• Register
  • Help
Results 1 to 8 of 8

Topic: Question For Sample Developers

Share/Bookmark
  1. #1

    Question For Sample Developers

    Hi, first of all thanks for your time and any suggestions you can add here.

    My question mainly has to do with work-flow and efficiency during the capture and edit process of making a libray, nothing to do with the mapping.

    I\'m just curious as to how you giants go about the physical capturing of the samples .... and the consequential naming/organizing. Do you do most of the work initially in a sequencer/recorder ... in order to keep the timing and organization of each note articulation right in front of you? Or maybe, do you use a montage in Wavelab ?

    I\'m just trying to save myself a little time, cut some corners, if you will. I\'m finding that the actually playing is much! easier than the editing and organization .... pheeeew ... not to mention the mapping.

    Any and all suggestions will be greatly appreciated :-)

    passacaglia

  2. #2

    Re: Question For Sample Developers

    [img]images/icons/rolleyes.gif[/img]

  3. #3

    Re: Question For Sample Developers

    Hey,

    Now I\'m not a sampling giant (I think you need 10+ years for that [img]images/icons/smile.gif[/img] ) but I can answer your question.

    About the naming/organizing. Of course you need to have a plan before you start recording samples, so you know what articulations, how many notes per octave, how many dynamics, etc.

    When you recorded all that onto tape or HD, you\'re crazy (I don\'t mean you as such [img]images/icons/smile.gif[/img] ) if you\'re gonna slice and edit all those samples manually and individually. Unless it\'s just a few.

    Common is to autoslice/split up the big takes and recordings into individual files or marker fields. You can use programs like WaveLab or SoundForge for this. At some point, before or after the slicing, you have to do the selecting - what samples are good, which have to go. If you need any renaming, most of that can be batched (automated). WaveLab has pitch detection that helps naming the samples accordingly (of course still needs checking and fixing). And more general renaming, like adding numbers, or removing characters, can all be automated with the right little renaming tools.

    Does this answer any part of your question? Or not at all? [img]images/icons/smile.gif[/img]

    I have NO clue though what you mean with a \"physical capture\" [img]images/icons/smile.gif[/img]

    Maarten

  4. #4

    Re: Question For Sample Developers

    I\'m just trying to save myself a little time, cut some corners, if you will. I\'m finding that the actually playing is much! easier than the editing and organization .... pheeeew ... not to mention the mapping.
    <font size=\"2\" face=\"Verdana, Arial\">Just a few words of caution; there are no real ways to cut the corners. You will need a lot of time, a lot! For a grand piano library (1.500 samples) you will need 4 to 6 months at least when you work really fast and non-stop (40 hours a week).

    Try using Wavelab for simple tasks (cutting samples from the main recording, naming and detecting the pitch).

    Michiel Post

  5. #5

    Re: Question For Sample Developers

    I\'m not a giant but I have made my own custom sample libraries for my projects. I generally record into Pro Tools at 24 bit on multiple tracks with various mikes set up. Then I go into mixdown mode and blend the result to a 2-track stereo bounce (this is where you can convert to 16bit as well if you wish).

    Usually when I record I\'ll name the audio track various things that will help me keep things straight later on (since the track name gives the audio file its name).
    Have a notepad handy and take lots of notes as to what\'s what.

    In case you haven\'t already done so, you need some pre-production here, pre-plan what you;re going to record, how many layers of what, how many articulations etc. Then set a naming convention. For example, you could assign 01 to sustains, 02 to marcatos, 03 to pizzicatos and so on, then you could call a given sample 02A#4mf.aif and that would tell you that you\'re working with a Marcato A#4 at a mf velocity layer. I\'ve seen it done different ways.

    One big tip, use numbering in the names and ideally have as many digits as you\'ll have in the whole instrument starting with a 0...it will help you keep the files in order in the finder. On the Mac if you call a file 1, then 2,3 and so on, file 10 will sit right after file 1. On the other hand, if you call file 1 \"01\", then file 10 will appear correctly after file 09...got it?

    I would also strongly urge you not to be the king of the world on your first project. Start small and learn from your mistakes. As you get better at it you\'ll develop your own methods which in some ways will rival the great ones\' methods.

    On that subject, don\'t ex[ect too much info to trickle down from the pros, afterall they would only be helping a new competitor!

    Send me a private message if you\'d like to chat some more!

  6. #6

    Re: Question For Sample Developers

    Common is to autoslice/split up the big takes and recordings into individual files or marker fields. You can use programs like WaveLab or SoundForge for this.
    <font size=\"2\" face=\"Verdana, Arial\">I just wanted to say that one of the things that autoslice will do which drives me nuts is that the thereshold for the slice is either too much or too little and hence problematic. I have taken on re-editing the raw samples for a lot of libraries that I have due to the fact that a LOT of them have as much as 20ms of dead space up front! With everyone constantly whining about latency issues it seems bizarre that a lot of commercial samples have all of this \"built in\" latency. I have also found some samples with cut off heads!

    Either way, I choose to go the way of the monk, hand trimming all of my samples.

    Oh....and let\'s get rid of the usual 2-5 seconds of dead space at the end of the samples while we\'re at it. I have monitored with my speakers blasting (Mackies 824) and full level on my mixer and I have found a lot of dead air at the tail end of the samples. I don\'t mind 1-2 seconds just to be safe, but I have seen as much as 10 seconds on some commercial libraries. Overall an 100meg instrument could be reduced to about 60 megs with some manual trimming without compromising the actual sound!

  7. #7

    Re: Question For Sample Developers

    WOW!

    thanks gentleman!!! hopefully this thread can stay alive a little while for a proper discourse ;-)

    exactly what i needed.

    passacaglia

  8. #8

    Re: Question For Sample Developers

    What a little bump can cause [img]images/icons/wink.gif[/img]

Go Back to forum

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •