Bug with Reverb

[stuey]

It = Solina in my previous comment.

Solina hasn't been updated since 19th Nov which was version 1.0.0. 409. That's a month ago.

Thanks, Stuart   

[kickwizard]

Buffer size i found personally on my setup i needed 256 samples @ 48000hz

Not because i had crackling or dropouts but i was stacking 3 instances of solina with 4 instances of the matrix and 3 instances of spark and 128 was not sufficent to do the job  and i was getting realtime vst peformance overload in cubase

The higher the sample rate the quicker it goes though the buffer

Personally i dont think 256 is a massive buffer size and gives me sub 10 ms latency

48000hz suits my production skills and space requirments

Whats everyone else running ?

[stuey]

48k?

Are you working with broadcast or dubbing video?

Thanks, Stuart

[kickwizard]

No im just a hobbist by "suits my production skills" i mean i dont need to go higher for mastering like some do like 192

I have only been back tinkering for 2 years after a long break so still a lot to learn

I noticed from my point of view 48k is more forgiving when being encoded in lossy compression for soundcloud and such

Biggest plus really is it goes faster though the buffer so slightly less latency than 44100

[stuey]

Hi,

Are you going from 48k delivering at 44.1?

There's an entire thread that can be devoted to sample rate and src, but it depends on so many different factors. Different audio apps will give you different src which will in turn sound different on different apps. That part is subjective.

But as a 'rule', one should really be going in multiples of the base sample rate. To add even more confusion, high end clocks will deliver a better more transparent sound at 44.1 than lower end gear will at 4 x the sample rate (given that most low end gear might not even go to 96k)

But, if it sounds good then keep doing it

Thanks, Stuart

[boof]

Im not a mac user and havent had the issues people have described in this post
Solina is not ram heavy for me

Nor me, I just think, like other VI's it needs a decent amount installed.

Thanks, Stuart

To be clear, the cracklin' reverb issue has nothing to do with ram, unless 32gb isn't enough.

It's a buffer issue as was stated in my 3rd post on this thread. Increase your buffer, it goes away.

Thanks, Stuart

That's not a fix, thats a workaround.  I'm just throwing in my 2 cents here to continue to draw attention to it for new users (like myself) who are heading to these forums for answers.

I've reported the issue to Arturia and reported it here.  Job done

Indeed.
I have uninstalled it entirely until I start to see people posting feedback that the issue is resolved. There is no reason why I should have to push the buffers up to some level that starts to show (hear) noticeable latency. I also try to keep it at 128 or 256 with 512 being my overhead room for when I do have projects that have allot of VSTs and VSTis.

I haven't read every post on this thread  so apologies if I missed it but has Arturia even acknowledged this as an issue and have stated that they will fix it?

JB

[stuey]

There is no reason why I should have to push the buffers up to some level that starts to show (hear) noticeable latency.

There is, otherwise it won't work correctly.

I haven't read every post on this thread  so apologies if I missed it but has Arturia even acknowledged this as an issue and have stated that they will fix it?
 

Well, not really. There's never admission of any issues, just updates.

Thanks, Stuart

[jackn2mpu]

Hey.. No problem. Glad you got something sorted.

It did notice that the Solina V is a bit memory heavy. In Pro Tools I had to increase my hardware buffer which I rarely do with 48Gb to spare.
So it may be that although you have a fair chuck of RAM, the Solina V wants more.

Thanks, Stuart

I don't have a dog in this fight as such, just cruising through, but felt I had to write to correct something. Hardware buffer in PT has nothing to do with your memory - those are two separate things. Hardware buffer affects latency which is due to the audio I/O used on a computer system, not ram.

Also buffer changes depend on which version of PT one is running. PT11 only has a setting for record buffer as playback buffer is taken care of automatically inside PT. In PT10 changing buffer size would affect both recording and playback which is why during mixing you'd see people raise their buffers to allow for more cpu intensive stuff like eq/compression/etc.

[stuey]

Virtual instruments have never liked Pro Tools, whatever the version.

Maybe some crossed wires here in that sometimes "we" talk about standalone and the AAX (or whatever other flavour of plug) in the same thread, and they work differently.

RAM affects the standalone and i've tested it, and obviously the more you have the better. I found significant differences with the intel macs, one with 8Gb and one with 48Gb - But that goes without saying, right? With 8Gb of RAM i found i could not really play it without it cracking or freezing under about 512, although that still has issues. Increasing to 1024 solved that. With the 48Gb machine, I could happily go to 32 and it not affect the sound.

As far as the plug in inside a DAW, there's so many things that affect it. Buffer, playback settings, track count, Core count, other plugs, voices, delay compensation, even having the bloody plug in GUI open.

Again, with the same two machines I had totally different experiences with the exact same set up of Pro Tools running the same session.

Playback / Mixing you can easily take it to 1024 and be happy that you are going to play everything fine, I've never had issues there. Recording and trying to get decent latency is different.

From my tests, the main reason this is so is due to having the GUI open, closing the plug GUI when playing results is a far smoother experience.

Thanks, Stuart
 

[jackn2mpu]

Virtual instruments have never liked Pro Tools, whatever the version.

Maybe some crossed wires here in that sometimes "we" talk about standalone and the AAX (or whatever other flavour of plug) in the same thread, and they work differently.

RAM affects the standalone and i've tested it, and obviously the more you have the better. I found significant differences with the intel macs, one with 8Gb and one with 48Gb - But that goes without saying, right? With 8Gb of RAM i found i could not really play it without it cracking or freezing under about 512, although that still has issues. Increasing to 1024 solved that. With the 48Gb machine, I could happily go to 32 and it not affect the sound.

As far as the plug in inside a DAW, there's so many things that affect it. Buffer, playback settings, track count, Core count, other plugs, voices, delay compensation, even having the bloody plug in GUI open.

Again, with the same two machines I had totally different experiences with the exact same set up of Pro Tools running the same session.

Playback / Mixing you can easily take it to 1024 and be happy that you are going to play everything fine, I've never had issues there. Recording and trying to get decent latency is different.

From my tests, the main reason this is so is due to having the GUI open, closing the plug GUI when playing results is a far smoother experience.

Thanks, Stuart
 

I know standalone and plugin versions like aax64 in PT work differently. No problem there. What I was trying to correct was the assertion that buffer setting in PT and ram are related and they aren't. Being you tried this on two different computers I submit that your comparison is invalid UNLESS they both had everything the same save for amount of ram. Same cpu, same audio I/O, same version of operating system, etc. The difference between 8 gig and 48 gig in standalone won't affect performance - it can't. Remember that Yosemite takes all memory that isn't being used for itself, unlike previous versions of OSX.

You shouldn't have to close a plugin gui for any daw work smoothly. To do so indicates poor programming on the part of the plugin developer. Most of the time one wants to have a plugin gui open, especially a synth, if one wants to or needs to make changes while playing - it's that simple.

[stuey]

Being you tried this on two different computers I submit that your comparison is invalid UNLESS they both had everything the same save for amount of ram. Same cpu, same audio I/O, same version of operating system, etc.

Of which they were. You can't run comparison tests unless they are.

The difference between 8 gig and 48 gig in standalone won't affect performance - it can't. Remember that Yosemite takes all memory that isn't being used for itself, unlike previous versions of OSX.

I'm using last mountain Lion release, stable for ages don't see the point in upgrading. 

You shouldn't have to close a plugin gui for any daw work smoothly. To do so indicates poor programming on the part of the plugin developer. Most of the time one wants to have a plugin gui open, especially a synth, if one wants to or needs to make changes while playing - it's that simple.

Agreed, but my real world testing proves otherwise

Thanks, Stuart

[stuey]

And also, Pro Tools 10 isn't qualified with anything higher than Mountain Lion.

Doing some more tests with Solina, I can use it in Pro Tools at 192khz and it stands up fine with certain reverbs but others just die straightaway.
Dropping it down to 44.1khz and the second i lower the samples to even 512, I get break up.

I can record at 192khz normally and it's fine but even at sample rates 4 times lower the reverb breaks the audio up.

The usual caveats apply in that the faster the drives you use, the better results you will get. I'd be interested in what Arturia tested this on, what machine-drive set up they used.

Thanks, Stuart

[jackn2mpu]

And also, Pro Tools 10 isn't qualified with anything higher than Mountain Lion.

Doing some more tests with Solina, I can use it in Pro Tools at 192khz and it stands up fine with certain reverbs but others just die straightaway.
Dropping it down to 44.1khz and the second i lower the samples to even 512, I get break up.

I can record at 192khz normally and it's fine but even at sample rates 4 times lower the reverb breaks the audio up.

The usual caveats apply in that the faster the drives you use, the better results you will get. I'd be interested in what Arturia tested this on, what machine-drive set up they used.

Thanks, Stuart

True about PT10 not being qualified on anything higher than ML however it does run and run quite well. The only real problem is the uninstaller program doesn't work and would have to be uninstalled manually which is not impossible and Avid even gives steps on how to do that.

As to test setup on what Arturia used that would be nice to know. But like I said it runs just fine here on my setup which is a mid-2011 27 inch iMac with a 3.4 GHz quad core i7 with 32 gig ram and a 2 gig 7200 rpm system drive. My external drives are a LaCie Thunderbolt connected drive with two physical 7200 rpm drives (one for PT to record to and one for samples). I have a Glyph GPT50 2 gig 7200 rpm drive connected via the Apple Thunderbolt to FW800 adapter that is a backup for the same drive in the LaCie. Also have a second Glyph GPT50 2 gig 7200 rpm FW800 drive as a backup for the other LaCie drive half. Sometimes I switch the roles and use the Glyphs as mains and the LaCie as backup. I use an Avid 11 Rack connected via USB2 as my audio interface. When tracking I run at 64 samples and mix at 1024 samples; all sessions are 24 bit 44.1 KHz sample rate.

And it doesn't matter if a vi gui is open or closed - things run just fine. However I close all vi and fx gui's before closing a session because PT sometimes quits to desktop. Doesn't crash it quits.

[stuey]

Thunderbolt connected drive

I'm guessing that's what they test on, TB connected drives will give you far better throughput than say USB or FW. My audio drive is 6TB raided via eSATA. It's me 300 odd Mb second throughput. I think the guys running USB and internal drives for their audio also with I/O on the same bus are going to be running into issues because of data rates. I don't record at 192k I just did this as a full on test which it stood up to quite well. Dropping back down to my 88.2k means I can run multiple arturia plugs at the same time.

Also, Spark doesn;t work with anything higher than 96k because of the internal audio samples.

Cheers, Stuart 

[ben arturia]

An update will come ASAP to fix this reverb bug.
Please be patient.