u-he

core · Post by **core** » Fri Jun 26, 2015 7:56 am

Reveal Sound also says to use fixed size buffer for Spire to sound correctly. I thought that setting was on by default in FL Studio...

keel · Post by **keel** » Fri Jun 26, 2015 8:05 am

This is interesting. Does this apply on FX's too, because I swear that some plugins sounds different on FL and S1. So should we use use fixed buffering on all the synths and effects in FL? Any words from IL about this?

exmatproton · Post by **exmatproton** » Fri Jun 26, 2015 8:12 am

keel wrote:This is interesting. Does this apply on FX's too, because I swear that some plugins sounds different on FL and S1. So should we use use fixed buffering on all the synths and effects in FL? Any words from IL about this?

It applies to everything inside the Fruity wrapper. However, it seems that fast rate LFO's and modulation stuff is affected the most.

I always enable this option, so i did not think about this at first.

FL Studio is awesome, but the use of wrappers has it's downsides...

EvilDragon · Post by **EvilDragon** » Fri Jun 26, 2015 8:16 am

FL is buffering things dynamically, which some plugins don't like, that's why the option is there. I think dynamic buffer sizes have the benefit of reducing CPU usage, but the caveat is that they might not catch all the audio-rate changes that could happen within a single ASIO buffer in the plugin.

KingTuck · Post by **KingTuck** » Fri Jun 26, 2015 3:19 pm

I've only been running fixed size buffers with plugins that need it to function (ie some Tone2 stuff and Synthmaster etc). I tested out the patch that the OP posted with buffers on and off. Not really any audible difference for me here at all. Perhaps because of how my ASIO settings are? Does anyone have a patch that would more immediately show the difference? Setting up a hard vibrato on the pitch of an oscillator also doesn't produce different results for me.

KingTuck · Post by **KingTuck** » Fri Jun 26, 2015 3:28 pm

Would people affected by this please try different project timebase settings? Both with and without the size buffer option ticked.

This is definitely setting dependant. I can't reproduce it in any synth currently.

exmatproton · Post by **exmatproton** » Fri Jun 26, 2015 4:28 pm

KingTuck wrote:Would people affected by this please try different project timebase settings? Both with and without the size buffer option ticked.

This is definitely setting dependant. I can't reproduce it in any synth currently.

I have my timebase on the highest setting, and when i disable the wrapper buffer setting, the patch sounds almost the same as the one the OP gave as an example. When i enable the buffer setting, the audio seems the same as in other DAW's.

woodsdenis · Post by **woodsdenis** » Fri Jun 26, 2015 4:32 pm

In the OP files the Studio One file is 1 db louder than the FL one also.

Peaks of Studio One L-3.76/R -3.8 FLoops L -4.78/R -4.44

KingTuck · Post by **KingTuck** » Sun Jun 28, 2015 1:20 am

Weird. Does anyone know if the use of variable size buffers like the FL Wrapper uses is uncommon? Is it simply a byproduct/necessary consideration of wrappers opposed to native VST support? Any other DAWS use them? As a precaution I've turned on the buffers in Hive, I still can't replicate the bug, but to be safe.

exmatproton · Post by **exmatproton** » Sun Jun 28, 2015 11:12 am

KingTuck wrote:Weird. Does anyone know if the use of variable size buffers like the FL Wrapper uses is uncommon? Is it simply a byproduct/necessary consideration of wrappers opposed to native VST support? Any other DAWS use them? As a precaution I've turned on the buffers in Hive, I still can't replicate the bug, but to be safe.

I think it has something to do with the language FL Studio uses; Delphi.
It has some major upsides, like a truly butter smooth UI (imho really important, for a good/smooth experience). But is has it's downsides as well. All data has to pass the wrapper (I/O) first, before it gets processed, so it takes some processor cycles for sure...

The functions the wrapper is able to add, weighs heavier for me, compared to the extra cycles it's adding. Modern processors are quite able to proces the data fast enough. But i can imagine some purists don't like this at all!

KingTuck · Post by **KingTuck** » Sun Jun 28, 2015 10:11 pm

Interesting. I knew of FL Studio's Delphi status, but I didn't know much of that's implication on VST implementation. I find it a bit confusing about plugins that recommend turning off threaded processing in the wrapper settings. Would this truly ever yield different sound output? As I take it the idea is that some plugins will perform better without it, however, in all cases of stress testing I ALWAYS get better performance with that option on.

The new Spire beta for example, requires fixed size buffers to function. It doesn't seem to perform better without threaded processing (although Reveal recommend you turn that option off), and also it doesn't seem to need the maximum buffer size option (which further increases plugin latency beyond what fixed size buffers necessitate). Again, that being what Reveal recommend. Weird stuff.

Elektronisch · Post by **Elektronisch** » Wed Jul 01, 2015 7:33 am

Even fixed size buffers helps, still on Studio One same patch sounds less muffled, a little bit cleaner and has a bit more crisp.

exmatproton · Post by **exmatproton** » Wed Jul 01, 2015 7:41 am

themachinelt wrote:Even fixed size buffers helps, still on Studio One same patch sounds less muffled, a little bit cleaner and has a bit more crisp.

Maybe upsampling when rendering (so offline rendering at 88.2 or 96) can help with that. The folding back in FL Studio @ 44.1 kHz can be heavy on some plugins.

Downsampling afterwards can be done in Audition for instance.

Ivan_C · Post by **Ivan_C** » Wed Jul 01, 2015 3:52 pm

Would it be possible that these differences are related with different panning laws in each DAW ?

KingTuck · Post by **KingTuck** » Wed Jul 01, 2015 5:40 pm

themachinelt wrote:Even fixed size buffers helps, still on Studio One same patch sounds less muffled, a little bit cleaner and has a bit more crisp.

Are you just qualifying this by ear? These differences are best served by actual measuring, going by ear is useless (although certainly cause to measure things and check what's up).

Wolfen666 wrote:Would it be possible that these differences are related with different panning laws in each DAW ?

Possibly, I'm unsure how fixed size buffers would come into play in that case though.

It would probably be helpful if Urs posted Viktor's findings on this.

u-he

KVR Audio

U-He Hive sounds different in FL Studio and Studio One?