Crg

panup


I made a null test with X1 and X2. Result: 16 bit/44100 Hz mix was perfect silence. Audio Engines of X1 and X2 output do identical mixes.

I not sure I understand this null test thing. Wouldn't every setting have to be identical in each track?

Yes, that is the point.  Eliminate (or make identical) all of the variables so that you can isolate and test only the differences between the X1 and X2 engines.  In other words, if everything is identical between a project in X1 and the same project in X2, a null test will reveal whether or not there is a difference that can be attributed to their respective audio engines.

So... if the default settings in X1 And X2 are different, you would hear something?

panup


+1 Same here. More things can be done during playback without dropout.

Yes, that is exactly what I experienced.
And what Stevec said: When gapping occurs it seems to be less obtrusive. That must mean dropouts are smaller.
 
The audio engine has improved.

Exactly!    They also seem less "sharp", if that makes sense, as though they're somehow softer.
 

I agree with everyone saying the audio engine feels more gapless - I was thnking that when doing my first mix in X2. And with the improved workflow I say that I can probably work 20%-40% more efficient in X2 than X1.

ProjectM


I agree with everyone saying the audio engine feels more gapless - I was thnking that when doing my first mix in X2. And with the improved workflow I say that I can probably work 20%-40% more efficient in X2 than X1.

Darn you Project I was just about to post it sounds 20 to 40% better. LOL

Actually you are not too far off!

LOL! Well dunno if it sounds 20-40% better.

But here's some "math": if I work 20-40% more efficient with the software, in other words faster, if I spend the same amount of time with a project I will add 20-40% more work to it and by that, have a finished track that sounds 20-40% better.

Yeah I guess it does sound 20-40% better

Project this may sound weird but I agree with your math. I knew the new math would come in handy some day. 

im not convinced with this null test thing, i cant understand how one file can sound better that the other file, rendered in an improved or different engine, and with that improved sound quality, there would be phase differences from the other file, of the exact same source, when doing a null test to see if it cancels out

There are some major updates/rewrites to the internals and engine that greatly improve realtime operations. The following areas were updated. 

1. Realtime synth triggering and timing. The mechanism to send realtime events to instruments was reworked to make MIDI timing buffer size independent. Synths can now be triggered with the same timing accuracy irrespective of the audio buffer size. 

2. Made the following operations fully asynchronous and seamless, minimizing interruptions in audio. (this makes gaps almost imperceptible). Many changes were made to make them more lightweight.

- Inserting virtual instruments, dragging to create new instruments from browser, also minor improvements for inserting audio plugins. 
- undo and redo of inserting effects and synths 
- Unloading and reloading virtual instruments in the rack. This operation can now be done during playback
- Quick Freeze and unfreeze no longer interrupt audio, so you can actually A/B the results during playback
- Changing loop points on the timeline doesn't require a gap
- Audio metronome no longer requires a gap to turn it on or off
- Preferences dialog can now be opened when browser or groove matrix preview is active (it will stop preview and automatically resume when you close the dialog)

3. Improvements for operations that require small interruptions in audio playback for re-synchronization. For these we now adopt a new "adaptive smoothing" technique to minimize the perceived effect of the operation. This kicks in whenever any perceivable interruption in audio is detected and subtly smooths out the transition. This makes many edit operations feel seamless now.  
     
4. Non linear playback engine rework. This component is responsible for all non linear playback of audio and MIDI including events triggered from the browser and Matrix.

Among hundreds of improvements, a big change was reworking the scheduler, making it synchronous with the audio engine. This makes triggered events from the matrix sample accurate and independent of the audio buffer size.

5. Previewing events in browser now caches the first 16 events and will avoid reloading them when triggered. This makes auditioning loops seamless.