From "secrets" in other DAWs concerning audio engines:
 
-------------
Studio One - Audio Dropout Protection
Cubase - ASIO Guard
Reaper - Anticipative FX processing
Sonar - <not exist> (or?)
 
Impact:
While not without some side effects, that technology allows to stay with lower ASIO buffer length. Relevant during recording and avoid switching driver settings / freezing heavy tracks.  
-------------
PDC (Plug-in Delay Compensation) approach
 
Sonar  - global?
Reaper - signal chain independent, with "shifting backward" when possible.
Studio One - ? can someone confirm ?
 
What I mean:
 
Create a project with 2 tracks. On one track put Cakewalk LP EQ with "high" settings (0.7+ sec plug-in delay).
On the second just open ProChannel EQ (in other DAWs some other effect without delay, or LP EQ without LP mode).
Now start play and change some parameter of the second track FX.
In Sonar, resulting sound is delayed by the first track EQ. In Reaper, resulting sound is not delayed (so the same as without the first track FX). Interesting that volume changes on the second track has immediate effect... So may be just a bug, or there is some option?
 
Impact:
Global:
a) the need to explicitly use "PDC override" during recording and resulting sound is desynchronized with the rest of the project (not only mastering plug-ins have delays, while other delays are usually small, they still can be relevant for recording)
b) using plug-ins with big delays on tracks reduce "responsiveness" of all tracks/plug-ins (during mixing)
Signal chain independent:
a) PDC can be "full automatic" and live chains are in sync with the rest
b) delayed plug-ins on tracks do not influence other tracks, so can be used even on backing tracks for future recording
 

Jeff Evans
I have been enjoying solid stability from day one for me in 2010.  It was young then but did the job real well for me. It has evolved into a magnificent program now.  They have quite a long road map and they are adding things in all the time as they travel. V4 will be pretty awesome I can guarantee it.  There might even be an update prior too. 
 
I push the audio side hard for days on end as I have just done mixing 9 songs.  Many hours.  Never flinched once. Had Scheps Omni here there and everywhere.  Good source of bundled plug-ins too.  In composing music mode I drive the external setup pretty hard all day long too.  Up to 10 hardware synths and never an issue.  Plays back perfect from anywhere.  Great external midi timing to the metronome.  The midi timing externally feels independent.  It just does its thing.  As you work the audio side harder e.g. with virtual synths as well as audio, with all the external midi going on for me, you just adjust the buffers and things to get playback smooth with the audio.  The external midi is always in sync with that. 
 
I love the gapless performance.  I am someone who is often doing something while it is in play or record.  The other day I was visually editing audio events while the cursor was just behind playing everything I had just edited.  All without any issues. You can loop sections drag effects in and audition patches while it is in play or  record.  I was dropping EQ's and things onto individual events all while in play and looping.  Setting up reverbs at the same time and tuning the spaces etc..
 
It never crashes for me under any of these conditions.  My RME audio interface and Steinberg Midex 8 midi interface all work seamlessly with the software.  I would use it all live too without hesitation.  Although Ableton is fantastic live.  It has a similar solid audio gapless engine as well.  It is becoming more linear in its own way though.  It would be the only other DAW I would be interested in learning.  It's way different.  Jarre is doing his current albums using Live controlling external midi as well as in the box.  
 
I am excited about the integration between Presonus digital mixers and Studio One.  (DAW Mode)  Some real interesting things are happening there.  

Jeff - I hope I'm not reading too much into your post, but are you saying you really, really like Studio One?

azslow3
 
PDC (Plug-in Delay Compensation) approach
 
Sonar  - global?
Reaper - signal chain independent, with "shifting backward" when possible.
Studio One - ? can someone confirm ?
 
 

see this:
https://www.youtube.com/watch?v=nkKT7ATxJOE
 
it's related.

Well another post disappeared.. I didn't say any bad words. Not sure what's up.
 
Delay compensation in Studio One seems to be simply a temporary bypass while recording. Works best with the fx built into the Presonus interfaces and obviously works better on faster interfaces. It basically bypasses the parts of the interface that cause a delay. It doesn't work well on everything. Primarily only on their interfaces with built in effects. In other words you still hear yourself playing through the effects but they are bypassed from the audio while recording and engaged while playing back. It isn't adding any additional resources , only making more intelligent use of the resources at hand. Its really a more complicated routing scheme. You can get a similar result if you record with hardware based effects on any other program. 
Using an RME interface will certainly give you low latency on just about anything. Using a low end interface won't give the same result. Don't be fooled. There are advantages in some instances primarily using high end Presonus hardware. If you use Studio One with an interface that isn't compatible you won't notice much of an improvement.
It really all depends.

batsbrew

azslow3
PDC (Plug-in Delay Compensation) approach
Sonar  - global?
Reaper - signal chain independent, with "shifting backward" when possible.
Studio One - ? can someone confirm ?

see this:
https://www.youtube.com/watch?v=nkKT7ATxJOE
it's related.

Sorry, it is not... It is related to the first, "Audio Dropout Protection".
 

Starise
Delay compensation in Studio One seems to be simply a temporary bypass while recording.

It can happened it is Sonar like... But delay compensation I am talking about is not audio interface related. Plug-ins with delays do not have the delay because they are "slow", they just need "future" information to made the decision for "now". When used in the recording chain, such plug-ins introduce the "latency" (up to 1 second or even more!) with any interface and any DAW. My question is how DAW deal with such plug-ins which sit outside of current chain you work with (for recording or mixing).
 
I have describe a very simple test how to check any DAW for that.

azslow3
 
Plug-ins with delays do not have the delay because they are "slow", they just need "future" information to made the decision for "now". When used in the recording chain, such plug-ins introduce the "latency" (up to 1 second or even more!) with any interface and any DAW. My question is how DAW deal with such plug-ins which sit outside of current chain you work with (for recording or mixing).

I don't think you would use a plugin with an intentional delay caused by look-ahead buffers, like the L-Phase plugins, for recording or mixing. Or for comparing PDC or dropout protection, or whatever.
 
They are intended for mastering audio where you are not monitoring a live input or a soft synth.  So you are no longer in a real-time situation when mastering, where large look-ahead buffers should not be an issue.  Everything should play back perfectly in sync, according to the delay incurred by the largest plugin buffer in the project.

I have been playing around with low latency monitoring as well.  Those figures in the video are a bit misleading and they are not good either. He is getting very poor latency performance there.  I can set an input latency of 1.5 mS and output latency of 2 mS using the RME set to 32 samples.  It still plays back a massive session with ease. 
 
Firstly the fastest interface ever from the Presonus range with Studio One is the Quantum. Latencies under 2 mS can be achieved.  Others though, work super well and they re RME and also Focusrite. 
 
With my RME I get the best performance with dropout protection set to Minimum and just monitoring the fastest or best buffer setting of 32 samples. If I have got a big instrument like Serum for example being played live from my controller, I can get 4 mS of latency so the synth feels fast to me.  When I use a higher dropout protection and the green low latency instrument monitoring comes into play, the best latency I can get is 6 mS.  Still fast but higher than the 4 mS without using dropout protection.  So in some cases the non use of LLM actually yields a better result.  This is like the other way around at the moment and they are refining what is going on here.  The lowest instrument latency should come when the green LLM is engaged.  This would also apply to using nice guitar plugins while tracking.  You could do it and it would be fast. I know that with vocals I can get  the feeling of zero latency monitoring now.  If I increase my normal audio buffer setting to a higher value e.g. 64 or 128 samples then I get a lower LLM green mode while using a medium to high dropout protection. But if you interface can still play a big session at 32 samples then normal low buffer setting use beats the LLM green mode.
 
I recently recorded a band to a 24 channel multi session. On Logic using a Focusrite on thunderbolt at its lowest 32 sample buffer setting. With vocal being monitored through the software and live at the same time there was not even phasing in my case.  Meaning the vocals are getting through the DAW at 2 mS and under in my case.  A PA was used live in the room while recording.  I was able to create a special PA buss inside Logic and feed the PA from there.  Any track I simply turned up an send for was routed to the PA with no noticeable delay of any kind.  I had guitar players going through as well through amp/speaker simulators and effects such as reverb and delay. 
 
Being able to use Studio One on a Mac allows it to talk via thunderbolt to a Focusrite Clarett interface.  Here I can 1.5 mS of roundtrip latency which is amazing.  This is the fastest it has ever felt to me. i.e. playing virtual synths live.  No matter how big the virtual synth is.  Also on the Mac you can use two different audio interfaces at once.  One for inputs and the other for outputs.  I tested this with a RME Fireface 800 working on the Firewire 800 port at the same time as a Focusrite Clarett on the thunderbolt port.  That all worked sweet.  You cannot use all the inputs at once though, unless you create an aggregate device in the Apple audio setup, then you can. 
 
I think they are refining the whole low latency monitoring setup. 
 
 

abacab

azslow3
Plug-ins with delays do not have the delay because they are "slow", they just need "future" information to made the decision for "now". When used in the recording chain, such plug-ins introduce the "latency" (up to 1 second or even more!) with any interface and any DAW. My question is how DAW deal with such plug-ins which sit outside of current chain you work with (for recording or mixing).

I don't think you would use a plugin with an intentional delay caused by look-ahead buffers, like the L-Phase plugins, for recording or mixing. Or for comparing PDC or dropout protection, or whatever.
 
They are intended for mastering audio where you are not monitoring a live input or a soft synth.  So you are no longer in a real-time situation when mastering, where large look-ahead buffers should not be an issue.  Everything should play back perfectly in sync, according to the delay incurred by the largest plugin buffer in the project.

Do the test, and you will understand...
Dropout protection is different, it has almost nothing to do with that (except technical implementation which is in fact similar).
 
Sure no-one will use L-Phase for recording. But there are quite some compressors, limiters, effects which have 2-5ms delay. There was many threads with an advise "till you start mixing, do not use such plug-ins" (not an easy task in Sonar since it does not show current delay, nor which plug-ins have it).
But the whole point, WHEN you have such plug-ing, up to L-Phase (f.e. if your backing track is complete and you want play along...), some DAWs will keep everything in sync with your (low latency) live input. And some DAWs will not.
 
In practice, if you ever had to hit PDC button, you can benefit from the technology. If you do not use PDC button, you do not need that technology. But the button is there for a reason, right?
 

i pretty much run my setup the same way jeff describes....
 
i really have zero issues with latency,
and zero requirement for plugin compensation.
 
 

Those are impressive numbers Jeff. I wanted to clarify that I really intended to mention low latency monitoring in Studio One when I said delay compensation which is often mentioned in relation to plug ins. Most software also has plug in delay compensation built in and it usually works just fine. Of course, this all figures into the equation as well.
It's a play on words since all of it is designed to minimize the effects of hearing a delay when using effects or when
playing soft synths. If I'm using an older interface I simply turn off effects or monitor through hardware while recording. Those who play soft synths can record the midi and feed the audio output form their keyboards to stay in sync with the rest of the mix, then use that midi track for a synth of your choice. Not necessarily mandatory  to always use the soft synth during recording if you have high latency.
 With the new series 3 mixers you can "take hold" of the hardware effects inside of the mixer as if they were a part of Studio One. This applies to automation as well. The main drawback being if you port the mix to a collaborator in another studio you will need to print those effects first and he/she can't use them. The advantage being you can take the load off of your computer and practically eliminate delay. This all only applies to the most recent Presonus mixers.
The dust has barely settled on those. The other have built in effects but limited control/integration.