Reason 9.5.1 now available!

[dvdrtldg]

I'm at work so can't test but has anyone with a mac noticed and improvement?

I just loaded a huge session that had gotten really slow & crackly before the upgrade, now it's running fine. Still a big CPU load but definitely way better. This is on a 2017 MacBook Pro with 2.7 GHz Intel Core i7

[spikey]

THANK YOU THANK YOU THANK YOU!!!!!


Please PLEASE let us know on this! (BTW- If it's related to the UAD side only, why does my Cubase 9x and Logic X run just fine with the UAD plugins?)

IMac 2.6 Ghz, 16 gigs Ram, Sierra OS.

They didn't say it wasn't something for Propellerhead to fix, but just that it's not an overall Reason performance problem: it's affecting UAD's plugins specifically, so Propellerhead has contacted UAD so they can learn more about what might be going wrong and figure out why Reason isn't handling them as well as it should be.

Well ok then, it's not an "overall Reason performance problem" (for Reason), but it IS affecting UAD plugins specifically. How that? No matter how it's worded Reason still buggers up UAD plugins even after the last patch, and (as I stated) these same plugins work fine in Logic X and Cubase 9. Im sure they will eventally fix it, but right now it's still broken and not on the UAD side.

[tumar]

With just one instance there is no way HT would help with any plugin, unless that plugin is coded specifically to spread out a single instance across threads. Diva does this to some extent, but doesn't always play well with a hosts HT implementation.

HT comes into play and can have a big benefit when you are using lots of different tracks and plugins. But with a single instance, it can't help, and might even hurt. I think that's what you are seeing.

Checked again. NI Kontour, patch "At the gates". The only difference is option "Use hyper-threading audio rendering". When ON, playing any notes produces ticking noise, sounding a bit like muffled fireworks. When OFF, everything goes fine. Theoretically you're right, but I think in practice Propellerheads messed up something with code.

[swamptooth]

With just one instance there is no way HT would help with any plugin, unless that plugin is coded specifically to spread out a single instance across threads. Diva does this to some extent, but doesn't always play well with a hosts HT implementation.

HT comes into play and can have a big benefit when you are using lots of different tracks and plugins. But with a single instance, it can't help, and might even hurt. I think that's what you are seeing.

Checked again. NI Kontour, patch "At the gates". The only difference is option "Use hyper-threading audio rendering". When ON, playing any notes produces ticking noise, sounding a bit like muffled fireworks. When OFF, everything goes fine. Theoretically you're right, but I think in practice Propellerheads messed up something with code.

tried that kontour preset...
nothing out of the ordinary on my end. will be testing more over weekend.

[CephaloPod]

With just one instance there is no way HT would help with any plugin, unless that plugin is coded specifically to spread out a single instance across threads. Diva does this to some extent, but doesn't always play well with a hosts HT implementation.

HT comes into play and can have a big benefit when you are using lots of different tracks and plugins. But with a single instance, it can't help, and might even hurt. I think that's what you are seeing.

Checked again. NI Kontour, patch "At the gates". The only difference is option "Use hyper-threading audio rendering". When ON, playing any notes produces ticking noise, sounding a bit like muffled fireworks. When OFF, everything goes fine. Theoretically you're right, but I think in practice Propellerheads messed up something with code.

OK, I just checked that patch. I compared the CPU usage in Activity Monitor with hyper threading ON and OFF. According to Activity Monitor, it did use a lot more CPU with HT OFF. Like 2X as much. Also, at least according to Activity Monitor, with HT ON, the load was distributed across 8 cores, even with only one instance of Reaktor 6 (but using more total CPU). Not sure how that works since I didn't think that was possible with only once instance. Not sure how any of this works anymore!

[jonheal]

Checked again. NI Kontour, patch "At the gates". The only difference is option "Use hyper-threading audio rendering". When ON, playing any notes produces ticking noise, sounding a bit like muffled fireworks. When OFF, everything goes fine. Theoretically you're right, but I think in practice Propellerheads messed up something with code.

OK, I just checked that patch. I compared the CPU usage in Activity Monitor with hyper threading ON and OFF. According to Activity Monitor, it did use a lot more CPU with HT OFF. Like 2X as much. Also, at least according to Activity Monitor, with HT ON, the load was distributed across 8 cores, even with only one instance of Reaktor 6 (but using more total CPU). Not sure how that works since I didn't think that was possible with only once instance. Not sure how any of this works anymore!

It's magic. Don't bother trying to look behind the curtain.

[CephaloPod]

OK, I just checked that patch. I compared the CPU usage in Activity Monitor with hyper threading ON and OFF. According to Activity Monitor, it did use a lot more CPU with HT OFF. Like 2X as much. Also, at least according to Activity Monitor, with HT ON, the load was distributed across 8 cores, even with only one instance of Reaktor 6 (but using more total CPU). Not sure how that works since I didn't think that was possible with only once instance. Not sure how any of this works anymore!

It's magic. Don't bother trying to look behind the curtain.

Probably good advice at this point!

[AttenuationHz]

Question does not using Kontakt solve the problem? If you take any VSTi/VST connect that to a mix channel copy paste the mix channel and the VST 5 times then double that to 10 copies. Select the 10 and ctrl move the 10 to copy/paste them. How many instances of that VST can you reach before computer too slow appears. Bear in mind at least one bar of DSP is needed to play move menu's sequence etc. Are all cores used?

[DJMaytag]

I am still seeing the "VST Windows get keyboard focus" or whatever the option is called. In another topic, it was noted that this was NOT supposed to be seen for MacOS users, and we can't click/unclick that option anyway.

[Galaxy]

I am still seeing the "VST Windows get keyboard focus" or whatever the option is called. In another topic, it was noted that this was NOT supposed to be seen for MacOS users, and we can't click/unclick that option anyway.

Noticed this too. Didn't know we're not suppose to see this? Why wouldn't Mac users have a menu option that pc users have? Also can't figure out how to un grey out this option. What's it even for?

[ProfessaKaos]

Got an i7 6800k which supports hypethreading and my Reason is running better CPU wise without Hyperthreading turned on inside Reason.
Dont know if i need to activate hyperthreading on my pc tho ?

Same issue here with my 6800k 6 cores 12 threads.

[Heater]

I'm at work so can't test but has anyone with a mac noticed and improvement?

I just loaded a huge session that had gotten really slow & crackly before the upgrade, now it's running fine. Still a big CPU load but definitely way better. This is on a 2017 MacBook Pro with 2.7 GHz Intel Core i7

Thanks.

I've notice a small improvement on my MacBook Pro 2015 i5 model. I'll keep testing.

[Carly(Poohbear)]

I am still seeing the "VST Windows get keyboard focus" or whatever the option is called. In another topic, it was noted that this was NOT supposed to be seen for MacOS users, and we can't click/unclick that option anyway.

Noticed this too. Didn't know we're not suppose to see this? Why wouldn't Mac users have a menu option that pc users have? Also can't figure out how to un grey out this option. What's it even for?

It's to give the keyboard focus to the VST device, e.g. I loaded a VST that required me to login to activate it from the VST device it self, I'm not able to type into the device as Reason was capturing the keystrokes (i.e keyboard shortcuts), with this option enabled I then could type in my username and password etc.

[Ostermilk]

Something to check on Windows: make sure hyperthreading is enabled in you BIOS. Mine was, but it could just as easily have been set to disabled.

It's a good point. Most PC's with HT capable CPU's will likely have HT enabled in the BIOS by default but it's certainly worth checking anyway.

The other thing to check of course is whether your CPU actually supports Hyperthreading, mine for example is an i5 6600K which does not. It would be pointless then for me to enable HT in Reason.

Even so I've tested my machine with Reason's HT support turned on and off and I'm happy to report, here at least, that the changes haven't caused any issues, i.e performance certainly hasn't got any worse.

[MattiasHG]

Thanks for all the feedback! It's worth repeating that Hyper Threading doesn't mean that anything takes less resources, only that Reason has more resources to work with. It's also important to point out that HT will not always give a performance boost, it might do the opposite and that depends on your set-up. That's why it's possible to turn off.

As I mentioned, hyper threading only really makes sense when you have a lot of parallel processes, like a song project with multiple tracks and devices that are not in series.

Finally, watching your CPU monitor is not always the best way to understand what's going on. In fact, Reason can "sure up" some idle CPU so it's ready to use when it's needed—making it look like you're running something very heavy when just looking at the CPU monitor. To test what works for you, try playing projects! Play them with HT off and HT on, how do they sound? Do you hear glitches or drop outs?

There are so many different moving parts in performance. Try increasing or lowering the buffer size in both modes, what happens? Are you running at 44.1kHz? Are you running other apps? Do you have any system utilities or config files for your CPU that could interact?

In the end, there will likely never be a scenario where everyone on every imaginable system has great performance but we'll keep an eye on things and improve where we can. Make sure you enable sending us usage statistics in the Reason preferences, that way we learn more about the performance issues and you're helping the entire Reason community.

So, something weird is happening with u-he Diva. I have a song that uses several instances, and get very different behavior based on hyper threading settings in Reason and Mulitcore setting in Diva. In order of best playback:

1. Reason Hyperthreading OFF; Diva Multicore OFF
2. Reason Hyperthreading ON; Diva Multicore OFF
3. Reason Hyperthreading ON; Diva Multicore ON (pretty much unplayable)

So, in this scenario, Hyperthreading definitely made things WORSE.

When plug-ins have their own multi-core implementation, incl. hyperthreading, then it's really like running another heavily optimized application so this can interfere with the host. You missed the scenario with Diva multicore ON and Reason Hyperthreading OFF. Since this is one instance of a heavy plug-in, that'd likely give you the best result.

[RoryM0]

Is this an update for Reason Essentials too?

[JiggeryPokery]

I don't particularly understand hyperthreading, but a quick test on the Ryzen 1800X here with the R8 Benchmark file, at 11ms latency.

This is a completely non-scientific test, I had other progs running in the background so it's not indicative of full Ryzen performance when just Reason is running, but it's kinda interesting.

Hyperthreading disabled, Task Manager's CPU Utilization is at 22% when the track starts. First hint of crackle at 22s in, track stops at 48s.

With Hyperthreading enabled, Task Manager's CPU Utilization figures instantly doubles, so, 44%, but the does track play longer, up to 1'07, but first hint of crackle is still around 22s in. (Obviously sounds terrible long before 1'07)



Repeating the test running at 30ms latency:
Hyperthreading off: crackles at 44s; song stop at 55s

Hyperthreading on: crackles at 1'06; song stop at 1'10, which is pretty impressive given I think I only got to about 1'16 in the past with less stuff running in the background, but crucially, it doesn't start to crackle until the song stop position at 11ms latency.


So a higher the Utilization figure alone is not necessarily indicative of the overall performance being worse; it's arguably quite the opposite, especially if one is prepared to increase latency slightly as the arrangement gets more complex, or more FX added.

My layman conclusion: Hyperthreading may well be very useful to enable on the Ryzen, but don't panic about the higher Utilization figure in Task Manager.

[tumar]

I'm back.
1. iMac 27" i7 3.5 ghz (Maverick) - after two restarts and permissions-repair everything sounds fine, no crackles, one bar CPU with huge pads etc.
2. Macbook Pro 15" 2015 i7 2.2 ghz (El Capitan) - this is f*** incredible! Wow, just wow! Tested with Kategra's "R8 Benchmark 44.100" file. Reason 9.5: "Computer too slow to play song" at 0:00:38:081. Reason 9.5.1 with HT switched on: stopped at 0:01:01:579 !!!


EDIT: But still Macbook Pro is pitifully weak to run new Reason for live gigging. Two-three NI Massive patches in Combinator and crackling, crackling...

[Loque]

Looking in the above test results, i might give it another try with different device setups and more tracks. Guess HT helps most, if already mention. I gonna open some old crackling projects and see what happens...

[tumar]

I gonna open some old crackling projects and see what happens...

I did so - they were still crackling. But... I put many Combinator into my projects. Uncombining and auto-routing (to send each instrument on his own mixer track) helps a bit.

[esselfortium]

They didn't say it wasn't something for Propellerhead to fix, but just that it's not an overall Reason performance problem: it's affecting UAD's plugins specifically, so Propellerhead has contacted UAD so they can learn more about what might be going wrong and figure out why Reason isn't handling them as well as it should be.

Well ok then, it's not an "overall Reason performance problem" (for Reason), but it IS affecting UAD plugins specifically. How that? No matter how it's worded Reason still buggers up UAD plugins even after the last patch, and (as I stated) these same plugins work fine in Logic X and Cubase 9. Im sure they will eventally fix it, but right now it's still broken and not on the UAD side.

Please read what I said instead of rolling your eyes at something you've decided you want to hear. It's not an "overall Reason performance problem," in that it relates specifically to something Propellerhead needs to fix about their handling of these plugins. It doesn't affect other plugins in general, so it's not an overall problem, but one relating specifically to Reason not playing nicely with some functionality UAD is making use of. That's literally all I said. I wasn't shifting the blame to UAD, just clarifying.

[sonicbyte]

The performance of my Macbook pro 2013 i5 has been improved at least 50%... I'm working right now with an epic orchestra track with lots of Kontakt instances with strings libraries, percussion, etc and now I can hear the whole song without a click or pop with much lower buffer size than before .... I can't believe it, thank you @propellerheads !!

[chimp_spanner]

Just a quick question - is anyone else getting sporadic loss of keyboard shortcut control? Most Ctrl/Option functions. Might just be something on my PC though. But a Reason Restart seems to cure it.

[supadee9000]

I've read this post with great interest, in particular the hyper hyperthreading issues. I thought I would chime in with my humble 2 cents: I have just started working on a remix - 60 audio stems! I also have a bunch of RE's going (4 synths, a couple of drum machines, couple of reverbs).

Before the update it was going pretty well: no DSP spikes. But after the update: instant DSP spikes and I couldn't play the session anymore.

After reading this thread, I turned hyperthreading off, and now it plays again. It is still slightly glitchy, but it plays. Thankfully: I was searching the net for how to revert to the previous update that was working.

Im on a 2.7ghz macbook pro, 12 gb ram, SSD.

[DJMaytag]

How do you turn Hyperthreading on and off? Is this a Reason setting or a BIOS setting?