|
|
|
05-10-2018, 05:02 PM
|
#1
|
Human being with feelings
Join Date: Jan 2012
Posts: 1,180
|
v5.81pre12 - May 10 2018
v5.81pre12 - May 10 2018
- + Automation items: pasting points in AI-only mode creates a new automation item
- + Envelopes: clear old point selection when duplicating points via mouse modifier [p=1987813]
- # Automation items: improve behavior when pasting points into exactly-fitting items and when pasting at the end of an item
- # Track selection: notify MIDI editor/projbay/etc from various actions of item selection change when item selection undo is disabled [p=1987752]
- # linux: fix jack MIDI control surface support crash
Changelog - Pre-Releases
Generated by X-Raym's REAPER ChangeLog to BBCode
|
|
|
05-10-2018, 05:19 PM
|
#2
|
Human being with feelings
Join Date: Sep 2013
Posts: 657
|
Quote:
[*]+ Envelopes: clear old point selection when duplicating points via mouse modifier [p=1987813]
|
Thank you very much!
|
|
|
05-10-2018, 05:22 PM
|
#3
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Next build will likely remove some of the long term testing features and be called 5.81rc1, I think. I’m on the fence about the Live FX multiprocessing improvements going into 5.81 — any thoughts?
|
|
|
05-10-2018, 06:02 PM
|
#4
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
Yes. Any improvment about Reaper lag when lots of Windows FX are opened/docked ?
|
|
|
05-10-2018, 06:33 PM
|
#5
|
Human being with feelings
Join Date: Jul 2009
Posts: 7,568
|
Quote:
Originally Posted by Justin
Next build will likely remove some of the long term testing features and be called 5.81rc1, I think. I’m on the fence about the Live FX multiprocessing improvements going into 5.81 — any thoughts?
|
not sure how to properly test - What should we look for?
|
|
|
05-10-2018, 06:45 PM
|
#6
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by EpicSounds
not sure how to properly test - What should we look for?
|
I suppose if you use "Live FX multiprocessing" and have seen improved performance in the prereleases (for armed tracks etc especially) with PDC/folders/routing, when compared with the release versions, that'd be a +1. If you've seen worse performance, that'd be a -1, and if you've seen instability/hangs/any other bad sort of regression that'd be -10.
|
|
|
05-10-2018, 06:53 PM
|
#7
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
It seems better for armed track which sends audio to lots of tracks (with FX). There is less glitch sound with heavy CPU loaded projets.
But I don't see any improvment when lots of Windows FX are opened/docked. Sometimes, I have to close all Windows FX if I want to listen a sound without glitchs.
Last edited by ovnis; 05-11-2018 at 07:53 AM.
|
|
|
05-10-2018, 08:33 PM
|
#8
|
Human being with feelings
Join Date: Jan 2016
Location: Los Angeles, CA
Posts: 3,116
|
Justin I have a project that is the perfect example of the PDC/Sends thing that I can make a video of. Basically a VOX folder is sending to 3/4 of an INST folder (masking treatment), and there's heavy/big PDC plugins like Soothe, GW Mixcentric on the folders. When engaged, that send creates 200% RT CPU. The moment it's disabled, back to minimal CPU use. If you print the VOX bus and put it in the INST lane (essentially recreating the send routing) it's minimal CPU. Something about the send/folder relationship is crippling things horribly.
Would this be helpful at all? It seems directly tied to your improvements.
|
|
|
05-11-2018, 01:23 AM
|
#9
|
Human being with feelings
Join Date: Jul 2015
Location: Stockholm, Sweden
Posts: 1,343
|
For me the Live processing improvements works fine and with no bugs so far.
A typical mastering project of mine is:
Pitching track sending to a LOOP track as in analog loop. On this loop track I usually have a couple of quite CPU intense plugs with lots of PDC, like a Linear ohase EQ for example, then I have a ReaInsert plugin for the analog loop. Now this loop track goes to another POST track with more plugins to do fine tunings of the analog bounce.
The way I do it is that I record the output of the loop track and this is when I get lots of RT CPU. In a specific and really heavy project where I also had some really heavy RX plugins on the pitching tracks the RT CPU when from around 80% down to 65% with one of these pre releases. Btw, bypassing reainsert doesnt seem to change the RT CPU much.
One thing though. If I instead do an online render of the loop track the RT CPU is almost zero, or in the single digits. What is the reason for this? These to things are essentially doing the exact same thing...
Why don't I just render instead of recording then? I could, and have done that but I prefer recording for workflow reasons.
__________________
Magnus Lindberg Productions - VRTKL Audio - Redmount Studios
magnuslindberg.com
|
|
|
05-11-2018, 05:15 AM
|
#11
|
Human being with feelings
Join Date: Oct 2017
Location: Black Forest
Posts: 5,054
|
@Justin, concerning the sluggish MIDI Editor, consider this fixed.
I just put a KVM in between my monitor (displayport) and my PC and the MIDI editor now behaves completely fluent. It is really scary, what impact displayport can have!
|
|
|
05-11-2018, 07:54 AM
|
#12
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by ferropop
Justin I have a project that is the perfect example of the PDC/Sends thing that I can make a video of. Basically a VOX folder is sending to 3/4 of an INST folder (masking treatment), and there's heavy/big PDC plugins like Soothe, GW Mixcentric on the folders. When engaged, that send creates 200% RT CPU. The moment it's disabled, back to minimal CPU use. If you print the VOX bus and put it in the INST lane (essentially recreating the send routing) it's minimal CPU. Something about the send/folder relationship is crippling things horribly.
Would this be helpful at all? It seems directly tied to your improvements.
|
Sorry, is this issue on 5.80 or 5.81pre12 or both?
|
|
|
05-11-2018, 08:17 AM
|
#13
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by mlprod
For me the Live processing improvements works fine and with no bugs so far.
A typical mastering project of mine is:
Pitching track sending to a LOOP track as in analog loop. On this loop track I usually have a couple of quite CPU intense plugs with lots of PDC, like a Linear ohase EQ for example, then I have a ReaInsert plugin for the analog loop. Now this loop track goes to another POST track with more plugins to do fine tunings of the analog bounce.
The way I do it is that I record the output of the loop track and this is when I get lots of RT CPU. In a specific and really heavy project where I also had some really heavy RX plugins on the pitching tracks the RT CPU when from around 80% down to 65% with one of these pre releases. Btw, bypassing reainsert doesnt seem to change the RT CPU much.
One thing though. If I instead do an online render of the loop track the RT CPU is almost zero, or in the single digits. What is the reason for this? These to things are essentially doing the exact same thing...
Why don't I just render instead of recording then? I could, and have done that but I prefer recording for workflow reasons.
|
Tracks which are record armed are automatically not processed by “anticipative FX rendering” when not rendering. Also, any tracks they send to inherit that limitation, which is why you see the increase in RT CPU...
|
|
|
05-11-2018, 08:39 AM
|
#14
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
Quote:
But I don't see any improvment when lots of Windows FX are opened/docked. Sometimes, I have to close all Windows FX if I want to listen a sound without glitchs.
|
Somone can explain me what causes this issue, please ?
It can happen when only one window FX is visible ! If some plugs are not directly useable (when they are docked and not visible for exemple), it should take less ressource than if they are directly useable, no ?
|
|
|
05-11-2018, 09:00 AM
|
#15
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by ovnis
Somone can explain me what causes this issue, please ?
It can happen when only one window FX is visible ! If some plugs are not directly useable (when they are docked and not visible for exemple), it should take less ressource than if they are directly useable, no ?
|
Hmm might be plugin dependent...
|
|
|
05-11-2018, 10:00 AM
|
#16
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
It seems to happen with fabfilter plugs (pro Q2, for exemple).
|
|
|
05-11-2018, 10:13 AM
|
#17
|
Human being with feelings
Join Date: Jul 2014
Location: London
Posts: 733
|
I can't imagine this being related to the pre at all, but I haven't had this happen before.
Strange behaviour with a region copy. Look at the track at the bottom as I make a region copy. An item gets duplicated in place but doesn't move with the region.
Apologies for the large gif
|
|
|
05-11-2018, 10:19 AM
|
#18
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
256 instances of fabfilter ProQ2.
1 Z3ta.
All Windows FX are docked (Reaper lags) :
All Windows FX are closed (Reaper is OK) :
|
|
|
05-11-2018, 10:23 AM
|
#19
|
Human being with feelings
Join Date: Nov 2007
Location: France
Posts: 919
|
Quote:
Originally Posted by Justin
Next build will likely remove some of the long term testing features and be called 5.81rc1, I think. I’m on the fence about the Live FX multiprocessing improvements going into 5.81 — any thoughts?
|
Is it related with hanging notes with some plugins.
If it is, I can test and see if there are improvements.
|
|
|
05-11-2018, 10:39 AM
|
#20
|
Human being with feelings
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
|
It's not.
|
|
|
05-11-2018, 10:59 AM
|
#21
|
Human being with feelings
Join Date: Jul 2015
Location: Stockholm, Sweden
Posts: 1,343
|
Quote:
Originally Posted by Justin
Tracks which are record armed are automatically not processed by “anticipative FX rendering” when not rendering. Also, any tracks they send to inherit that limitation, which is why you see the increase in RT CPU...
|
Ok cool. Btw bypassing plugins on the POST track also lower RT CPU a lot being after the recorded track. Normal?
__________________
Magnus Lindberg Productions - VRTKL Audio - Redmount Studios
magnuslindberg.com
|
|
|
05-11-2018, 11:00 AM
|
#22
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by Stroudy
I can't imagine this being related to the pre at all, but I haven't had this happen before.
Strange behaviour with a region copy. Look at the track at the bottom as I make a region copy. An item gets duplicated in place but doesn't move with
Apologies for the large gif
|
Can you send the .rpp to support at cockos? Thanks!
|
|
|
05-11-2018, 11:55 AM
|
#23
|
Administrator
Join Date: Jan 2005
Location: NYC
Posts: 15,721
|
Quote:
Originally Posted by ovnis
256 instances of fabfilter ProQ2.
1 Z3ta.
|
256 instances of the UI opened?
This might be something they could optimize -- not updating the UI if the window is not visible?
|
|
|
05-11-2018, 01:50 PM
|
#24
|
Human being with feelings
Join Date: Aug 2007
Location: Luxembourg/Spain
Posts: 1,922
|
Quote:
Originally Posted by Justin
256 instances of the UI opened?
This might be something they could optimize -- not updating the UI if the window is not visible?
|
AFAIK, they also use opengl so maybe the gpu driver is involved too..
__________________
Reaper for Linux Documentation (WIP). Software: Archlinux/KDE, Fabfilter FX, Komplete 8, Nebula, Schwa/Stillwell, T-racks Max/Amplitube/SVX, etc. Gear: i7-2600k/4700HQ/16GB, RME Multiface/Babyface, Behringer X32, Genelec 8040, etc. :)
|
|
|
05-11-2018, 03:11 PM
|
#25
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
Quote:
256 instances of the UI opened?
|
256 instances shared between 34 windows FX docked.
Quote:
This might be something they could optimize -- not updating the UI if the window is not visible?
|
Good idea. Not updating the UI if the plug is not visible inside the visible Windows FX, too ? Maybe it's already the case.
Quote:
they also use opengl so maybe the gpu driver is involved too..
|
True. It seems better when I disable the Fabfilter GPU-accelerated graphics acceleration via regedit.
With 765 instances, I have succeeded to crash Reaper.
Last edited by ovnis; 05-11-2018 at 03:26 PM.
|
|
|
05-11-2018, 03:17 PM
|
#26
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
If I don't use "Only Allow one Fx chain window open at a time" is it because, when I open a new FX chain window, there is lag time + ugly visual animation when the plug appears.
With the new optimisation, is it not possible to pre-load FX (and never unload) chain window to obtain zero latancy when we open a new FX chain windows ?
Last edited by ovnis; 05-11-2018 at 03:29 PM.
|
|
|
05-11-2018, 03:41 PM
|
#27
|
Human being with feelings
Join Date: Apr 2014
Posts: 151
|
Quote:
Originally Posted by ovnis
256 instances shared between 34 windows FX docked.
Good idea. Not updating the UI if the plug is not visible inside the visible Windows FX, too ? Maybe it's already the case.
True. It seems better when I disable the Fabfilter GPU-accelerated graphics acceleration via regedit.
With 765 instances, I have succeeded to crash Reaper.
|
On my Win 10 x64 system, under a heavy load (end of mix, lots of tracks and VSTs)... Opening even 1 Fabfilter plugin window noticeably slows computer. CLose window and its fine again. Ive always immediately chalked it down to Fabfilters fancy GUI displays and the CPU/GPU straining to update that in realtime with the plugs..seems like there is something there
|
|
|
05-11-2018, 03:47 PM
|
#28
|
Human being with feelings
Join Date: Oct 2011
Posts: 2,924
|
Quote:
Originally Posted by Endisol
On my Win 10 x64 system, under a heavy load (end of mix, lots of tracks and VSTs)... Opening even 1 Fabfilter plugin window noticeably slows computer. CLose window and its fine again.
|
Same thing here.
|
|
|
05-13-2018, 03:39 AM
|
#29
|
Human being with feelings
Join Date: Jan 2007
Posts: 496
|
Quote:
Originally Posted by Endisol
On my Win 10 x64 system, under a heavy load (end of mix, lots of tracks and VSTs)... Opening even 1 Fabfilter plugin window noticeably slows computer. CLose window and its fine again.
|
So you guys happen to have Nvidia cards?
I have noticed this with all Waves and Goodhertz plugins. (they both use OpenGL)
As soon as the GUI is open, the RT cpu meter jumps +10%.
I have found the solution on my system.
You have to turn off threaded optimization in Nvidia control panel:
https://prnt.sc/jh901r
Maybe this is useful info for the developers too?
|
|
|
Thread Tools |
|
Display Modes |
Linear Mode
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
All times are GMT -7. The time now is 11:27 PM.
|