Reason 9.5 & above CPU Stress Tests (2017, two different songfiles included)!

This forum is for discussing Propellerhead's music software. Questions, answers, ideas, and opinions... all apply.
Socialnotworking
Posts: 3
Joined: 24 Aug 2018

Post 24 Aug 2018

2.jpg
1.jpg
My first test, have only just finished configuring the rig so not sure if everything is optimal yet

Home built Z370 Rig cost me about £930 ($1200) all in
CPU model (Intel i7 8700k, OC 4.7 Ghz) I may test again without the OC as I have heard conflicting stories about the effects it has with apps like Reason

Also back in the R5 days they advised you to set everything to performance in windows and optimise apps for background services which I haven't done.
Out of interest I gave the windows background and optimisation a try and it actually made the performance worse on the benchmark
16GB DDR4 2400 RAM
No Audio Card ( I have a Saffire6 USB but it sucks so I have my Onkyo Amp and speakers hooked up via HDMI from the GPU Card (GTX 1060 6GB)
NVIDIA HD Audio 44,100 buffer 8,192 samples
95% CPU
RE ----> (20.930 Seconds)
Simple FX Chain ----> (1minute 25.782 seconds )
You do not have the required permissions to view the files attached to this post.

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 24 Aug 2018

Socialnotworking wrote:
24 Aug 2018
(Intel i7 8700k, OC 4.7 Ghz) I may test again without the OC as I have heard conflicting stories about the effects it has with apps like Reason
Not sure what conflicting stories you've heard, but overclocking a CPU - especially in Reason - almost always yields better results than a stock clock. This is assuming you're overclocking properly...

That said, try turning Hyperthreading off (or on). That, on the other hand, can change your performance drastically (YMMV).
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

Socialnotworking
Posts: 3
Joined: 24 Aug 2018

Post 25 Aug 2018

The two regularly sited complaints about OCs on DAWs are increased noise especially when using external interfaces to record audio into Reason and the challenge increased temperatures present to silent system design. DAW designers tend to want everything as cool and quiet as possible because the last thing you want is the sound of CPU fans ruining your delicate guitar or vocal track 😊

I did run at various OC settings and the difference between 4.0ghz and 4.8ghz wasn't huge, it was way less than you would think on the 8700k.

One bit of advice I was given was to install R10 on a secondary SSD along with the patch collections as allegedly this does help load times for hypersamples.

Turning off HT in R10 reduced the performance whereas the biggest increase came from using the DX primary sound driver with 8k samples.
In the heavy test this increased time by 5 seconds over the Asio I use for the Saffire.

👍

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 25 Aug 2018

Socialnotworking wrote:
25 Aug 2018
The two regularly sited complaints about OCs on DAWs are increased noise especially when using external interfaces to record audio into Reason
A cpu's clock has nothing to do with "noise" in a DAW, although some lower quality audio interfaces might have issues if their drivers (or hardware) are buggy. This is also often related to how your USB port handles system overclocks. Some behave better than others. Indeed, YMMV...
Socialnotworking wrote:
25 Aug 2018
and the challenge increased temperatures present to silent system design. DAW designers tend to want everything as cool and quiet as possible because the last thing you want is the sound of CPU fans ruining your delicate guitar or vocal track 😊
True, depending on your cooling solution, your silent rig might have issues. That said, this is easily addressed in any small home recording studio. A simple mic filter (such as SE Electronics Reflexion series) can have a massive effect on room noise. I've had great success with this on my overclocked system. I've also made a small acoustically damped panel that I can sit in front of my PC if the noise bleeds into any live recordings. A little effort goes a long way.
Socialnotworking wrote:
25 Aug 2018
I did run at various OC settings and the difference between 4.0ghz and 4.8ghz wasn't huge, it was way less than you would think on the 8700k.
Indeed, YMMV. Overclocking from my stock 3.5 Ghz to 4.5 Ghz has shown massive improvements on my old 3770k.
Socialnotworking wrote:
25 Aug 2018
One bit of advice I was given was to install R10 on a secondary SSD along with the patch collections as allegedly this does help load times for hypersamples.
You're part of the way there. Installing Reason and any samples/refills/whatever on an SSD will speed up load times all around, not just on hypersmaples. More like - any samples. ;) This is compared to older mechanical hard drives, of course. If you can install your OS on an SSD, you'll also have a massive speed increase over older mechanical drives. But if you're indeed running a true "silent system design", you already have SSD's. Right?
Socialnotworking wrote:
25 Aug 2018
Turning off HT in R10 reduced the performance whereas the biggest increase came from using the DX primary sound driver with 8k samples.
In the heavy test this increased time by 5 seconds over the Asio I use for the Saffire.

👍
Indeed, HT performance is hit or miss with Reason. In some instances, it improves performance noticeably on my system (usually with large projects that have complex instrument/effect chains and/or Combi tied to their own mix channels). But when I'm auditioning presets or just loading up one or two VST, if I turn HT off I get much better performance. Weird!

That said, you must use ASIO drivers in Reason if you want to record audio. DX sound is not acceptable for a serious pro recording setup. Latency is laughable.

:thumbs_up:
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

Socialnotworking
Posts: 3
Joined: 24 Aug 2018

Post 25 Aug 2018

If I am honest I don't really care that much about achieving peak performance. I am interested in how its done but i use my DAW for photo editing and occasional music production and for my needs it works fine as it is. I started using Reason in 2008 and have been through lots of different hardware configurations but I don't spend that much time on music nowadays. I appreciate the rig could run better but personally I would get no more benefit for what I need. Yes I run a system and two storage SSDs and im air cooled because I get the results I need at relatively low OC speeds and didn't want the expense or added build complications. I only really went for the K because I wasn't sure if I was going to do an upgrade in the future for gaming but I probably won't. In my original post you will see i quoted the build cost which was way below what SCAN and others charge for the same spec ready built and I am pleased with the result. I am aware of the latency issues on the DX but that has no impact on the test result as I am not playing anything. When I am playing I always use the Asio. I really just wanted to share my results withing the context of my specific setup to help others considering their next rig.

JunkMaster
Posts: 5
Joined: 25 Aug 2018

Post 25 Aug 2018

Socialnotworking wrote:
25 Aug 2018
The two regularly sited complaints about OCs on DAWs are increased noise especially when using external interfaces to record audio into Reason and the challenge increased temperatures present to silent system design. DAW designers tend to want everything as cool and quiet as possible because the last thing you want is the sound of CPU fans ruining your delicate guitar or vocal track 😊

I did run at various OC settings and the difference between 4.0ghz and 4.8ghz wasn't huge, it was way less than you would think on the 8700k.

One bit of advice I was given was to install R10 on a secondary SSD along with the patch collections as allegedly this does help load times for hypersamples.

Turning off HT in R10 reduced the performance whereas the biggest increase came from using the DX primary sound driver with 8k samples.
In the heavy test this increased time by 5 seconds over the Asio I use for the Saffire.

👍
@Socialnotworking

I use the native Focusrite Saffire ASIO driver here so to hear that in this instance the DX Primary Sound Driver worked better in your tests is quite interesting.

My computer is also overclocked but using an i7 Ivy Bridge-E cooled by a Corsair H100i.

Under normal Reason operating conditions, were you able to notice any difference?

User avatar
Troyvasanth
Posts: 60
Joined: 04 Jun 2016
Location: Mumbai

Post 16 Oct 2018

Hey Guys,
Test settings 44.1Khz buffer 1024 samples and cpu 95%
SPECS:
Cpu - Ryzen 2700x (stock speed and stock Prism cooler)
Motherboard - Asus prime 470x pro
Ram - Corsair vengence lpx C16 DDR4 3000Mhz (D.O.C.P profile enabled)
OS drive - Samsung 970 evo M.2 nvme 250gb ssd.
Audio Interface - Focusrite 2i4 (1st gen)

Simple Fx chain - 01:08:870
Complex FX chain - 00:16:370
No crackling of audio, It just simply stops.
Reason 9.5 simple fx.jpg
Reason 9.5 complex fx.jpg
You do not have the required permissions to view the files attached to this post.

Ahead74
Posts: 1
Joined: 29 Oct 2018

Post 29 Oct 2018

Tried the "Reason 9.5 2017 COMPLEX RE benchmark song V2"

WIN 10
I7-8700K CPU @ 3.70 GHz
32 GB RAM

No pops and crackles, but song stops at 0:00:48

//Anders

User avatar
Carly(Poohbear)
Posts: 2291
Joined: 25 Jan 2015
Location: UK

Post 09 Nov 2018

i97940@4Ghz (base speed 3.1GHz)
Test settings 44.1Khz buffer 1024 samples and cpu 95%

Complex HT OFF 1:21
Complex With HT all the way through.

Simple HT OFF 2:00
Simple with HT 1:14

PoohBear

----------------------------------------------------------------------------------------------------------
Tips & Tutorial's viewtopic.php?f=5&t=7503909
PoohBears Reason Channel https://www.youtube.com/channel/UCqCNsA ... p4cQF4j-8A
----------------------------------------------------------------------------------------------------------
My Soundcloud Page ....... Nektar Mappings
----------------------------------------------------------------------------------------------------------

User avatar
candybag
Posts: 46
Joined: 26 Feb 2018
Location: Sweden

Post 09 Nov 2018

i9 9900k @ 4,9 GHz
16 GB RAM
Focusrite Scarlet 2i2
Buffer setting @ 1024
44000 Hz

COMPLEX RE with HT on: 56,7 seconds
COMPLEX RE with HT off: 56,2 seconds
Yamaha HS7 - HD600 - Focusrite Scarlett 2i2 - Akai MPK261 - AT2050 - Auralex Project 2™ Roominator Kit
9900K - 16 GB - 3xXB270HU - GTX 1080 ti

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 09 Nov 2018

candybag wrote:
09 Nov 2018
i9 9900k @ 4,9 GHz
Are you overclocking your 9900K, or is that stock turbo clock? If you're overclocking, are you clocking all 8 cores at 4.9 Ghz? What cooler are you using? What are your core temps?

Just curious...

Thanks!
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

User avatar
candybag
Posts: 46
Joined: 26 Feb 2018
Location: Sweden

Post 09 Nov 2018

EnochLight wrote:
09 Nov 2018
candybag wrote:
09 Nov 2018
i9 9900k @ 4,9 GHz
Are you overclocking your 9900K, or is that stock turbo clock? If you're overclocking, are you clocking all 8 cores at 4.9 Ghz? What cooler are you using? What are your core temps?

Just curious...

Thanks!
Yes, that's across all cores. I used a preset overclock. Stock is 3,6 GHz i believe, with turbo on two cores.

Using Noctuas NH-D15, got about 2 mm of free space between the bottom fan and my GPU :S. During the test temps maxed out at 57 C. Bet it'll get even hotter under prolonged use!
Yamaha HS7 - HD600 - Focusrite Scarlett 2i2 - Akai MPK261 - AT2050 - Auralex Project 2™ Roominator Kit
9900K - 16 GB - 3xXB270HU - GTX 1080 ti

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 09 Nov 2018

candybag wrote:
09 Nov 2018
Using Noctuas NH-D15, got about 2 mm of free space between the bottom fan and my GPU :S. During the test temps maxed out at 57 C. Bet it'll get even hotter under prolonged use!
That NH-D15 is still the best air cooler in the industry (often beating many AIO water coolers). And 57 C is awesome - the 9900K's T-Junction is 100 C (meaning, it won't even throttle the die until around 95 C). Crank those babies to 5 Ghz man!!!!


https://www.intel.com/content/www/us/en ... 9900k.html
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

User avatar
candybag
Posts: 46
Joined: 26 Feb 2018
Location: Sweden

Post 09 Nov 2018

EnochLight wrote:
09 Nov 2018
candybag wrote:
09 Nov 2018
Using Noctuas NH-D15, got about 2 mm of free space between the bottom fan and my GPU :S. During the test temps maxed out at 57 C. Bet it'll get even hotter under prolonged use!
That NH-D15 is still the best air cooler in the industry (often beating many AIO water coolers). And 57 C is awesome - the 9900K's T-Junction is 100 C (meaning, it won't even throttle the die until around 95 C). Crank those babies to 5 Ghz man!!!!


https://www.intel.com/content/www/us/en ... 9900k.html
Haha yeah gonna try it after i've ironed out some wrinkles. Getting bluescreens with MEMORY_MANAGEMENT, hopefully it's just driver issues (never did a clean install after switching mobo+CPU.)

Edit: It's now 5,2 GHz across all cores. Strange thing is it only played 52 seconds this time (84 C package temp.) Before the extra OC i ran another test (4,9 GHz) where it played just past the 1 min mark, HT on in both cases.

Gonna try one more time later today, if the tradeoff is noisier fans and shorter playback it's a no-go :D
Yamaha HS7 - HD600 - Focusrite Scarlett 2i2 - Akai MPK261 - AT2050 - Auralex Project 2™ Roominator Kit
9900K - 16 GB - 3xXB270HU - GTX 1080 ti

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 10 Nov 2018

candybag wrote:
09 Nov 2018
Edit: It's now 5,2 GHz across all cores. Strange thing is it only played 52 seconds this time (84 C package temp.) Before the extra OC i ran another test (4,9 GHz) where it played just past the 1 min mark, HT on in both cases.

Gonna try one more time later today, if the tradeoff is noisier fans and shorter playback it's a no-go :D
Yeah, that’s weird. Let us know what happens after your second test at 5.2 Ghz.
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

LiberalTugboat
Posts: 10
Joined: 11 Oct 2017

Post 01 Mar 2019

Does anyone still have this file? The links on the first page no longer work.

User avatar
splangie
Posts: 200
Joined: 21 Dec 2017
Location: Park County, Colorado

Post 04 Mar 2019

LiberalTugboat wrote:
01 Mar 2019
Does anyone still have this file? The links on the first page no longer work.
If you need a copy of those files let me know.

I am running a i7-9700k and have all 8 cores running at 5.0. I can run both the files to completion. It starts to crackle a bit towards the end but no errrors are displayed. The highest core temp hits about 60 with a Noctua UH-14S. I don't have the option of turning off HT with a 9700k.

It appears that Reason 10 runs better on high end Windows based systems than the off the shelf MACs. Is that true?

Anyone runing a Mac Clone on 9700 or 9900 or bigger? Or a 2066 socket? Or will that even work. I ran one on a 3930k years ago.

Przemyslaw
Posts: 39
Joined: 05 Jan 2018

Post 05 Mar 2019

splangie wrote:
04 Mar 2019
LiberalTugboat wrote:
01 Mar 2019
Does anyone still have this file? The links on the first page no longer work.
If you need a copy of those files let me know.

I am running a i7-9700k and have all 8 cores running at 5.0. I can run both the files to completion. It starts to crackle a bit towards the end but no errrors are displayed. The highest core temp hits about 60 with a Noctua UH-14S. I don't have the option of turning off HT with a 9700k.

It appears that Reason 10 runs better on high end Windows based systems than the off the shelf MACs. Is that true?

Anyone runing a Mac Clone on 9700 or 9900 or bigger? Or a 2066 socket? Or will that even work. I ran one on a 3930k years ago.
Great result in terms of price/performance. What score in passmark benchmark on that oc system do You have ? And what the rest of computer specification ? And what latency in Reason during test ?

User avatar
splangie
Posts: 200
Joined: 21 Dec 2017
Location: Park County, Colorado

Post 05 Mar 2019

Przemyslaw wrote:
05 Mar 2019
splangie wrote:
04 Mar 2019


If you need a copy of those files let me know.

I am running a i7-9700k and have all 8 cores running at 5.0. I can run both the files to completion. It starts to crackle a bit towards the end but no errrors are displayed. The highest core temp hits about 60 with a Noctua UH-14S. I don't have the option of turning off HT with a 9700k.

It appears that Reason 10 runs better on high end Windows based systems than the off the shelf MACs. Is that true?

Anyone runing a Mac Clone on 9700 or 9900 or bigger? Or a 2066 socket? Or will that even work. I ran one on a 3930k years ago.
Great result in terms of price/performance. What score in passmark benchmark on that oc system do You have ? And what the rest of computer specification ? And what latency in Reason during test ?
It is on an Asus Prime Z390-A MB, with 2-8GB G. Skill 3000s, Samsung EVO 860 and just using UHD 630 from CPU for graphics. CPU scores 19260 and 3038 for single thread, memory gets 3496. That puts them all in 99th percentile. Vcore is set to 1.28, Vcio 1.2, just about everything manually and no offsets on AVX or anything else. RAM is stock timing. I stressed it with Prime95 Blend for 24hrs and Small FFTs for 14 with AVX off. The small FFTs for 14 hours was an accident, but it passed and temps hit 88 so that was nice to see. I ran the test out both to the Realtek digital from MB and USB to headphones. The input latency was showing 1 ms and 0 out. I did not have any control surfaces selected. I use a Tascam US2x2 when needed. Should the test be done using that as an out with a control surface as well?

In November the CPU was $369 after 30 dollar bundle discount with MB at 159 from Microcenter. Quite a bit but all performance issues for me are gone.

Latest test results -
https://www.passmark.com/baselines/V9/d ... 7882853643

Przemyslaw
Posts: 39
Joined: 05 Jan 2018

Post 05 Mar 2019

splangie wrote:
05 Mar 2019
Przemyslaw wrote:
05 Mar 2019


Great result in terms of price/performance. What score in passmark benchmark on that oc system do You have ? And what the rest of computer specification ? And what latency in Reason during test ?
It is on an Asus Prime Z390-A MB, with 2-8GB G. Skill 3000s, Samsung EVO 860 and just using UHD 630 from CPU for graphics. CPU scores 19260 and 3038 for single thread, memory gets 3496. That puts them all in 99th percentile. Vcore is set to 1.28, Vcio 1.2, just about everything manually and no offsets on AVX or anything else. RAM is stock timing. I stressed it with Prime95 Blend for 24hrs and Small FFTs for 14 with AVX off. The small FFTs for 14 hours was an accident, but it passed and temps hit 88 so that was nice to see. I ran the test out both to the Realtek digital from MB and USB to headphones. The input latency was showing 1 ms and 0 out. I did not have any control surfaces selected. I use a Tascam US2x2 when needed. Should the test be done using that as an out with a control surface as well?

In November the CPU was $369 after 30 dollar bundle discount with MB at 159 from Microcenter. Quite a bit but all performance issues for me are gone.

Latest test results -
https://www.passmark.com/baselines/V9/d ... 7882853643

Many thanks for the description. It's very interesting that you went through the whole reason test complex. I have i7 7820x, after OC 19990 points in the passmark and I reach 56 seconds in this complex test. The remaining specification is comparable to yours. With the fact that it's great if you carried out the test on Tascam US2x2 drivers (preferably with 1024samples latency in Reason audio preferences). The control surface is unlikely to affect the test.
You have a very strong single core processor, I wonder if this determines such a huge performance. Your 3038 to 2489 in my single thread processor. That's 20% more.

User avatar
jam-s
Posts: 668
Joined: 17 Apr 2015
Location: Aachen, Germany

Post 05 Mar 2019

splangie wrote:
04 Mar 2019
I am running a i7-9700k and have all 8 cores running at 5.0. I can run both the files to completion. It starts to crackle a bit towards the end but no errrors are displayed. The highest core temp hits about 60 with a Noctua UH-14S. I don't have the option of turning off HT with a 9700k.
Are you sure you did the test correctly? Crackling sound hints that you have no CPU limit set.

From the 1st post: "CPU Usage Limit must be set to 95% in Reason's settings"
If you're in Aachen, come and visit us at the Voidspace.

User avatar
EnochLight
Posts: 5468
Joined: 17 Jan 2015
Location: your mom

Post 05 Mar 2019

splangie wrote:
04 Mar 2019

I don't have the option of turning off HT with a 9700k.
Of course you do - under Reason’s Preferences. And there should be an option under your motherboard BIOS if you want to turn it off for Windows (which I don’t recommend).
Windows 10 64-bit | Reason 10.4 |  Studio One 4.5 | Asus Sabertooth Z77 | Intel i7 3770k Quad-Core @ 3.5 Ghz | 16 GB RAM | Mushkin Reactor 1TB SSD | RME Babyface Pro| Nektar Panorama P-4 | Akai MPC Live

MatsNo
Posts: 6
Joined: 29 Jan 2018

Post 05 Mar 2019

He doesn't have HT

MatsNo
Posts: 6
Joined: 29 Jan 2018

Post 05 Mar 2019

Also the fact that there's a maximizer on the master bus renders this test pretty useless, making it a bottleneck that favours high single core speed..

User avatar
splangie
Posts: 200
Joined: 21 Dec 2017
Location: Park County, Colorado

Post 05 Mar 2019

jam-s wrote:
05 Mar 2019
splangie wrote:
04 Mar 2019
I am running a i7-9700k and have all 8 cores running at 5.0. I can run both the files to completion. It starts to crackle a bit towards the end but no errrors are displayed. The highest core temp hits about 60 with a Noctua UH-14S. I don't have the option of turning off HT with a 9700k.
Are you sure you did the test correctly? Crackling sound hints that you have no CPU limit set.

From the 1st post: "CPU Usage Limit must be set to 95% in Reason's settings"
Yep, I was running with limit at none. That makes a difference and now neither song completes. At 95%, R9.5 Benchmark44...V2 stops at 51.5 with no pops or noise that I can hear. R9.5+2017 SimpleFX... stops at 2:27. I can hear some noise during the last few seconds or so.
Last edited by splangie on 05 Mar 2019, edited 1 time in total.

  • Information
  • Who is online

    Users browsing this forum: No registered users and 2 guests