Search | Statistics | User Listing Forums


Forums supporting
reViSiT (http://revisit.info)
and MIVI (http://mivi.nashnet.co.uk)
nashNET Forums ->  reViSiT - Tracking Software for VST hosts -> Testing & Development -> View Thread

You are logged in as a guest. ( logon | register )

reViSiT v1.1.3 Pro update released
Jump to page : 1 2
Now viewing page 1 [25 messages per page]
View previous thread :: View next thread
   reViSiT - Tracking Software for VST hosts -> Testing & DevelopmentMessage format
 
chrisnash
Posted 2009-06-11 12:17 AM (#14737)
Subject: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England

Hi everyone,

I'm delighted to present you all with reViSiT 1.1.3 Pro, featuring all the goodies mentioned recently in the Development Blog, including:

* Improved audio performance (up to 2x) and sound quality.
* Improved compatibility with many hosts, notably SONAR and those using variable buffer sizes.
* Improved host sync. engine.
* Fixes for the audio engine and keyboard handler.

Download reViSiT v1.1.3 Pro from http://www.nashnet.co.uk/experiment/download.asp.
(not registered? Register for free, as part of the reViSiT Experiment)

Enjoy!,
Chris

Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-11 8:53 AM (#14739 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
It sounds like quite a drastic update (already read it from the development blog) and will try it out tonight! Good to hear the program's stability/performance remains top priority as well, and not just adding new features. Nice one in advance Chris!
Bookmark and Share Top of the page Bottom of the page
fegi
Posted 2009-06-11 4:48 PM (#14740 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released



Veteran

Posts: 102
100
Location: Austria
nice update chris! but one problem so far: i can't record loop selections in my host anymore...revisit stops playback when the host is jumping back to the beginning of the loop selection.
Bookmark and Share Top of the page Bottom of the page
fegi
Posted 2009-06-11 5:23 PM (#14741 - in reply to #14740)
Subject: RE: reViSiT v1.1.3 Pro update released



Veteran

Posts: 102
100
Location: Austria
hmm...actually revisit is completely out of sync with reaper now playback doesn't even start at the right pattern most of the time.
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-11 7:35 PM (#14742 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
I have to say that the new update actually plays more out of sync than before as well unfortunately
for the rest it's all good for now though!
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-11 8:24 PM (#14743 - in reply to #14742)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
Yeah, the new sync. algorithm has a couple of problems - due to the conflict of trying to play a tracker pattern faithfully while trying to maintain sync. with the host. I'll take a look at it as soon as I can.

Chris
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-11 9:45 PM (#14744 - in reply to #14739)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England

Okay, I've switched back to the older sync. method (albeit a slightly cleaner version with the improved tempo support). It meant I had to disable the retro-rendering feature, but I'll look at renabling it down the road.

Just redownload reViSiT 1.1.3 Pro from the experiment site.

Best,
Chris

Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-17 9:28 AM (#14749 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
I also noticed revisit is playing parts of patterns it shouldn't play with midi triggered patterns, like the first note of a pattern is heard after the ending of it.
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-17 1:09 PM (#14750 - in reply to #14749)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
By design, reViSiT will repeat a MIDI-triggered pattern, if the triggering note is held for longer than the pattern itself. It's possible reViSiT is receiving or recognising the termination of playback too late. A number of factors can affect how accurate the note-off timing is, such as audio buffer size (and latency) and how the host delivers MIDI.

I have already found a timing bug in Cubase SX3 that affects the "punctuality" of its time information, but I will have another look at the reViSiT code in question to see if your problem lies elsewhere. In the meantime, I have a few of questions:

1) What host (and version) are you using?
2) What is your ASIO buffer size / latency?
3) How are you triggering the patterns? Manually (by hand) or using notes on a MIDI track? If the latter, how 'accurately' timed are those notes?

One easy option would be to disable the pattern repeat, but I'd like to rule out any other problems before I consider it. Does anyone rely on these repeats?

All the best,
Chris
Bookmark and Share Top of the page Bottom of the page
arnorhs
Posted 2009-06-17 3:04 PM (#14751 - in reply to #14737)
Subject: Re: reViSiT v1.1.3 Pro update released


Member

Posts: 10

Hi Chris

I'm having a problem opening older projects. Most of the projects I've got now have been created using revisit 0.95b or something like that, but when I open projects using 1.1.3, I get this error:
The plug-in "reViSiT" could not be found for VST Synth 1!

What can I do?
Bookmark and Share Top of the page Bottom of the page
arnorhs
Posted 2009-06-17 3:05 PM (#14752 - in reply to #14737)
Subject: Re: reViSiT v1.1.3 Pro update released


Member

Posts: 10

Can I have both plugins active in the VST folder?
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-17 3:21 PM (#14753 - in reply to #14750)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
1. I'm using windows xp sp3 with cubase sx3.1.1 build 944
2. my dma buffer size (latency) is 256 samples. I have a latency of about 5-6 ms with an m-audio asio driver
3.I trigger the patterns by notes on a midi track. they used to play quite accurately until 1.1.3 it seems, before they were usually correct or due to other heavy processing (cpu wise) had a delay of about 1/2 steps in the pattern on 170 bpm for example.

HOWEVER... at times the patterns play pretty good. I'm certain my cpu ain't it (it's a new bought quad core and until now hasn't reached the level of 50% usage yet) so I hope I have to adjust something in my cubase settings or my dma buffer size. I did just switch to a new computer and everything, so I actually hope it's not revisit but me having forgotten to configure some stupid small thing somewhere.
The reason I'm adressing the issue is that during an audio mixdown the inconsistently played patterns show up in the output as well, and I didn't experience this before. This version sometimes also plays part(s?) of different pattern it isn't supposed to be playing (for instane a kickbass playing which should be silent).

Edited by Reactor Grits 2009-06-17 3:36 PM
Bookmark and Share Top of the page Bottom of the page
arnorhs
Posted 2009-06-17 3:21 PM (#14754 - in reply to #14737)
Subject: Re: reViSiT v1.1.3 Pro update released


Member

Posts: 10

it seems I can have both plugins active, I'm making experiments in saving the revisit module via 0.95 and opening again using 1.1.3 inside the same project
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-17 4:33 PM (#14755 - in reply to #14754)
Subject: Re: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
Reactor Grits:

The sync. algorithm's now pretty similar to the one in 1.1.2, and very little has changed with MIDI triggering, so I'm not quite sure where I should start looking for the problem. It's interesting that you're using SX3, being the host I singled out for dodgy time-keeping, but if earlier versions of reViSiT work fine, I guess it must be in the 1.1.3 code. If it turns out to be SX3, then the official Steinberg line is not going to help your very much - "Yeah, that was fixed in v4"! Otherwise, your buffer size sounds fine, and I don't think it has anything to do with the CPU.
I don't know how much time I'll have tonight to look at the problem, but can you just confirm that you re-downloaded 1.1.3, after I replaced it with the newer build? Also, can you tell me more about the moments the "wrong" pattern appears to be played - are they reproducable, momentary, prolonged? Could it not still be a pattern looping back on itself after a period of silence? As for the first notes being repeated at the end, is it just briefly (e.g. just the first pattern frame), or do you get the whole row / several frames?

Arnor:

Make sure your reViSiT plugin file is always called "reViSiT.dll" - as the host might use the filename to match a plugin named in a project to that in the VST plugins directory. Hence, if the old project file used a plugin called "reViSiT095.dll" or similar, that's probably what it's looking for now. Note, simply renaming the newest plugin to match the old filename isn't the best permanent solution, as the reBUS plugin requires that the reViSiT plugin be called "reViSiT.dll" in order to work. But if this is the case, you could temporarily rename the file to match that in your old projects, just so that you can load your project and save/export your reViSiT module to a .ZIP file. Then you can rename it back to "reViSiT.dll" and load up the reViSiT module from the .ZIP - though you might have to re-wire any audio outputs you'd setup - and then save the project.
If this is not the case, let me know what host you're using and I'll delve deeper into the problem! If the project file's not too big, feel free to email it to revisit [at] nashnet.co.uk, and I'll take a look.

Best,
Chris
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-17 5:06 PM (#14756 - in reply to #14755)
Subject: Re: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
Thanks chris. 'wrong' patterns are played randomly I noticed. But like I said I just reinstalled a whole new pc which I might have to tweak a bit more, and when i got the new pc I didn't do much revisit since I'm finishing my semester at uni (still). I did replace the 1.1.3 with the new build. It might could have something to do with midi settings, yet I have no clue what it could be. I'll keep you posted when i get some more research done not concerning revisit in cubase. one thing I noticed is that cubase is glitching sometimes and that could be the timing problen. but the random playing with midi triggered pattersn not playing what they should is a mystery. maybe you have a link to the old revisit 1.1.2 so I can see or it is really revisit doing this?

I also noticed my pc sometimes plays patterns a tad to fast when they're triggered (only at the start, which is of course looping the first bit and then playing the next one).

Edited by Reactor Grits 2009-06-17 5:07 PM
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-17 5:31 PM (#14757 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
it seems that turning off 'multi processing' in the expert settings of the vst audiobay (in device settings) fixed things! RV is playing in sync now and seems to play patterns correctly again. I wonder why though.. anyway, sorry for the hassle chris, I'm just really glad for the moment things look better now!


Edited by Reactor Grits 2009-06-17 5:45 PM
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-17 6:34 PM (#14758 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
however, now I can't use multi processing and now cubase maxes out with a minimum of settings. anyone any ideas?
Bookmark and Share Top of the page Bottom of the page
arnorhs
Posted 2009-06-18 1:42 AM (#14759 - in reply to #14737)
Subject: Re: reViSiT v1.1.3 Pro update released


Member

Posts: 10

I'm having timing problems in some songs, but not in others.

After a few minutes revisit and cubase SX3 get out of sync with each other. This reminds me of what happened when I was using renoise tracker. I'd export a few tracks into cubase and then even though both programs had a tempo of 143 (to name any) ... cubase and the audio tracks were not in sync... then I discovered a small display window they have in revisit where you can read the "real tempo" of the tracker... which would usually be 143.3883 - then I'd put that time into cubase and everything would be in sync...

I don't know if it's related, but I'd think different tracking software might have similar timing issues...

Notice that this is a song I've imported from revisit 0.95 to 1.1.3 -- so the timing was fine before, but now it's off...

And the tempo I'm using is 143

I'm using M-Audio FW 1814, 44100 hz, with ASIO drivers running with a buffer size of 256
Cubase SX3
Bookmark and Share Top of the page Bottom of the page
arnorhs
Posted 2009-06-18 1:55 AM (#14760 - in reply to #14737)
Subject: Re: reViSiT v1.1.3 Pro update released


Member

Posts: 10

and btw. I can use both 0.95 and 1.1.3 at the same time, but not both buses... but that's probably not a supported feature, right ?
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-18 9:05 AM (#14761 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
I remember it's not safe to load multiple instances of revisit anyway man, tried it and things got nasty so won't do that anymore.

@ Chris: I read through the fixes for the 1.1.3 release, and saw you added support for multi-processing hosts. Isn't there something wrong with that, since unticking the option in cubase makes everything run smooth again (with about 25% of my cpu power)?
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-19 2:42 PM (#14762 - in reply to #14761)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
Reactor Grits:

Here's a link to reViSiT 1.1.2 Pro, so that you can check whether your problems are new with 1.1.3 or not:

http://www.nashnet.co.uk/experiment/reViSiT1.1.2Pro_X.exe
http://www.nashnet.co.uk/experiment/reViSiT1.1.2Pro_X.zip

I've tinkered a bit with the code to iron out some kinks, but I'm no closer to discovering why "multi processing" doesn't work. There was a similar issue with SONAR in previous versions, but I thought I'd nailed that with this version (Owen?).

One thing you might try is giving longer buffer sizes a try (though only in 1.1.3!) - the only thing I can think of is that multi-processing gives the plugins less time to do their processing, which can lead to glitches if they overrun.

It'd be useful if VST hosts published what the heck they do differently when "multi processing" is selected!

Cheers,
Chris
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-19 4:42 PM (#14763 - in reply to #14762)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
Arnor:

The songs going out of sync. was the problem I addressed earlier in this thread - and should be fixed in the new build. The original v1.1.3 took the host's time at the start of playback and then kept its own time. However, the new v1.1.3, that's available for (re-)download now, is once again slaved to the host at all times - I can't really see how it could gradually go out-of-sync (after a few minutes). Can you confirm you're using the newest build? The reViSiT.dll file date should be 11/06/09.

Best,
Chris
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-19 5:12 PM (#14764 - in reply to #14737)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
Thanks for the effort Chris. I tried running a project with 1.1.2 and the problem persists. Increasing the buffer size didn't do much as well, same goes for switching my asio drivers (to asio4all for example) or disabling all plugs except for revisit. Honestly, I'm not sure I ever had multi processing turned on with my last setup in the first place, so that makes deduction even harder (I might hadn't all along).
Maybe since I now have a quad core instead of a dual core cpu cubase has trouble handling it? I've scouted fora for answers for days now with no results. For the time being I'll just disable multi processing since things still work pretty smoothly then. When I do find the bugger some day, you'll be the first ones to know of course!
Bookmark and Share Top of the page Bottom of the page
chrisnash
Posted 2009-06-19 6:02 PM (#14765 - in reply to #14764)
Subject: RE: reViSiT v1.1.3 Pro update released



Developer

Posts: 746
50010010025
Location: England
Reactor Grits:

Just to confirm; disabling "Multi processing" solves all the problems you mentioned?

(a) the glitches
(b) first notes played after last notes
(c) random 'wrong' patterns during silences
(d) the tad-too-fast starts

I hope to have a new update out soon, and I'll do some extra testing of the MIDI-triggering, so it'll be useful to know what I should be on the lookout for.

For the record, I've had "Multi Processing" enabled in my copy of SX3, for a while now, and it seems to be behaving. Perhaps it's a quad-core thing...? You could try, running it with Multi Processing enabled, but then using Task Manager to switch the CPU Affinity for the cubasesx.exe process - selecting between 1 and 4 to see if anything gets fixed.

Best,
Chris
Bookmark and Share Top of the page Bottom of the page
Reactor Grits
Posted 2009-06-19 6:38 PM (#14766 - in reply to #14765)
Subject: RE: reViSiT v1.1.3 Pro update released


Regular

Posts: 60
2525
As far as I know, all 4 issues melt like snow in the sun when I disable multi processing. I messed adound with the affinty as well, and as long as I select one cpu (any) there's no trouble, but when I add more the problems rise again. If more people using quad core cpu's and have these trouble, it could be revisit, but otherwise I'll just have to keep on tweaking and hopefully get lucky of course.
Bookmark and Share Top of the page Bottom of the page
Jump to page : 1 2
Now viewing page 1 [25 messages per page]
Jump to forum :
Search this forum
Printer friendly version
E-mail a link to this thread

(Delete all cookies set by this site)