Importing audio file into Nave sounds very different from original source

edited March 3 in General Vote Up0

Tried a few different sources but the result always sounds very different in Nave. In WaveMapper it sounds much more like the real thing. Has anyone experience with this? Saw a YouTube where it worked with an organ sound, but organ sound is easier mimic. Maybe I’m doing something wrong, but I’ve been messing with it the whole day.


  • 15 Comments sorted by Date Votes
  • I cannot help here but would really like to play with this more also. Today I have some time to experiment.
    I have some JP8000 samples that I made before I sold that synth and would really like to create some wavetables with them.
    If you create anything interesting and successful maybe we could share the results?

  • Sure that would be great! I absolutely love Nave’s intuitive (and beautiful) UI and most importantly- potential seems vast. Hope we can figure out how to import audio and make wavetable out of it, that would make it my favorite synth of all. So far I’m nowhere with this.. googling and watching YouTube didn’t help. Ryan (on YouTube) made a video where he imported an organ and it just sounds very realistic and. “Mee Zanook” also imported a sound that resembles original. The stuff I import sounds like bzzzzz, or similar to pure sine sound.

  • I guess it is related to the Tuning of the sample. If the sample is out of tune to the wave table frames, the cycles will be cut in half.

  • well conceptually, I’d like to start with as close to the original that is being imported, and then... perform magic spells on it :wink:

    So I init the Nave before importing, but it still just sounds like something else. Are you saying that some sounds just won’t work, or that there’s a way to make it work by tweaking something?

  • as mentioned: the imported sample is cut into a sequence of frames.
    It's a shortcut to avoid builing the table from individual slices.
    The original Waldorf waves are a masterpiece of snippet stitching.
    Imho the whole process is not intended to replicate an original sound.
    There may be a doc explaing how the algorithm works, otherwise you may use trial and error with some very obvious sounds to get an idea about the method.

  • said:
    I guess it is related to the Tuning of the sample. If the sample is out of tune to the wave table frames, the cycles will be cut in half.

    So what key (Or single cycle frequency if you like) should they be in ?

  • Well no, I disagree. I think the point should be to sound exactly like the wave imported, now to “summarize” the original to a wavetable format is no simple matter so you can’t expect it to be extremely similar... maybe “close enough”? ;)

    So, let’s experiment! I’m using a small audio with speech in it to be able to tell how close it sounds.

    Here are my steps:
    Initialize Nave and load up a sound
    Don’t tap on the ribbon, just go back to the main view
    Set Wave (this is playhead start), Spectr. and Keytrack knobs to 0.
    Set the Travel to about +25 (this is the speed at which it plays)

    Now press a key on your midi controller and slowly turn the Spectr. Knob counterclockwise. The sound begins to resemble original. (You might need to adjust octave)

    So this is as far as I got, I get a very robotic sounding voice with most of the harmonic characteristic lost. I used a voice to be able to gauge how near I am to the original. But with other source sounds it just feels too different.

    This is not a well documented feature, so maybe we can come up with a good starting point ourselves?

  • edited March 5 Vote Up0

    This is because the genius who makes WAVEMAPPER(wolfgang), no longer works for the people who make NAVE. Nave is prettier, Wavemapper is better.

  • Nave has a more intuitive/usable interface.

  • It all depends how the WaveTable is created, some analyse the audio and recreate this as a sequence of it's harmonic frequencies, this can alter the sound from the original source because it's a more limited frequency sample set. Usually the more complex waveforms can sound radically different, things like speech.

  • @MrSmileZ said:
    This is because the genius who makes WAVEMAPPER(wolfgang), no longer works for the people who make NAVE. Nave is prettier, Wavemapper is better.

    The potential of WaveMapper is amazing, but the interface is very un-fun to use. It would be much more popular with proper UI design. But it actually speeds certain things up for you. Like the envelopes show up when you touch a control associated with it. We’re used to making a link in one place and then accessing envelopes directly elsewhere. So here they have no dedicated page like in (all?) other synths. But then, LFOs are hidden inside a Performance tab... I needed couple of days to get the sense of the whole app, and reading most of the manual and in-app help menus. I think if I use it for about a week I’ll be able to “think” more creatively with it. So far my thinking is 50% occupied with controls, and 50% with the process itself. Where as in Nave it’s mostly just the process itself. I don’t think I’ll ever get it to that level of fluidity, or not in a short time for sure.

    It’s hard to compare these two as they each offer certain unique features. WaveMapper has 3 wave OSC sources, you can even do some sort of FM (according to the manual), but I haven’t figured out how that works yet. Nave got 2 wave OSCs + 1 regular OSC and you can’t use any of them as sources, though in Mod matrix there are plenty of Destinations. WaveMapper has a few unique approaches to shaping the wave. Also from my “basic” experiments, wave analysis seems much better in WaveMapper. However, Nave got this (beautifully implemented) Spectral 3d editor which gives precise controls (where and how loud) to introduce randomness into the harmonics. This feature is what in my opinion, makes it different from WaveMapper, making it a different instrument so you can’t compare them in terms of better. In other words, you can’t create the same sounds in both (like a Les Paul vs Stradivari). WaveNaver would be a cool app :) I don’t regret buying both.

  • I own both as well, I like them both...
    Nave has the better GUI all the way
    Wavemapper is more than just wave table’s also like plex it can use different models and those can be tweaked in real-time...which is insaaaane. Wavemapper definitely sounds the best of the two as well...I find nave to be metallic and thin harsh and cold.

  • edited March 7 Vote Up0

    Nave indeed may feature a lot of expression in the clear, cold, glassy domain.
    But that's first of all related to it's (excellent) filter parametric.
    You easily swipe/slip into the top end which gets very pronounced this way - but it's always well defined and never harsh. Great to add some highs to the mix, but you definitely have to keep the peak tamed if you're after a more 'warm' sound.

  • edited March 7 Vote Up0

    Definitely recommend both Nave and PPG WaveMapper.

    If you want more flexibility importing audio then PPG is the winner.

    "3 types of synthesis - the optimal type for kind of each sounds
    Classic wavetables - for the typical PPG sounds
    Time compressed samples - more authentic sounds, allowing analysis of user samples
    Pure samples - classic sample playback
    Analyzer - convert your own samples into the new format or into wavetables"

  • I just wish you could import 24bit samples into PPG apps, not just 16bit. Nearly all sample packs come in 24bit these days and all my own rips are made at 24bit.

Sign In or Register to comment.