Need some advice on Sample Rate/Audio Buffer Size. i have been using GP for about six months and starting to notice some cracking. Thinking of changing Sample Rate/Audio Buffer Size if my Mac can handle it.
Here is my current set up:
Current Hardware:
Novation 61SL MkIII
Novation 61SL MkIII (Yes, I use two of the same controllers)
PreSonus AudioBox 96
MacBook Air 13-inch Laptop - Apple M4 chip - 24GB Memory - 512GB SSD (My first Mac, have been a PC guy for years)
Your sample rate and buffer size are already conservative: where would you go? I don’t think that a buffer size of 1024 would be playable.
The only thing that you could do in this department is lowering the sample rate to 44.100, but that’s not going to give a lot of headroom.
Have you looked at the cpu usage in Gigperformer? If that’s rather high, you could take a look which vst is eating your cpu. If it’s not too high, then maybe the crackling has nothing with buffer size/sample rate.
That’s not excessive. Could it be the case that the signal is too high? On the wire panel, you can easily spot connections with signals that are too high by setting the wire-style to ‘metered’.
Another thing: sometimes drivers don’t handle sleep mode of the system very well. Therefore (if you don’t do this already), I would recommend to always shutdown your computer and not using sleep and/or hibernate.
(And from here, maybe the mac-users and/or keyboard players have better suggestions).
There are posts in this forum indicating that, surprisingly, Macs can have issues (crackling) if the buffer is set high because it switches to efficiency cores on its own.
So, I would definitely try changing the buffer size 128 (I would also try 44.1 sample rate), as the post above suggested.
Have you tried predictive loading? that might help. I use an M5 Air and/or M4 pro, so both more powerful than the M4 Air, also started with an M1 Pro, and walks used 128 buffer size at 48 or 44.1 without issues, although you do need to keep an eye on the amount of ram used as well as CPU IME.
Ive recently started to get a clicking, which seems to be connected to having audio inputs active. It’s not random, it’s a regular clicking. Once ive disabled the audio inputs it goes away. Might be worth looking at that as well.
Hi, I’ve been playing for about 9 months and have learned a lot about GP in that time. I had an M3 with 32 GB of RAM. I had all my songs in a single gigfile with many sample-intensive VSTs (the M3 maxed out at 100 songs). Now I have an M5 with 127 GB of RAM (lots of RAM was a game-changer for me). Now I have 145 songs in a single gigfile and it runs very well. My sample rate is also set to 44.1 kHz. Audio buffer size is at 160 samples. That’s still playable without losing the feel. 512 samples is absolutely unplayable (for me). I used to get the crackling because the samples simply weren’t being reloaded quickly enough with the limited RAM.
My tips:
VSTs with sample libraries: Use the purge function if possible.
Predictive loading, if the time loss isn’t an issue for you.
Okay. You definitely have an advantage because you know what you’re talking about. 128 was always my standard, but I thought I wouldn’t notice a difference in feel at 160, so I’m making it easier on my computer. So, does the computer actually perform better at 128?
You probably didn’t make it easier on your computer Many algorithms, particularly DSP algorithms, work with data that can be represented by lengths that are powers of 2. If the data doesn’t fit into such lengths, then extra work has to be done to handle the “extra” bytes.
I always though the options were 32, 64, 128, 256, 512…with 128 on the low end and 512 at the upper end of what is generally considered reasonable (at least for keyboards)
If you do not load an existing gig file, it gives you the option to “Open File” bottom left. If you click on it, go to the location on your computer where your gig file is and double click. (I think that’s it).
OK, I was making it complicated. I just saved the new audio settings and just reopened the latest gig file. New audio setting are now in the gig file. Thank you!
Many algorithms, particularly DSP algorithms, work with data that can be represented by lengths that are powers of 2. If the data doesn’t fit into such lengths, then extra work has to be done to handle the “extra” bytes.