Jump to content

Ultrabeat cutting off transients


Rufuss Sewell

Recommended Posts

I've been having this problem with Ultrabeat cutting off anything before the firt zero crossing. I have a good kick sample that I'd like to host somewhere so that people can experiment and let me know if it's a real bug or if I'm doing something wrong.

 

Is there a file section on this site? It's only 28kb.

Link to comment
Share on other sites

Ok, here's a zip. As I said it's just a single .wav file. Try it in the finder window and then in UB. You'll hear a dramatic difference.

 

It's almost certainly because the sample does not start at a zero crossing and UB is starting the sample at the first zero crossing.

 

Utrabeat is starting the sample from the same place (if you bounce it, you can see is is in phase with the original), but it is making a zero crossing at the sample start.

 

As a test I used the pencil in the sample editor to make the first zero crossing in the sample really far in and bounced it out of Ultrabeat. It's still in phase, but the sample starts at a zero crossing. With a bit of a fade in, even though I had none in the ADSR.

 

See the attached pic, the waveform on top is the original sample, the lower one is after playing it in Ultrabeat.

673105870_Picture3.png.ecc38c0a316dbe98d228b33b2714e6a0.png

Link to comment
Share on other sites

Maybe it's a feature, not a bug...

How long is the period of this difference at the waveform's left end? Of course if the leading edge of a waveform is not as steep as it should be, there is less "klick" at the sound's beginning, but edges are always tricky with time-sliced waveforms, and Ultrabeat does more than just playing a sample: It also can change the sample's pitch and thus makes an additional sample rate conversion necessary. Of course if the sample isn't being transposed, there is no need for such thing, but I think Ultrabeat does the same calculations anyway. Be aware it's at least bad style to have a sampled waveform not start near zero; several playback algorithms might come to different interpretations.

Did you try to add some zero samples at the waveform's start?

Link to comment
Share on other sites

Maybe it's a feature, not a bug...

 

Hi Jope. Interesting. It could be that UB has to modify the sample this way in order to properly calculate envelopes, ie a deliberate quantization of the sample start. In general, it's good engineering practice to have your one-shot samples start with a zero crossing, but I realize thats time consuming.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...