64bit vs 32bit. Mac M2 Vi Major audio glitching with HI track count!

[grayedout]

I am new ish to Cubase 12 (latest version) and also I Have been testing out HALion 7 , hoping to move over to this as my main sampler for my own samples , still be using hundreds of Kontakt Libs for factory stuff …

anyway I have been testing track counts and found major problems with anything over 300 Tracks of HALion 7 , and around 480 tracks of Kontakt 7 . both with the same very simple patch loaded

I thought I would check in Logic (latest version) and was shocked to fine 600 + tracks was no problem at all HALion or Kontakt 7 .

I tried some setting in Cubase and found 32bit mode fixed this and can no load over 1000 + HALion 7s

IS there a reason to use 64bit , on a new M2 Mac ?

[st10ss]

What you mean with 32 bit mode?
32bit precision?
64bit precision is “doubled” precision with doubled system load.

If you have no idea what it’s all about, leave it at the default setting.

[Phil_Pendlebury]

@st10ss

64bit precision is “doubled” precision with doubled system load

This is absolutely not the case. There are many factors involved. It’s a huge subject tbh but certainly 64 bit precision does not mean it will double your system load.

There’s a nice (lengthy) discussion on it here if you want fine details:

Cubase Processing Precision, 32 Bit or 64 Bit?

Under the hood, there is very little difference between a VST instrument, and a VST effect. Whatever holds for one, usually holds for the other. There is a small effect in performance, but it’s unclear which direction the effect will have in each case! On the one hand, 64-bit means more bytes in each buffer, meaning the L1 cache will see higher pressure, meaning large sweeping processes may run slower. On the other hand, 32-bit may mean that plugins that prefer 64-bit, have to convert to 64 b…

[grayedout]

I am new to Cubase 12 , but was using it 25 + Years ago Cubase Audio 0.9 ! I am not saying 32bit gives twice as much power over all but with hundreds of HALion 7 or Kontakt 7 , there’s a massive performance gain !

I have been running lots of tests and found On my system HALion 7 run around 380 tracks in 64bit then things start glitching, in 32bit , I can run over 1000 ! on my 2023 MacBook Pro 96Gb 8Tb

I have read lots on the forum on 64 vs 32bit before writing this Thread , but couldn’t find anything on Vi or HALion and 32bit mode

[grayedout]

I agree and did read this , but I think (on a Mac a least) I have found a massive downside to 64Bit , if your running hundreds of Vis .

[grayedout]

Just ran some real world tests with a project I finished a couple of months ago , at the end of a score I print everything (Vis midi etc) for the final mix in Quad. only had a play for a hour but everything seem to be happy in 32bit.

this is how things look audio only running 32bit vs 64bit

I then re enable the Live Midi - Vi tracks , this is were 32bit made a big difference ,

also timed opening and closing the Project , they took the same time around 2.30mins

[Johnny_Moneto]

I kinda expected the opposite, 64bit giving you a better performance than 32bit.
How much memory (RAM) is in your computer?

[st10ss]

@Johnny_Moneto

I kinda expected the opposite, 64bit giving you a better performance than 32bit.

It’s not about 32bit vs 64bit architecture, it’s about calculation precision.

@Phil_Pendlebury

This is absolutely not the case. There are many factors involved. It’s a huge subject tbh but certainly 64 bit precision does not mean it will double your system load.

Of course, you are right, and I was simplifying things too much.

@grayedout

Again, if you are not understanding why there are differences, it would be safe for you to use 32bit precision.

And the decision of 32bit or 64bit precision is not a performance related decision.
At least to the people who know.

[Johnny_Moneto]

@st10ss

It’s not about 32bit vs 64bit architecture, it’s about calculation precision.

You chose to look at the world through your glasses. Doesn’t mean that you always see everything sharp.
Furthermore: The OP’s concern is actually performance, not precision.

[st10ss]

But the Apple M processors can’t work in 32bit anymore.
That’s related to RAM, the processing precision is not directly related.

@Johnny_Moneto

Furthermore: The OP’s concern is actually performance, not precision.

That’s why I answered it is safe to choose 32bit.

[grayedout]

Hmm 380 HALions vs 1000+. Not performance related … yer it is . With audio 64 vs 32 nothing in it , with VIs it’s massive , and needs some investigation. I am asking steinberg

[grayedout]

It’s not cubase running 32bit , it’s the audio engine side , if your not using hundreds of Vis , then forgot about it , I have just spotted that 32bit takes the lid of any CPU problems I had , with over 1000+ tracks of HALion and still going

I can not find any other thread on this , so stay tuned I am trying to find out why

[jmtapadas]

@st10ss

It’s not about 32bit vs 64bit architecture, it’s about calculation precision.

This. It is really not architecture related. Cubase doesn’t even run 32 bit plugins.

[st10ss]

@jmtapadas

Cubase doesn’t even run 32 bit plugins.

Sorry, that is completely off-topic!

[grayedout]

I think I should summarise this, including the reply from Steinberg below.
For anyone using lots of virtual instruments HALion 7 and Kontakt 7 , you will get double the amount if you run @ 32bit Float on a M1 or M2 Mac.

This is a list of
32Bit Vis List

64bit Vis list (as you see none of the main Vis or 32bit

how to show 64bit only Plugs

From Steinberg:

I suppose you mean the difference between audio processing precision to 32 bit float or 64 bit float.
Cubase 9.5 - 64-bit double precision audio processing
Processing Precision

By setting the audio processing precision to 32 bit float or 64 bit float, all channels are processed and mixed in 32-bit floating-point or 64-bit floating-point format.
A processing precision of 64 bit float can increase CPU load and memory consumption.

To show all plug-ins that support 64-bit float processing, open the VST Plug-in Manager from the Studio menu, and activate Show Plug-ins That Support 64-Bit Float Processing in the Display Options pop-up menu.