ReasonTalk.com

I got a new CPU - Ryzen 5950X and I loaded the old 9.5 benchmark and it played the whole song (with 256 sample buffer setting) and CPU only got to 52% load, no pops or clicks. So I thought.. why not make a new one so when times come to upgrade, we can compare which CPUs are better at Reason?

I recommend to try this "benchmark" ONLY if your computer has (at the very least) an 8 core CPU and 16 GB RAM. It could be that even 8 core CPU can't play it at all. Just having this project loaded without any playing loads my 5950X 16 core CPU to ~48%!

How to get any meaningful comparison:
Reboot your computer to have a clean start and close any background program that could interfere or create network traffic (windows updates, steam clients, browser, games etc.)
Start Reason and have the same Reason settings as shown below, load the bechmark play a few seconds then stop and start playing from beginning and pay attention to when you start to hear click and pops and note the time. On my system, this "benchmark" plays perfect to ~ 1 m 18 seconds then I can hear pops and clicks and ~1m 30 seconds there are more click and pops than music

2nd of January 2021 EDIT:
I've done second R10+ benchmark, this one is even more demanding, it's called Brutal#2 but behaves more like a Nuke on my computer, I have to wait anywhere from 10 to 50 seconds just to hit play or double stop so the cursor goes to start, extremely laggy transport bar.
On my computer I get the first crackle\pop sound @ 3 min 1 second, and then sporadically increasing, with consistent crackles and pops after 3 min 50 seconds.

These are the new settings that I used:


Here is the "Brutal#1" file: http://s.go.ro/o8g9b30b

Here is the "Brutal#2" file http://s.go.ro/t1tlfvil
sounds like this:



Please report the following when sharing your results:

Your CPU model (Intel i9 10900k, AMD Ryzen 3900X, 5950X, Threadripper 3990X, etc) and clock speed - as well as whether or not you are overclocking
The amount of RAM and frequency
Your audiocard Brand and Model
Your audiocard driver settings (sample rate 44100 Hz recommended and max buffer settings)
Reason version
Operating system and version
The time when you begin to hear pops and clicks




If your computer can't play any of these benchmarks at all (not even the first 8 seconds) without clicks and pops, but still want to see how your computer compares with others regarding Reason performance then it might be best to try the other benchmark files:

a. Reason 9.5 & above CPU Stress Tests (2017, two different songfiles included)! viewtopic.php?f=4&t=7501402
b. Reason 8.1 & above CPU Stress Test (songfile included)! viewtopic.php?f=4&t=7263897

This is fun!

NO OC
AMD Ryzen 9 5900X 12-Core Processor 3.70 GHz
DDR4-3200 Kingston 32GB HyperX Fury RAM running at 2400mhz
AMD Asus TUF B550M-Plus MB
SSD Samsung 970 EVO 1000GB NVMe M.2 SSD
SSL 2 soundcard

So first tried running the song at exactly your setting. = instant cracks. unplayable.

Then i switched On "Render Audio using audio card buffer size setting (improves plugin performance)
And voila, no cracks or pops until 2:42m , which is pretty good i think, however i noticed that the reason app was pretty "slow" when looking around the mixer, sequencer and watching the seconds timer on the transport bar, how was your experience?

Wonder what the "Render Audio using audio card buffer size setting (improves plugin performance)" actually does or why it failed with it off, may be my SSL 2 that gets confused lol

Thanks for another great benchmark test Kategra !

I have the same issue like Le Boeuf. Render Audio using audio card buffer size setting (improves plugin performance) - when switched off DSP is full instantly whithout playing the song. When swithed on I can run song smooth till 48-52 seconds. This function is always ON on my preferences setup.
Graphicaly reason is slower when looking around the mixer, sequencer and watching the seconds timer on the transport bar.

Your CPU model i7 7820x 4,3GHz OC (8-core cpu)
32GB 3800MHz CL17
Motu Track 16
sample rate 44100 Hz 1024samples
Reason version 10
Windows 10
The time when you begin to hear pops and clicks - 48-52 seconds

I'm to scared to open the project file, but would love to find out more about the making of it.

Is it an actual song or have you just added stuff until the machine starts going tits up?

Render audio using audio card buffer ON ... Cracks around 45 sec.

It's a Laptop
i7 8750 6xcores
16gb Ram 2667mhz
Win10
Reason11
RME Fireface UC

M

Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz 3.40 GHz with 16 GB Ram.

Fire Brigade is on route to put it out.

Kind of glad to see that you experienced the same issues.

It is also worth noting that i was running Reason on a 4K 60hz resolution. Hence the extra lag.

Przemyslaw wrote: ↑

09 Dec 2020

Thanks for another great benchmark test Kategra !

I have the same issue like Le Boeuf. Render Audio using audio card buffer size setting (improves plugin performance) - when switched off DSP is full instantly whithout playing the song. When swithed on I can run song smooth till 48-52 seconds. This function is always ON on my preferences setup.
Graphicaly reason is slower when looking around the mixer, sequencer and watching the seconds timer on the transport bar.

Your CPU model i7 7820x 4,3GHz OC (8-core cpu)
32GB 3800MHz CL17
Motu Track 16
sample rate 44100 Hz 1024samples
Reason version 10
Windows 10
The time when you begin to hear pops and clicks - 48-52 seconds

same issue like Le Boeuf over here, switched on render audio

CPU model AMD 3700x @4,3GHz OC'd (8-core cpu)
16GB 3200MHz CL17
Roland Quad Capture
sample rate 44100 Hz 1024samples
Reason version 11
Windows 10
first little clicks @1:24 seconds @ 1:37 a lot of of clicks and glitches
@2048 samples first glitches @ 1:49

2:32

I had to enable "Use hyper-threading audio rendering" and "Render audio using audio ..." as well.

Without these two options enabled i can't start song playback at all.

My configuration:

CPU: Intel Core i9-10900K, 10x 3.70GHz, BOX, Comet Lake-S, LGA1200 (BX8070110900K) 3700 MHz
Motherboard: GIGABYTE Z490 VISION D
RAM: G.Skill Trident Z F4-3200C14Q-64GTZSW (4x16GB) DDR4-3200 C14
PSU: be quiet! Pure Power 11 700W
CPU Fan/Cooler: NOCTUA NH-D15S
SSD: 2 x SAMSUNG 970 EVO Plus 1TB M.2 SSD PCIe 3.0x4 + SAMSUNG 860 EVO SSD 1TB 2,5 sata3

RME Babyface, 44.1 kHz, 1024 samples

Przemyslaw wrote: ↑

09 Dec 2020

Thanks for another great benchmark test Kategra !

I have the same issue like Le Boeuf. Render Audio using audio card buffer size setting (improves plugin performance) - when switched off DSP is full instantly whithout playing the song. When swithed on I can run song smooth till 48-52 seconds. This function is always ON on my preferences setup.
Graphicaly reason is slower when looking around the mixer, sequencer and watching the seconds timer on the transport bar.

Your CPU model i7 7820x 4,3GHz OC (8-core cpu)
32GB 3800MHz CL17
Motu Track 16
sample rate 44100 Hz 1024samples
Reason version 10
Windows 10
The time when you begin to hear pops and clicks - 48-52 seconds

Thanks!



CPU Ryzen 5950X (16 cores, SMT off - similar to Intel Hyper Threading deactivated)
Overclocking => PBO boosts speed ~4.4Ghz when Reason loads the CPU

Motherboard Gigabyte Aorus PRO X570
Memory 4x8 GB Corsair 3000 MHz CL15
SSD Samsung EVO 850 (Windows10, ver. 2004)
SSD ADATA SU800 (Reason 10)
Video Asus RTX2080
Monitor LG G-Sync 144Hz

Audio Interface RME Fireface 400
sample rate 44100 Hz ,
buffer 1024 samples

With Render Audio using audio card buffer size setting = OFF => plays ~ 1min 18 seconds perfect, then clicks and pops
With Render Audio using audio card buffer size setting = ON => plays the whole song no clicks and I don't notice additional GUI lag - which Is already pretty high.



Here is the screenshot with the CPU load with Render Audio = ON

-

Billy wrote: ↑

09 Dec 2020

I'm to scared to open the project file, but would love to find out more about the making of it.

Is it an actual song or have you just added stuff until the machine starts going tits up?

Indeed It's just a loop (let's call it X1, made of many devices), then I added the same amount of devices as X1 every two bars... X2, X3 ,X4 .....X50.
So, in the last loop, there are 50 times more devices triggered by the sequencer than in the first loop.
Please note that almost all the devices consume lots of CPU cycles even if no sound is going through them.. so we can't make a perfect benchmark song that puts only 5% load on the CPU in the first bars and 100% in the last bars.. it's more like 25% in the first bars and 100% closer to the end or 50% in the first bars and 100% closing to middle....something like that..

Kategra wrote: ↑

09 Dec 2020

Billy wrote: ↑

09 Dec 2020

I'm to scared to open the project file, but would love to find out more about the making of it.

Is it an actual song or have you just added stuff until the machine starts going tits up?

Indeed It's just a loop (let's call it X1, made of many devices), then I added the same amount of devices as X1 every two bars... X2, X3 ,X4 .....X50.
So, in the last loop, there are 50 times more devices triggered by the sequencer than in the first loop.
Please note that almost all the devices consume lots of CPU cycles even if no sound is going through them.. so we can't make a perfect benchmark song that puts only 5% load on the CPU in the first bars and 100% in the last bars.. it's more like 25% in the first bars and 100% closer to the end or 50% in the first bars and 100% closing to middle....something like that..

Sweet

Hi kategra, just curious.

You don't experience any xtra lag when rezising windows, transport bar and such in this song file?
I can play thru the whole thing with HT and Audio buffer on, but the responsiveness is really slooooooooow haha

Would be curious to see how my system would perform if i just OC'ed a smidge, im just not a OC expert lol.
but maybe i should try to run the "Auto Tuning" in the Asus AI utility?

Kategra wrote: ↑

09 Dec 2020

Przemyslaw wrote: ↑

09 Dec 2020

Thanks for another great benchmark test Kategra !

I have the same issue like Le Boeuf. Render Audio using audio card buffer size setting (improves plugin performance) - when switched off DSP is full instantly whithout playing the song. When swithed on I can run song smooth till 48-52 seconds. This function is always ON on my preferences setup.
Graphicaly reason is slower when looking around the mixer, sequencer and watching the seconds timer on the transport bar.

Your CPU model i7 7820x 4,3GHz OC (8-core cpu)
32GB 3800MHz CL17
Motu Track 16
sample rate 44100 Hz 1024samples
Reason version 10
Windows 10
The time when you begin to hear pops and clicks - 48-52 seconds

Thanks!



CPU Ryzen 5950X (16 cores, SMT off - similar to Intel Hyper Threading deactivated)
Overclocking => PBO boosts speed ~4.4Ghz when Reason loads the CPU

Motherboard Gigabyte Aorus PRO X570
Memory 4x8 GB Corsair 3000 MHz CL15
SSD Samsung EVO 850 (Windows10, ver. 2004)
SSD ADATA SU800 (Reason 10)
Video Asus RTX2080
Monitor LG G-Sync 144Hz

Audio Interface RME Fireface 400
sample rate 44100 Hz ,
buffer 1024 samples

With Render Audio using audio card buffer size setting = OFF => plays ~ 1min 18 seconds perfect, then clicks and pops
With Render Audio using audio card buffer size setting = ON => plays the whole song no clicks and I don't notice additional GUI lag - which Is already pretty high.



Here is the screenshot with the CPU load with Render Audio = ON

-
Bechmark 2020 R10 Render audio.jpg

Le Boeuf wrote: ↑

10 Dec 2020

Hi kategra, just curious.

You don't experience any xtra lag when rezising windows, transport bar and such in this song file?
I can play thru the whole thing with HT and Audio buffer on, but the responsiveness is really slooooooooow haha

Would be curious to see how my system would perform if i just OC'ed a smidge, im just not a OC expert lol.
but maybe i should try to run the "Auto Tuning" in the Asus AI utility?

I don't notice lag difference Render Audio using audio card buffer Render Audio using audio card buffer OFF or ON setting.
I don't notice lag resizing windows, transport bar or the sequencer.
But I do notice extra lag when I scroll in the Rack window or the Mixer window for this project song, irrespective of audio buffer size or the Render Audio setting.

Regarding overclocking, the cooling is extremely very important. I use NOCTUA NH-15D huge air cooler. If you have a high end CPU cooler it may be worth it to overclock.
When I overclock a PC, I always do it from BIOS settings.
For AMD, research what these settings do: PBO, PTT, EDC, EDC and if your motherboard supports changing them to manual values or if they have a setting like "motherboard" try that and basically if lets the motherboard feed more current to CPU than what AMD said is standard for your CPU.

Kategra wrote: ↑

08 Dec 2020

Le Boeuf wrote: ↑

09 Dec 2020

Wonder what the "Render Audio using audio card buffer size setting (improves plugin performance)" actually does or why it failed with it off, may be my SSL 2 that gets confused lol

There is no point in setting a buffer size of 1024 and then turning off "Render Audio using audio card buffer size".

This makes your sound card work with a buffer size of 1024 but Reason with a buffer size 64, this is very conflicting as Reason has to do 16 collections per the sound card collect . The only real time to turn that option off is for timing issues, e.g. I have turned it off when making video's as I have to use DX drivers to reduce CV latency.

Please update your original post so that people leave that option on for a fairer test results across systems.

Thank you for taking your time to make a new test....

CPU i7-4790K

Motherboard Gigabyte Gamming 3
Memory 2x8 GB Corsair 2400
SSD Kingston 960 gb
Windows10
Video on board intel HD 4600
Reason 11 suite
Audio Interface Focusrite 18i20
sample rate 44100 Hz ,
buffer 1024 samples

My old PC play 20 seconds, but I don't need to change it yet, it has worked very well in my projects ...

Carly(Poohbear) wrote: ↑

30 Dec 2020

There is no point in setting a buffer size of 1024 and then turning off "Render Audio using audio card buffer size".

This makes your sound card work with a buffer size of 1024 but Reason with a buffer size 64, this is very conflicting as Reason has to do 16 collections per the sound card collect . The only real time to turn that option off is for timing issues, e.g. I have turned it off when making video's as I have to use DX drivers to reduce CV latency.

Please update your original post so that people leave that option on for a fairer test results across systems.

Thank you for taking your time to make a new test....

It's not conflicting, my understanding is, turning off Render Audio using audio card buffer size setting (which default is OFF) is needed to have accurate and 100% repeatable events, like CV, in Reason.

But anyone can test how they like and post both results; when it will be time to upgrade, I will be interested in the results with the setting OFF because my current CPU can't play the whole project with the setting OFF

I believe that trying all benchmark songs on this forum, anyone can gauge the relative Reason performance of their computer and then have a pretty good idea on what to get to increase performance without having bad surprises.

arqui wrote: ↑

31 Dec 2020

CPU i7-4790K

Motherboard Gigabyte Gamming 3
Memory 2x8 GB Corsair 2400
SSD Kingston 960 gb
Windows10
Video on board intel HD 4600
Reason 11 suite
Audio Interface Focusrite 18i20
sample rate 44100 Hz ,
buffer 1024 samples

My old PC play 20 seconds, but I don't need to change it yet, it has worked very well in my projects ...

I presume that you activated this setting Render Audio using audio card buffer....

AMD Threadripper 3970x @ 4.10 GHz
64gb G Skillz Ram at 3600 Mhz (2x32gb)
Tascam 16x08
Sample rate 4410hz
Buffer at 2048
Reason 11 Suite (latest update)
Windows 10 Pro
No pops or clicks but a very sluggish graphical UI whether song is playing or not

Kategra wrote: ↑

31 Dec 2020

arqui wrote: ↑

31 Dec 2020

CPU i7-4790K

Motherboard Gigabyte Gamming 3
Memory 2x8 GB Corsair 2400
SSD Kingston 960 gb
Windows10
Video on board intel HD 4600
Reason 11 suite
Audio Interface Focusrite 18i20
sample rate 44100 Hz ,
buffer 1024 samples

My old PC play 20 seconds, but I don't need to change it yet, it has worked very well in my projects ...

I presume that you activated this setting Render Audio using audio card buffer....

Yes, you are correct, all three options activated

Kategra wrote: ↑

31 Dec 2020

Carly(Poohbear) wrote: ↑

30 Dec 2020

There is no point in setting a buffer size of 1024 and then turning off "Render Audio using audio card buffer size".

This makes your sound card work with a buffer size of 1024 but Reason with a buffer size 64, this is very conflicting as Reason has to do 16 collections per the sound card collect . The only real time to turn that option off is for timing issues, e.g. I have turned it off when making video's as I have to use DX drivers to reduce CV latency.

It's not conflicting, my understanding is, turning off Render Audio using audio card buffer size setting (which default is OFF) is needed to have accurate and 100% repeatable events, like CV, in Reason.

If you think setting your Audio buffer size to 1024 and then using Reason with a buffer size of 64 is not conflicting, then that would be a new meaning to the word conflicting (1024 does not equal 64). Buffer size does not have anything to with accurate and 100% repeatable events including CV in Reason, of course doing testing with different buffer sizes will change things but then you are changing the parameters of a test.
"Render Audio using Audio Card Buffer Size" is ON as default.

Carly(Poohbear) wrote: ↑

31 Dec 2020

If you think setting your Audio buffer size to 1024 and then using Reason with a buffer size of 64 is not conflicting, then that would be a new meaning to the word conflicting (1024 does not equal 64). Buffer size does not have anything to with accurate and 100% repeatable events including CV in Reason, of course doing testing with different buffer sizes will change things but then you are changing the parameters of a test.
"Render Audio using Audio Card Buffer Size" is ON as default.

Carly, you're right on everything, with small exception that "Render Audio using Audio Card Buffer Size" with a different value (>64) sample rate for the Audio Buffer can change the sound when making feedback loops and automating or CV controlling parameters in a VST.

Here is a quote from Reason Studios

While this is all good news, some things might actually sound different with these rendering changes and that's important to understand. When automating or CV controlling parameters in a VST, it will use the same resolution as the buffer size. This is also true for any feedback connection, for example sending a signal out of a device and then back in to the same device.

In Reason 10.2 and earlier, the buffer size was fixed at 64 samples and thus had almost no noticeable delay. Now, with larger buffer sizes you might hear a difference when controlling a VST parameter or hear a delay when creating a feedback loop. In most cases, you probably won't notice it, but for full compatibility with your older songs you can either bring the buffer size down to 64 samples or untick the new option in the preferences.
https://www.reasonstudios.com/blog/reason-103-is-here

Kategra wrote: ↑

01 Jan 2021

Carly(Poohbear) wrote: ↑

31 Dec 2020

If you think setting your Audio buffer size to 1024 and then using Reason with a buffer size of 64 is not conflicting, then that would be a new meaning to the word conflicting (1024 does not equal 64). Buffer size does not have anything to with accurate and 100% repeatable events including CV in Reason, of course doing testing with different buffer sizes will change things but then you are changing the parameters of a test.
"Render Audio using Audio Card Buffer Size" is ON as default.

Carly, you're right on everything, with small exception that "Render Audio using Audio Card Buffer Size" with a different value (>64) sample rate for the Audio Buffer can change the sound when making feedback loops and automating or CV controlling parameters in a VST.

Here is a quote from Reason Studios

While this is all good news, some things might actually sound different with these rendering changes and that's important to understand. When automating or CV controlling parameters in a VST, it will use the same resolution as the buffer size. This is also true for any feedback connection, for example sending a signal out of a device and then back in to the same device.

In Reason 10.2 and earlier, the buffer size was fixed at 64 samples and thus had almost no noticeable delay. Now, with larger buffer sizes you might hear a difference when controlling a VST parameter or hear a delay when creating a feedback loop. In most cases, you probably won't notice it, but for full compatibility with your older songs you can either bring the buffer size down to 64 samples or untick the new option in the preferences.
https://www.reasonstudios.com/blog/reason-103-is-here

Hence my statement " of course doing testing with different buffer sizes will change things but then you are changing the parameters of a test." <- This will also include any Reason file created with a different buffer size could be impacted.

At the end of the day we both want the same thing, accurate results and the best way in this case is to get everyone to run with the same audio settings..
You would be better off asking for everyone to run with buffer size of 64 (does not matter if "Render Audio using Audio Card Buffer Size" is on or off).
It would also be useful to ask to run the test twice, with and without "Use HT Audio Rendering"

CPU model: AMD Ryzen 3900X 3.8GHz 12 Core - not O/C
Amount of RAM and frequency: 32GB 3600Hz DDR4
Audiocard: Focusrite Saffire PRO 14
Your audiocard driver settings: 44.1, buffer 2048
Reason version: Reason 11 Suite (just upgraded for relatively peanuts!)
Operating system and version: Windows 10 Home Version 20H2
The time when you begin to hear pops and clicks: It got to the end but just started clicking at the last few seconds.

I did turn on the Performance options in Preferences - why make it hard on the PC?

Grum.

Carly(Poohbear) wrote: ↑

01 Jan 2021

Hence my statement " of course doing testing with different buffer sizes will change things but then you are changing the parameters of a test." <- This will also include any Reason file created with a different buffer size could be impacted.

At the end of the day we both want the same thing, accurate results and the best way in this case is to get everyone to run with the same audio settings..

You would be better off asking for everyone to run with buffer size of 64 (does not matter if "Render Audio using Audio Card Buffer Size" is on or off).

It would also be useful to ask to run the test twice, with and without "Use HT Audio Rendering"

^^ THIS ^^

I've dropped out of these benchmarking threads these days because my machine is admittedly an ancient artifact, but making sure everyone tests at the exact same settings and parameters is the only way to get solid data. Keep the science simple, peeps!