their music creation/performance, and if they have any that they particularly recommend, or have any tips to get the most out of the technology. The quality of the track separation appear to vary considerably depending on the source recording/mix. Sometimes I find it has been useful to upload separated left/right channels of a file instead of the original stereo version. Cheers, James x
  9. This doesn't answer your question, however the P-125a is the version of the P-125 *without* the useful USB audio interface. If you wish to connect your digital piano to a computer or phone/tablet and have the sound from the device output through the instrument's speakers, the USB audio interface is a really useful feature. Kind regards, James x
  10. Well spotted, I was just about to make the same observation. Marco Parisi is a phenomenal musician, and an excellent ambassador for ROLI products.
  11. It's not directly related to Apple Vision Pro, however I watched an interesting take on this topic from Ben Jordan a few weeks ago: Cheers, James x
  12. You may be interested to learn that in Japan, Nord instruments were once distributed by Korg (until October 2016). This contract was then taken over by Yamaha, which remained the Japan distributor until October 2023. Today, Nord instruments are distributed throughout Japan by Kyoritsu group: https://kyoritsu-group.com/news/48055
  13. Wow, that's awesome! The next time you see him, please do pass on my warm regards! Cheers, James x
  14. Sounds awesome, definitely gonna have to check this guy out! Cheers, James x
×
×
  • Create New...