Loop start point issue on certian transients/hits

[Art vandelay]

I'm having a very frustrating issue with drums loops in SparkLE. I'm trying to break up a loop into multiple slices across dfferent pads, but Spark seems to have a problem with certain hitpoints, because when I -for instance- add a drum loop and change the start point to the snare hit, Spark seems to add a little delay in front of it (enabling the metronome confirms the snare is a little late).

Even when I fully zoom in to the snare hitpoint and make sure it is absolutely hitting the transient of the snare, Spark isn't detecting it properly and seems to add a bit of silence in front of it, causing the snare to fall 'out of sync'.

It's not the sample, of that I am sure, because I also noticed that when I zoom in a bit more and keep moving the start point up a bit, I hear no difference untill I go far enough that I do hear a difference, but then it has cut off the transient altogether.

It's as if Spark -for some reason- doesn't allow the transient of the snare to be the absolute starting point of the sampler/pad...very frustrating.

Is this a known bug? Can anyone else reproduce?

[Terrym]

Hi , I am not having this problem ,what samples are you trying to load into spark whats your set up ? are dragging from cubase the samples? or from your hd.


terrym

[Art vandelay]

I'm adding the sample from my HDD into Spark standalone. I've attached an example loop. To reproduce:

1. Simply import the loop to a pad in Spark
2. Set the start point of the loop to the first snare hit (and set the end point to the end of the snare hit)
3. Place the Snare hit in the stepsequencer on steps 1,5,9,13
4. Enable metronome
5. Playback

You will notice a delay from the snare hitting after the metronome.

Furthermore, I've attached a screenshot showing how Spark adds a bit of silence in front of the snare hit. I was able to demonstrate this by exporting the snare hit, re-importing it into Spark and viewing the waveform, which shows a noticable gap in front of the snare hit, where there shouldn't be any!
 

[Terrym]

HI art, thanks for the info ,so i tested your sample and cannot reproduce the snare missing the metronome .it seems to be in time for me .
but the exporting the pattern as a wav where you have used a sample loop and edited the loop does leave a very small gap at the beginning of the exported loop .
I did try a few of my own loops  and it did the same even though you could just trim the front off to solve this issue.

i have attached my spark test project load it into your spark and see if its still out of time .
ill do some more testing as my other projects export ok so it may be a bug with just a single loop in a project .before i report this for you to arturia , ben is on leave so i will email JB or kevin about this.
hope this helps a bit


terrym

[Art vandelay]

Thanks for your support. That's very weird that you can't reproduce. I tried your test project, and I'm getting the same issue with that project. 

I have a Presonus Firebox (with ASIO driver loaded) audio interface and get very good latencies with Cubase 7.5, so that can't be the problem.
I also tried the onboard audio interface with different kinds of settings/drivers, but still the same problem.
I'm also getting this issue on my other system (Intel Core i3 laptop with Windows 8 ). My main DAW is a Core i5 running Windows 7 ultimate.

Note that I'm getting this issue with different samples/projects. It's not just related to a single loop or project.
For clarification purposes, I've attached an audio sample I recorded so you can hear what I'm actually hearing (the snare and the metronome out of sync).

Also, I'm running Spark 2.0.1 (build 400, rev: e345467) 32 bit.

[Art vandelay]

One more thing I noticed is that when I import a loop which is causing issues, then export the loop (in its entirety) as audio, and re-import it into Spark, I don't get this issue. I can then set the start point the the first snare hit and it'll be perfectly in time with the metronome.

For some reason, Spark doesn't like the Original audio file and exporting it from Spark seems to 'fix' it. I did notice that my audio interface driver was running at 48000 sample rate and that the audio file is at 44100 sample rate. So I then changed my audio interface driver to 44100 sample rate to match the sample rate of the audio file, but unfortunately it hasn't solved my issue.

What also caught my attention, is that spark creates additional files when I import audio files. For instance, if I have a sample called 'drumloop1.wav', Spark will create a file called 'drumloop1.wav.44100'. And it will create a file with '.48000' at the end when I have my audio interface driver set to 48000 sample rate. Why is this?

UPDATE:

One more update before I head off for the night: I found out why my 'export audio -> re-import audio' workaround works...it's because the original drum loop was 2 bars long and when I exported it, I only exported it 1 bar long. I just tested this, and indeed...if I export the same loop also 2 bars long, and re-import it into Spark, I get the same issue. Export it as a 1 bar loop and re-import it, the issue disappears.

Seems to me that Spark has trouble with properly handling hitpoints in drum loops longer than 1 bar...atleast in my situation. Hope this additional info can help get closer to a solution, because it's a pretty massive dealbreaker for me t.b.h.