Migrating Desktop Samples Into Obsidian Workflow Questions

T4HT4H
edited February 17 in General chat

Just been working on building soundsets in obsidian and it’s pretty cool. One thing I’m loving is that it remembers the piano octave of the sample patch when you save it, that is 🔥! Just was wondering for cataloging purposes, are custom tags in obsidian possible?

«1

Comments

  • Also, is there a way to have obsidian tempo sync a sample? Play faster/slower at different tempos?

  • And another question, I'm currently taking some desktop stuff i have and migrating over to obsidian. The .wav files have the note mappings (C3, etc) however they also have velocity mappings (for instance 1-32, 33-54, 55-71, 72-90, 91-127). Im aware obsidian can automap the notes but can it also automap the velocities on the .wav files? My initial gut feeling says it probably can't, but I'm wondering what the best workflow method on something like this would be?

  • edited February 15

    I'm afraid i do not have much positive answers for you this time :(

    Just was wondering for cataloging purposes, are custom tags in obsidian possible?

    No at the moment.

    People use hidden "trick" for better cataloging of many patches (which was not supposed to be revealed until Matt will not have proper UI for this, but more people already found it, so it's not secret anymore :))

    You can create manually subfolders - using Apple Files app in "Instruments > Obsidian > User" folder, and put your patches there.
    Then in Obsidian just refresh patches database (in patch browser - at left site tap "MANAGE" and then "Refresh") ..

    Don't forget that that every new patch you create or import goes always directly to "User" folder, so you need always move it manually using Apple Files to proper subfolder and do again "Refresh" in Obsidian patch browser...

    Also don't forget that Obsidian patch browser shows patches hierarchically - if you choose "All" - you see all patches from "User", "Factory", "IAP" .. same way if you choose "User" and you have there some subfolers, you see also patches from all those subfolders.

    Im aware obsidian can automap the notes but can it also automap the velocities on the .wav files?

    No, just notes.

    Also, is there a way to have obsidian tempo sync a sample? Play faster/slower at different tempos?

    So you basically asking about timestretch ? I mean - faster tempo, sample plays faster but with preserved pitch.

    This is not possible for now, although time stretch is on the list (probably together with introducing of audio tracks - but that's just my guess).

    Do you need this for short "single hit" samples or for some longer loop, for example drum loop ? Especially with drumloops surprisingly good works trick with sample start automation, check it here:
    (in that video obsidian is playing drum loop sample and you can hear how it changes speed but not pitch when song tempo is increased)
    https://blipinteractive.co.uk/community/index.php?p=/discussion/355/little-trick-for-timestretched-drumloop

  • T4HT4H
    edited February 15

    @dendy said:
    I'm afraid i do not have much positive answers for you this time :(

    Just was wondering for cataloging purposes, are custom tags in obsidian possible?

    No at the moment.

    People use hidden "trick" for better cataloging of many patches (which was not supposed to be revealed until Matt will not have proper UI for this, but more people already found it, so it's not secret anymore :))

    You can create manually subfolders - using Apple Files app in "Instruments > Obsidian > User" folder, and put your patches there.
    Then in Obsidian just refresh patches database (in patch browser - at left site tap "MANAGE" and then "Refresh") ..

    Don't forget that that every new patch you create or import goes always directly to "User" folder, so you need always move it manually using Apple Files to proper subfolder and do again "Refresh" in Obsidian patch browser...

    Also don't forget that Obsidian patch browser shows patches hierarchically - if you choose "All" - you see all patches from "User", "Factory", "IAP" .. same way if you choose "User" and you have there some subfolers, you see also patches from all those subfolders.

    Im aware obsidian can automap the notes but can it also automap the velocities on the .wav files?

    No, just notes.

    Also, is there a way to have obsidian tempo sync a sample? Play faster/slower at different tempos?

    So you basically asking about timestretch ? I mean - faster tempo, sample plays faster but with preserved pitch.

    This is not possible for now, although time stretch is on the list (probably together with introducing of audio tracks - but that's just my guess).

    Do you need this for short "single hit" samples or for some longer loop, for example drum loop ? Especially with drumloops surprisingly good works trick with sample start automation, check it here:
    (in that video obsidian is playing drum loop sample and you can hear how it changes speed but not pitch when song tempo is increased)
    https://blipinteractive.co.uk/community/index.php?p=/discussion/355/little-trick-for-timestretched-drumloop

    Thanks I’ll try the cataloging trick. Will this let me have multiple tags at once. For instance instead of just ‘User, Vocal’ can I do something like ‘User, Vocal, Female, Reverse’? If so how do I do that?

    Thanks for the info on the auto mapping.

    Yes on the time stretching. Basically I have instruments for other Vst’s on desktop that are loops that aren’t drums they’re vocals at 150bpm .wavs, 90bpm .wavs etc, so they can be played like single shots when composing if that makes sense. But when the tempo of the song is changed on desktop in the VST the sample playback changes with it automatically. I never paid attention to how useful this is but I miss it. Just wondering if there’s any possibility on this.

    Thanks so much for your help with all of this @dendy it’s much appreciated!

  • Thanks I’ll try the cataloging trick. Will this let me have multiple tags at once. For instance instead of just ‘User, Vocal’ can I do something like ‘User, Vocal, Female, Reverse’? If so how do I do that?

    no, this is not about tags - this is just about putting patches to different folders .. try it with few patches and you will immediately understand it .. just create some subfolders in "User" dir, put there 1-2 patches.. then in Obs. refresh database and watch left directory tree - your subfolder will appeart there, and if you choose it at right side you will see just those 2 patches .. if you choose "User" - you will see all patches including those 2 in your subfolder

    Thats all.

  • T4HT4H
    edited February 15

    Also....as I'm going through breaking down these VST's to get them ready for NS2 obsidian, I'm wondering if there's a desktop solution since some of the zoning gets complex, and honestly some of it's just easier with a mouse and keyboard. Is it possible, for example, to make them into an EXS24 instrument and then import it (.exs file) into Obsidian, similar to how can i do VirSyn AudioLayer? I know AudioLayer is an option but due to Auv3 crashing tendencies and memory limits due to iOS I'm wanting to double down and go all in on obsidian. Any help is appreciated @dendy . Thanks

  • edited February 15

    if you can share some of that vocal samples - just one or two as example - i can try a few combinations of that sample strart automation trick, maybe i found for you some setup which will for some limited bmp range works also with vocals...

  • edited February 15

    regarding import - i'm afraid there is no simple way. Currently really just that simple automap by note in file name ...

    there is one more trick how to import a bit faster 3 velocity layers (which is basically maximum what Obsidian can handle, obviously)

    Prepare your samples into 3 subfolders - each witch properly named files ending with root note.

    This way, you simply use automap in each of 3 Obsidian's sample oscillators, just always from corresponding folder.

  • T4HT4H
    edited February 15

    @dendy said:
    if you can share some of that vocal samples - just one or two as example - i can try a few combinations of that sample strart automation trick, maybe i found for you some setup which will for some limited bmp range works also with vocals...

    @dendy I just PM'd you the .wav sample. Thanks

  • @dendy said:
    regarding import - i'm afraid there is no simple way. Currently really just that simple automap by note in file name ...

    there is one more trick how to import a bit faster 3 velocity layers (which is badically maximum what Obsidian can handle, obviously)

    Prepare your samples into 3 folders - each witch properly named files ending witj root note.

    This way, you simply use 3x automap - once in each Obdidisn sample oscillator, just always from proper folder.

    THATS A REALLY GOOD IDEA!

  • edited February 15

    another hints , check hamburger menu in oscillators section - you can copy ALL oscillator settings from one oscilator to other oscillator

    And this is even cooler : with zone selected, tap on that wrench (hope i used proper english word for that icon :-))) - there are some functions which you can apply on all zones, like set volume crossfades between all zones in active oscillator, or turn on/off looping (or switch it to spectral looping mode) .. this saves a lot of taps ;)

  • @dendy said:
    regarding import - i'm afraid there is no simple way. Currently really just that simple automap by note in file name ...

    there is one more trick how to import a bit faster 3 velocity layers (which is basically maximum what Obsidian can handle, obviously)

    Prepare your samples into 3 subfolders - each witch properly named files ending with root note.

    This way, you simply use automap in each of 3 Obsidian's sample oscillators, just always from corresponding folder.

    Happy to report this worked perfectly!

    @dendy said:
    another hints , check hamburger menu in oscillators section - you can copy ALL oscillator settings from one oscilator to other oscillator

    And this is even cooler : with zone selected, tap on that wrench (hope i used proper english word for that icon :-))) - there are some functions which you can apply on all zones, like set volume crossfades between all zones in active oscillator, or turn on/off looping (or switch it to spectral looping mode) .. this saves a lot of taps ;)

    I will definitely look into this!

    @dendy said:

    Thanks I’ll try the cataloging trick. Will this let me have multiple tags at once. For instance instead of just ‘User, Vocal’ can I do something like ‘User, Vocal, Female, Reverse’? If so how do I do that?

    no, this is not about tags - this is just about putting patches to different folders .. try it with few patches and you will immediately understand it .. just create some subfolders in "User" dir, put there 1-2 patches.. then in Obs. refresh database and watch left directory tree - your subfolder will appeart there, and if you choose it at right side you will see just those 2 patches .. if you choose "User" - you will see all patches including those 2 in your subfolder

    Thats all.

    WOW, I just tried this with one Sample and Folder and this SubFolder is Perfect! It’s Clean, sleek, and is really gonna help with organization. I love it. I’m finding more and more about NS2 that is just incredible. Additional unlimited ‘tagging’ would also be nice to see in the future, as it’s been a power feature of Native Instruments .NKS format for finding ‘that perfect sound’ among 1000’s of sounds. Hopefully @Blip Interactive will agree.

  • This was a wonderful exchange of idea’s. Great stuff!!

  • Great tips dendy!!!!

  • @SlapHappy said:
    Great tips dendy!!!!

    Yes, thank you!

  • @drez said:
    This was a wonderful exchange of idea’s. Great stuff!!

    Agreed! This app is revolutionary and @dendy has been on it! Now, If only I could type my file names with a physical keyboard I wouldn’t need the periodic breaks to rest my hands as I’m migrating these samples lol

  • And just to add to this for anybody contemplating making the jump from Desktop to NS2 and wanting to migrate huge desktop sample libraries over to it, the benefits if you’re willing to put the countless hours into the process are worth it! NS2 can handle as many Obsidian instances as you can throw at it, AND you are in an integrated solid state hard drive environment (iPad) where the sample patches load instantly, even faster than SSD USB3 Externals on desktop. So far I’ve exprienced NO LOADING TIMES in obsidian!

  • Could you script it out on your desktop to put everything into a specific folder structure and then just drag that into the nanostudio folder to give you a head start?

  • T4HT4H
    edited February 15

    @drez said:
    Could you script it out on your desktop to put everything into a specific folder structure and then just drag that into the nanostudio folder to give you a head start?

    Yup. Thats what I'm doing actually. Im spending hours finding .wavs and setting everything in its newly created folder that represents each patch. Then, I use RENAME IT to batch erase excess tags off of file names after the note indications on each .wav (instead of 'big piano C3 dry', it becomes 'big piano C3') so NS2 can recognize the notes for automapping. Im also putting all those into their own folders and then .zipping them after in batches, at which point i airdrop them to the iPad and get to work in NS2 to get that side rolling. Tedious, tedious stuff so any shortcuts you can recommend are greatly appreciated!

  • @dendy once I have imported my files into NS2 in the Library folder, and then have turned them into an obsidian instrument and saved it as an 'obsidian' patch, which then saves it in the Obsidian User Folder, am i free to delete the original files from the 'Library' folder without it affecting my obsidian patch for anything?

  • edited February 17

    @T4H yeah - if you save custom patch, all used samples are saved together with patch so you can delete them from their original location..

    Same way it works with Slate banks - if you save bank, all used samples are bundled (copied) into bank.

    But note this is not the case if you just save whole project - until you don't save patch manually, and you save just project, all samples in patches in that project are linked from their original locations. To pack samples with project, use "archive" feature in project load/save menu...

  • @dendy said:
    @T4H yeah - if you save custom patch, all used samples are saved together with patch so you can delete them from their original location..

    Same way it works with Slate banks - if you save bank, all used samples are bundled (copied) into bank.

    But note this is not the case if you just save whole project - until you don't save patch manually, and you save just project, all samples in patches in that project are linked from their original locations. To pack samples with project, use "archive" feature in project load/save menu...

    Perfect! Im like 30gbs of samples in and I need that to be cleared out of the library so I can have that space back lol. I’m like 20,000 individual samples in, and with the auto mapping features they have turned into a few hundred obsidian patches. Lots have the 3 different velocity level samples. So far they are working great! Loading patches has actually slowed slightly, but I think that has to do with my iPad being near capacity so I’ll report back after deleting the source files. Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

  • @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

  • @Stiksi said:

    @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

    I assumed so, thanks for confirming. Multisampling it is until I get the time to understand making synth patches from scratch a bit better. Also, I just discovered the Auto-Sampler inside of Apple’s MainStage and its incredible, it samples your vst or hardware for you on it’s own, playing the notes and all by itself and then recording it!

  • @T4H said:

    @Stiksi said:

    @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

    I assumed so, thanks for confirming. Multisampling it is until I get the time to understand making synth patches from scratch a bit better. Also, I just discovered the Auto-Sampler inside of Apple’s MainStage and its incredible, it samples your vst or hardware for you on it’s own, playing the notes and all by itself and then recording it!

    Yes, it’s almost like cheating, I would use the heck out of that, but I don’t have space for all those samples 😀

  • @Stiksi said:

    @T4H said:

    @Stiksi said:

    @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

    I assumed so, thanks for confirming. Multisampling it is until I get the time to understand making synth patches from scratch a bit better. Also, I just discovered the Auto-Sampler inside of Apple’s MainStage and its incredible, it samples your vst or hardware for you on it’s own, playing the notes and all by itself and then recording it!

    Yes, it’s almost like cheating, I would use the heck out of that, but I don’t have space for all those samples 😀

    Now if only I could find a way to make it switch patches on its own and just run down the whole vst catalog and autosample while I do other things I’d be happy lol

  • @T4H said:

    @Stiksi said:

    @T4H said:

    @Stiksi said:

    @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

    I assumed so, thanks for confirming. Multisampling it is until I get the time to understand making synth patches from scratch a bit better. Also, I just discovered the Auto-Sampler inside of Apple’s MainStage and its incredible, it samples your vst or hardware for you on it’s own, playing the notes and all by itself and then recording it!

    Yes, it’s almost like cheating, I would use the heck out of that, but I don’t have space for all those samples 😀

    Now if only I could find a way to make it switch patches on its own and just run down the whole vst catalog and autosample while I do other things I’d be happy lol

    I have been using Renoise as a VST sound module for NS2 and it has handy multi-sampling which is super nice. If you are of the scripting bent there may be some ways to automate the workflow somewhat (no specific ideas).

  • @Audiogus said:

    @T4H said:

    @Stiksi said:

    @T4H said:

    @Stiksi said:

    @T4H said:
    Also @dendy is it possible to take other synth presets from ‘Massive’ ‘Diva’ ‘Serum’ etc and convert them for use in obsidian?

    I’ll butt in: not directly, the architecture of those synths are different. Multisampling is the easiest way, but it’s always cool to try to recreate sounds in another synth by trial and error. For example, I’ve had some success creating sounds for Obsidian by following tutorials for Reason’s Thor and Ableton’s Operator. You learn so much about a synth that way.

    I assumed so, thanks for confirming. Multisampling it is until I get the time to understand making synth patches from scratch a bit better. Also, I just discovered the Auto-Sampler inside of Apple’s MainStage and its incredible, it samples your vst or hardware for you on it’s own, playing the notes and all by itself and then recording it!

    Yes, it’s almost like cheating, I would use the heck out of that, but I don’t have space for all those samples 😀

    Now if only I could find a way to make it switch patches on its own and just run down the whole vst catalog and autosample while I do other things I’d be happy lol

    I have been using Renoise as a VST sound module for NS2 and it has handy multi-sampling which is super nice. If you are of the scripting bent there may be some ways to automate the workflow somewhat (no specific ideas).

    Thanks for the tip I appreciate it. I just discovered SambleRobot and this allows batch autosampling using midi program change. It’s sort of complicated for the initial set up, but so far with the SampleRobot demo I’ve been experimenting with the batch autosampling in omnisphere, And some Kontakt 5 multi’s. After the initial set up, SampleRobot will automatically sample each sound ‘key by key’ and then move on ‘patch by patch’ by itself.

  • A quick update for those making the migration from desktop with big vst libraries wanting to get them into obsidian and beyond. I bought the full version of SampleRobot 6 Pro. I’m using virtual midi and the loopback application to route the midi and trigger sounds directly from Omnisphere into SampleRobot. It involved manually setting up midi learning in Omnisphere with my selected patches 1-128 however I started with 70 to see how it does. So far it’s working and sampling every 3 semitones per patch and then changing patches and repeating again. Also, you can load a text file of your patch names and it’ll automatically name and number the .wav files for u with the patch and note it’s in. For Example (001 Full Choir C3.wav). This is a great workflow once you get it up and running and I’m happy I can get my day back to myself and let that process run automatically. It’ll take around 4.5 hours to complete!

  • @T4H said:
    A quick update for those making the migration from desktop with big vst libraries wanting to get them into obsidian and beyond. I bought the full version of SampleRobot 6 Pro. I’m using virtual midi and the loopback application to route the midi and trigger sounds directly from Omnisphere into SampleRobot. It involved manually setting up midi learning in Omnisphere with my selected patches 1-128 however I started with 70 to see how it does. So far it’s working and sampling every 3 semitones per patch and then changing patches and repeating again. Also, you can load a text file of your patch names and it’ll automatically name and number the .wav files for u with the patch and note it’s in. For Example (001 Full Choir C3.wav). This is a great workflow once you get it up and running and I’m happy I can get my day back to myself and let that process run automatically. It’ll take around 4.5 hours to complete!

    Very interesting thread. Thanks.

    What takes 4.5 hours to complete? The initial setup?

Sign In or Register to comment.