Battery 4 is skipping when Directed Out

AlienObserver

Hello all!

I'm having a weird issue. I load up a factory kit (Garage LE in this example) and I want to direct out the Splash samples. If I do this, some samples start choking/skipping. If I switch the output to master, the samples are behaving normally again. If I change back to Direct Out, they start skipping again... The (even) weirder thing is that the samples that get skipped are the same ones every time I press play. If however i move the start cursor slightly, the "problematic" samples are playing fine, but others "break" and start having problems...

I have a screen capture video showing exactly this, but apparently I'm not allowed to post links yet..

Any idea what causes this and how to solve this?? My DAW is Reaper and I'm using battery 4.3.0

Thank you in advance

Uwe303

Hello,

have you tried to set the bpm down, just for researching.

AlienObserver

Thanks for your reply!

I tried to set the bpm down, doen't solve the issue.

This issue is happening even in standalone (tho, I can't record and pin point a note.. But if i keep pressing the splash, some will eventually skip (i'm not triggering them super fast, I'm playing around 55bpm in standalone, I'm not reaching the 32 voices, not even close, i'm reaching 5 max). (this is true for other samples as well, not just the splash. It happens on the ride as well.)

Plus, I forgot to mention the (even-even) weirder thing: When routing out of the master, as I mentioned, everything is fine. If I route to Direct out (let's say 3/4) the problem appears. BUT if i route the sample to Bus1 and the Bus1 is routet to direct out 3/4, the problem isn't there... Basically, if the sample is routed to a Direct out, the problem appears. If it's routed to the Master or a Bus, the problem isn't there, even if the Bus itself is routed to a direct out.....

Uwe303

i tried a bit in maschine and everything works as expected, no matter hoe i route it, through bus or direct out or master - i asked a colleague moderator to give you the right to post links but it will take a bit time, maybe it then get´s a bit clearer. Maybe you can meanwhile describe your system a bit, os and hardware ....

AlienObserver

Thank you so much for your time and attention!

My system is as follows: Windows 10 Pro 22H2, CPU i7 7700k, Battery (4.3.0) is installed on C drive (samsung 960 evo), Battery library is installed (stupidly now that I think about it) on D drive (wd1003fzex) (I can try to relocate the library to the E drive (a samsung 860 qvo) (but again, if it was a hard drive issue, I would think that the problem would be present all the time, even if i routed through the master, not just be when i use direct out..). My audio interface is an Audient id14 mk1 (usually running on 1024 buffer size and 48kHz sample rate)

As I said, this issue is happening even in standalone. But, my main issue is that is happening when working in my DAW, Reaper v6.82.

I'll try to rewrite the issue to give as much details as possible. (thank you so much for the request for me to be able to post the video, I think that it will help as you'll see exactly the problem).

What is happening is as follows: I create a track. I put Battery 4 on it (I have the option to chose Battery as a VSTi and VST3i. I've tried both, and the issue is the same). I create tracks to route the Battery outputs (in Reaper you can do this automatically, on the FX window right click Battery and then "build multichannel routing for output for selected FX. I've done this, and I've tried to create the new tracks myself and then manually send the audio to the new tracks. Both methods result in the same outcome).

Now, in Battery, I load up a factory Kit. In my current project I'm working on, I load up the Garage LE Kit. So far, everything works as expected. Then, I select a sample (this issue is obvious in samples that have some sustain, like the splash and rides) and right click, Output, Direct Out (doesn't matter which one I choose, let's say 3/4 to keep things simple). I trigger the sample with my midi keyboard, everything works, that is the audio is indeed being routed into the track I've set to receive audio from the 3/4 output of Battery.

Now, I can record a midi item. At (what it seems like) random, some hits get "skipped". The sample is getting triggered, but it sounds like it stops abruptly (I've loaded up the default kit, there isn't any choking on the splashes or the rides, and they aren't getting choked by other samples, the problem exist even if i play just the splash at about 40bpm. So the sample starts normally and then suddenly it stops with no other sample getting triggered. What I found (I've captured this on the screen capture) is that if i head over to the editor tab on battery, i see the sample getting triggered, I see the play line activated and going through the sample, and then it just skips a part of it, it jumps in front, so the sustain isn't played completely, but instead the play head jumps towards the end of the sample.

If i have recorded the midi item, whenever I hit play, the same notes have the problem. If I right click the cell and change the output to the master again and hit play, there is no issue whatsoever. If I change back to direct out 3/4, the issue re appears, with the same notes skipping. If I route to a Bus, the problem disappears (as I said, even if that bus is routed to direct out 3/4).

Thank you again for your attention!

Uwe303

i tried again in maschine with some longer samples in the kit and everything worked as expected, then i made a reaper project and everything works perfect there too. No jumps or noises, clicks, whatever

AlienObserver

I uninstalled Battery and then reinstalled it, I relocated the Battery Library to my E drive (the samsung 840qvo, instead of the HDD D drive) and the problem still exists...

Do you have any suggestions on what else could I try?...

AlienObserver

https://drive.google.com/file/d/1BQQ-pNIfFmeuB47iF7aNKBWWVJnwao78/view?usp=drivesdk

The video showing the problem

AlienObserver

Update:

I installed Battery on my laptop as well to test this. The problem appears on my laptop as well, with a clean and new install of Battery and its library.

The laptop is running Win 10 Pro 22H2 (so, same as my main pc, although the laptop at the time of testing run 19045.3324 OS build, and my desktop is running 19045.3448) but has a weaker i5-5200U CPU, and I was using an audient EVO 8 for the test, again on 1024 buffer at 48kHz..

Uwe303

Wow that's absolutely strange, can you maybe uploaded the reaper project, and tell me what kit/samples you use, maybe also upload the samples. Then I can test on my systems. One with win 11 the other with win 10 all latest updates.

AlienObserver

This issue is present on all my projects. I created one super simple project as an example to send you.

I'm using the stock Battery 4 library (native access says its the 1.0.1) and using the Garage LE Kit.

Battery_Problem.zip

I've screen captured how Battery behaves on my end on this exact project and I'm putting the video in the zip file. As you can see, if i play from the beginning of the project, the 4th, 10th and 11th hits are problematic. I've placed markers just before the 4th and 10th hits, to mark them (and change the play cursor position accurately). In the video:

1) The Splash is already routed to Direct out 3/4, and in Reaper track 2 (named "Dir.Out3/4") is set to receive the 3/4 output of the Battery track. Playing from the start of the project, the 4th, 10th and 11th hits have the problem.

2) I then route the cell to master out, press play and the problem is gone.

3) Route the cell to Bus 1, and route the Bus 1 to direct out 3/4. Press play and everything is normal, no issue.

4) Route the cell again to direct out 3/4, the problem reappears..

5) Move the cursor to marker 1, and press play from there. Now the previously-problematic 4th hit has no issue, but the previously-good 6th, 8th and 9th hits have the problem, and the 10th is good...

6) I moved the cursor to marker 2 and pressed play from there, again the 10th and 11th hits are problematic....

7) From marker 2, I route the cell back to master, and now the 10th and 11th hits have no issue... I routed again to Bus 1, and again no problem...

I did this to demonstrate that which hits are problematic are somehow dependent on the starting position. This however doesn't only occur when playing back a recorded midi item, it happens on real time play as well, I just don't know which hit is going to get skipped. I play normally, and some random hits get skipped.

Thank you so much for your attention and time. I really hope you can find what causes this, because it drives me insane... As I said, I'm using the stock Library and stock kit, without any changes. If however you want the samples, please give me instructions on how to best do this. Do you want the nbkt file of the kit? Or the actual wav sample?

AlienObserver

Hello again.

I did more experimenting today, and I found some interesting things.

I was using the project that uploaded yesterday, and experimenting with settings and kits. In the project that I uploaded, the problem only happens when the cell is set on DFD (on the Main tab/ Engine). If I deactivate the DFD button, the problem goes away (and if i reactivate it, the problem reappears). I then experimented with other Kits (all from the stock Battery 4 Factory Library). I found that the problem can't be reproduced with all the kits. But, I found that the same problem can be reproduced on the Alkaloid Kit, specifically with the Crash 1 and 2 samples. So, if you load up the Alkaloid Kit, (drag and drop the Crush 1 sample from cell A12 to D12, only to trigger it by the midi item already recorded on the project), press the DFD button (loading the kit, it comes with DFD deactivated, while the Garage LE kit comes with DFD activated) and direct it out to 3/4, the same problem appears, on the same exact hits. (Obviously, I can't and I didn't test ALL the kits of the library, that would take too much time. I just found it interesting that this is not happening just to the Garage LE kit that I originally found the problem, but other kits as well.)

I read that the DFD button is a "direct from disk" option, to save on RAM. While this would make some sense as the cause for the problem, there are some problems with this theory. Firstly, as I said I relocated the Battery Factory Library to my E drive, a Samsung 860 qvo SSD. I've never had a problem with this drive, and it's fast enough to not have such problems. But most importantly, this wouldn't explain why the problem goes away when the sample is in DFD but directed to the Mater or a Bus, rather than a direct out. If the problem was in fact the ssd, I would imagine the problem to be present regardless of how the sample is routed in Battery...

Uwe303

I don´t get this visual jumping stuff, but there is definitely something wrong, after switching back and forth the routing the samples get cut, until i switch of dfd or change the engine to strech. To me it seems that there is no release phase anymore.

Uwe303

The native team has noticed the issue. There was one thing, after opened your project there was a gap in the velocity mapping between the first and second sample, have you changed it maybe before saving the project.

AlienObserver

Thank you so much for your time, attention and effort to get this shorted! I really appreciate it!

I'm assuming you mean there's a gap in the velocity mapping for the Splash that I used in that project, correct? No, I haven't changed it. Opening the project I sent, i see that the first sample (from weakest to strongest) has a velocity mapping from 1 to 74, the 2nd has 75-99, the 3rd has 100-114, the 4th has 115-123 and the 5th has 124-127. That's weird tho that you see a gap... But I guess you can test this, by loading another kit, and then loading the Garage LE Kit again. When changing kits, the problem still exist on my end, it happens with a fresh load of the Garage LE Kit...

Uwe303

Strange - I had to bring the lowest velocity of the second sample down to 75, but it was quickly done. But anyway the bug is confirmed and noticed by native. And in my case I have to switch the routing back and forth then the issue appears, not after loading the kit directly. For now I would use the other modes or disable dfd mode, there is enough ram in todays computers.