Reason10.3 timing

This forum is for discussing Propellerhead's music software. Questions, answers, ideas, and opinions... all apply.
User avatar
wendylou
Posts: 200
Joined: 15 Jan 2015
Location: Saint Joseph, MO

Post 14 May 2019

I had a song.I did on my last album that had a dual arpeggiator combinator consisting of some delays and two RPGs. When I opened that today in the latest Reason 10.3d78 build 9745, the timing was off and quite clearly so. I had multicore processing on and multithreading off, but turning off multicore processing made no difference. I looked at my buffer settings, which was previously set to max 2048 for my prior playback of a song with a bazillion devices. So I now lowered it to the minimum of 64. Voila, the arp plays back with the correct timing. The song does not use many devices, and it barely loads the DSP, so it's not struggling or anything. The delays and everything in the combinator are stock devices. But with a dual arpeggiator running, timing is crucial for the interplay of beats. Has anyone else experienced timing issues in the latest Reason 10.3? It's clearly a buffer issue but did not occur in previous versions. I missed spotting this during beta testing.
:PUF_balance: http://www.galxygirl.com -- :reason: user since 2002

zoidkirb
Posts: 120
Joined: 18 Nov 2018

Post 14 May 2019

Check the blog posts on propellerheads for the 10.3 release, this stuff is explained. Basically the cv was always fixed at 64 buffer size but now the larger buffer size allows you to improve performance but at a cost of potential longer cv timings

User avatar
Loque
Posts: 5651
Joined: 28 Dec 2015

Post 14 May 2019

Did you read the release notes from the blog?
https://www.propellerheads.com/blog/reason-103-is-here

The buffer size has an impact on automation and CV signals. May this be your problem?
:reason: 10, Win10 64Bit.

User avatar
wendylou
Posts: 200
Joined: 15 Jan 2015
Location: Saint Joseph, MO

Post 15 May 2019

zoidkirb wrote:
14 May 2019
Check the blog posts on propellerheads for the 10.3 release, this stuff is explained. Basically the cv was always fixed at 64 buffer size but now the larger buffer size allows you to improve performance but at a cost of potential longer cv timings
Loque wrote:
14 May 2019
Did you read the release notes from the blog?
https://www.propellerheads.com/blog/reason-103-is-here
The buffer size has an impact on automation and CV signals. May this be your problem?
Well color me daft, I did not see that before and you are correct, CV is indeed the issue – the two RPGs control two subtractors via 8 CV lines total. Duh! :o Thanks!!!
:PUF_balance: http://www.galxygirl.com -- :reason: user since 2002

PhillipOrdonez
Posts: 471
Joined: 20 Oct 2017
Location: Colombia

Post 15 May 2019

wendylou wrote:
15 May 2019
zoidkirb wrote:
14 May 2019
Check the blog posts on propellerheads for the 10.3 release, this stuff is explained. Basically the cv was always fixed at 64 buffer size but now the larger buffer size allows you to improve performance but at a cost of potential longer cv timings
Loque wrote:
14 May 2019
Did you read the release notes from the blog?
https://www.propellerheads.com/blog/reason-103-is-here
The buffer size has an impact on automation and CV signals. May this be your problem?
Well color me daft, I did not see that before and you are correct, CV is indeed the issue – the two RPGs control two subtractors via 8 CV lines total. Duh! :o Thanks!!!
I thought it shouldn't be an issue unless you were on a feedback loop though?

User avatar
friday
Posts: 184
Joined: 17 Jan 2015

Post 15 May 2019

PhillipOrdonez wrote:
15 May 2019
I thought it shouldn't be an issue unless you were on a feedback loop though?
I assumed the same assumption?

User avatar
dioxide
Posts: 1105
Joined: 15 Jul 2015

Post 15 May 2019

Yeah this tripped me up the other day also. Something to do with CV and Thor's sequencer.

When it comes to feedback loops, this covers a lot. It includes wiring devices into themselves and also putting Send effects into Mix Channels.

User avatar
nooomy
Posts: 372
Joined: 16 Jan 2015

Post 15 May 2019

dioxide wrote:
15 May 2019
Yeah this tripped me up the other day also. Something to do with CV and Thor's sequencer.

When it comes to feedback loops, this covers a lot. It includes wiring devices into themselves and also putting Send effects into Mix Channels.
Wait are all CV signals effected by the buffer size or is it just CV signals sent to VST's?

64 is super low and my computer is way to slow to produce a whole song in 64.

User avatar
ScuzzyEye
Posts: 1305
Joined: 15 Jan 2015

Post 15 May 2019

nooomy wrote:
15 May 2019
Wait are all CV signals effected by the buffer size or is it just CV signals sent to VST's?

64 is super low and my computer is way to slow to produce a whole song in 64.
I've had conflicting answers to this. But I believe there's one CV value change per audio batch, with the render audio using audio buffer size box checked.

You can still use a larger buffer, but revert to the 64-frame batch size of pre-10.3, by unchecking the box in the preferences. You'll lose the performance increase that 10.3 allowed, but it won't be as bad as running with a 64-sample buffer.

User avatar
nooomy
Posts: 372
Joined: 16 Jan 2015

Post 15 May 2019

ScuzzyEye wrote:
15 May 2019
nooomy wrote:
15 May 2019
Wait are all CV signals effected by the buffer size or is it just CV signals sent to VST's?

64 is super low and my computer is way to slow to produce a whole song in 64.
I've had conflicting answers to this. But I believe there's one CV value change per audio batch, with the render audio using audio buffer size box checked.

You can still use a larger buffer, but revert to the 64-frame batch size of pre-10.3, by unchecking the box in the preferences. You'll lose the performance increase that 10.3 allowed, but it won't be as bad as running with a 64-sample buffer.
Im still on reason 9, but this sounds pretty bad.
So basically all the CV will be out of sync if i use something else than 64-sample buffer? Unless i deactivate the "VST fix patch"...

Its like you have to chose if you want CV in sync or a 10-80 % performance boost on your rack extensions and VST's...

WTF??? :| :thumbs_down: :thumbs_down:

zoidkirb
Posts: 120
Joined: 18 Nov 2018

Post 15 May 2019

I haven't had any perceptible changes to my old projects that involved a fair amount of cv cabling. I think if you're creatively using cv to trigger things like arpegiators that trigger something else then that could lead to a real problem as things absolutely need to start 'on time' to work. Seems like moving forward we will intuitively work around any new limitations, but there will always be a chance that old projects may behave strangely when loaded in 10.3. In that case we either have to re think our cv spaghetti, or simply uncheck the new buffer settings in preferences

User avatar
dioxide
Posts: 1105
Joined: 15 Jul 2015

Post 15 May 2019

The problem is CV that involves a feedback loop. I need to test more to be honest but an example would be wiring a Kong pad into another Kong pad on the same device. This means the CV trigger is pushed back into the next batch (apparently) and it causes a delay, which can screw the timing up entirely if you're using a large buffer. Not a great state of affairs IMO, but hey people wanted improved VST performance...

User avatar
wendylou
Posts: 200
Joined: 15 Jan 2015
Location: Saint Joseph, MO

Post 15 May 2019

Well here's the CV in my dual arp Combinator, all stock devices. It's just sequence control, mod wheel, and pitch wheel, plus one Subtractor LFO that adjusts the Freq on an ECF-42. So no feedback loops I am aware of. It was out of sync using a buffer size of 4096 until I learned about this issue. 64 buffer sounds perfect, Alternately, now I know I can disable audio card buffer setting. But as this issue was CV inside a Combinator from another author, one will have to be cognizant and look inside any Combinators you may load to see if there is much CV usage, then set your buffer to 64 or disable audio card buffer setting as needed.

cv.jpg
You do not have the required permissions to view the files attached to this post.
:PUF_balance: http://www.galxygirl.com -- :reason: user since 2002

User avatar
MattiasHG
Posts: 256
Joined: 16 Jan 2015

Post 16 May 2019

wendylou wrote:
15 May 2019
Well here's the CV in my dual arp Combinator, all stock devices. It's just sequence control, mod wheel, and pitch wheel, plus one Subtractor LFO that adjusts the Freq on an ECF-42. So no feedback loops I am aware of. It was out of sync using a buffer size of 4096 until I learned about this issue. 64 buffer sounds perfect, Alternately, now I know I can disable audio card buffer setting. But as this issue was CV inside a Combinator from another author, one will have to be cognizant and look inside any Combinators you may load to see if there is much CV usage, then set your buffer to 64 or disable audio card buffer setting as needed.


cv.jpg
You're sending a send fx signal out of the line mixer and back into a channel on the line mixer, thus creating a feedback loop since that channel can also be sent to the same fx again. Just like using a mix channel as a send return in the big mixer. Hope that makes sense!

As you know, this is the reason why we didn't want to increase the internal 64 sample buffer to begin with but now at least you have the option. :)

User avatar
diminished
Posts: 949
Joined: 15 Dec 2018

Post 16 May 2019

MattiasHG wrote:
16 May 2019
wendylou wrote:
15 May 2019
Well here's the CV in my dual arp Combinator, all stock devices. It's just sequence control, mod wheel, and pitch wheel, plus one Subtractor LFO that adjusts the Freq on an ECF-42. So no feedback loops I am aware of. It was out of sync using a buffer size of 4096 until I learned about this issue. 64 buffer sounds perfect, Alternately, now I know I can disable audio card buffer setting. But as this issue was CV inside a Combinator from another author, one will have to be cognizant and look inside any Combinators you may load to see if there is much CV usage, then set your buffer to 64 or disable audio card buffer setting as needed.


cv.jpg
You're sending a send fx signal out of the line mixer and back into a channel on the line mixer, thus creating a feedback loop since that channel can also be sent to the same fx again. Just like using a mix channel as a send return in the big mixer. Hope that makes sense!

As you know, this is the reason why we didn't want to increase the internal 64 sample buffer to begin with but now at least you have the option. :)
Using the send return would work though, right?

User avatar
MattiasHG
Posts: 256
Joined: 16 Jan 2015

Post 16 May 2019

diminished wrote:
16 May 2019
MattiasHG wrote:
16 May 2019


You're sending a send fx signal out of the line mixer and back into a channel on the line mixer, thus creating a feedback loop since that channel can also be sent to the same fx again. Just like using a mix channel as a send return in the big mixer. Hope that makes sense!

As you know, this is the reason why we didn't want to increase the internal 64 sample buffer to begin with but now at least you have the option. :)
Using the send return would work though, right?
Yes, it would.

mcatalao
Posts: 996
Joined: 17 Jan 2015

Post 16 May 2019

I think other possible "problem" is also that refeeding the pitch bend and mod wheel might be creating additional interference. This connection is not a feedback loop because all 4 cables go from output to input of the devices, but still lots of paralel cv connections between the 2 devices...

Could you try feeding the mod wheel through the combinator programmer to the devices, bypassing the RPG-8?

Hope it helps.

  • Information
  • Who is online

    Users browsing this forum: CommonCrawl [Bot] and 3 guests