Comparing DAW's

Sorry, not buyin' it. The test is flawed. The three mixes are obviously not volume-matched (which you can confirm by the spectral displays), which is why the nulled version sounds like a quiet version of the original mix. 

cclarry2
There was a thread a while ago that said "all of the major DAW's sound the same",
and had a bunch of tests, and blurbs, and etc...

And that thread had some right tips...
 

but this gentleman actually does a valid test between Reaper, Samplitude, and Sonar, and the results
speak for themselves.  Of course their will STILL be "naysayers", and we are all 
entitled to our opinion.  But I think the "Science" is pretty clear on the matter!

The "Science" is pretty clear. As long as it stay scientific...
This gentleman has no idea how to do correctly what he has done and that is the only result which speak
for itself. Any Noob can see and hear that within the first 6 minutes, bitflipper has already written why.
 
If someone is really interested in scientific differences between Sonar and REAPER, here are some facts:
I) Sonar always work sample accurate. REAPER by default does not and keeping sample accuracy in it is not a simple job. 
* What that means? Any waveform in Sonar is aligned with project dependent fixed sample rate. You can not start a clip in the middle of a sample, any material is converted to the project sample rate before insertion and you can not change the sample rate of the project. In REAPER there is no project sample rate, you can put any waveform with any sample rate at any position.
* What happens when the material is not sample aligned? The waveform has to be sample aligned before any processing and so REAPER does sample rate conversion or shifting. What is used depends from many parameters and operation. For example if material is just played and the target has the same rate, REAPER just shift the material to the nearest sample. If material has significantly different sample rate (f.e. 44.1 vs 96kHz) and the target is rendering, REAPER does full sample rate conversion according to the export settings, with low pass filtering when desired.
* What are possible artifacts from not aligning to samples? Not bit accurate output, f.e. Recording from played output, Glue and Render will not zero (up to full signal level!). Potential phase shifting during some operations, at sub-sample level. For signals over 15kHz with 44.1kHZ current rate some alien can hear the difference...
* How to null? Import all waveforms to the first measure. Do not use small clips or check they all are sample aligned position in REAPER.
 
II) Pan laws are not matching, even when they have (almost) the same names. Interesting that different Pan laws settings in REAPER can be closer to Sonar Pan laws.
* What that means? With the same material and Pan position, signal level will be different.
* How to null? Select matching settings. Note that "all at 0db and center" is not a warrantee you do this right.
 
Not particular DAW specific rules for comparison:
III) Many plug-ins have a "random" component. In some that is obvious (the reason for AUX in Sonar...), in other the difference in the output is subtle but exist.
* How to find such plug-ins? Duplicate the track, shift the material (sample accurate) and separately render both tracks. Import the result and try to null. If you can not find nulling position, there are some random/time dependent effects in the chain.
 
IV) some parameters apart from plug-in own "preset" can influence the result from it. Including information provided to the plug-in by the DAW.
* how to check? render one track which is known to null when rendered in both DAWs (so (I) and (II) is already checked). Do the same after inserting plug-in(s) in question. Check they also null.
 
V) each DAW has "fast" settings for playback. Some of them can be hardcoded and not exposed in parameters.
* How to avoid that? Compare rendered results only, carefully matching rendering settings.
 
----
Will DAWs null binary? No. Different programs do not match at bit level when using floating point arithmetic, even if the algorithm is the same. Null means very small difference in the signal level, close to the bit depth noise (-96dB for 16bit, -144dB for 24bit). Note that some DAWs show any deviation from absolute zero in level meters while other display the same signal as "no signal".
 

First time I ever saw Studio One was at a demonstration at NAMM. I actually won a copy in a raffle there. The presentation was done by a very excited fellow who bounced around like he'd had too much caffeine, and repeatedly stated that "I don't know much about software, but to my ear Studio One just sounds better!".
 
He really didn't need to add the "don't know much about software" part, as that was obvious. But to my amazement, most of the people in the audience nodded in agreement. Yes, they said, it does sound better! This in a very noisy trade show hall with sound played over a little PA system. The power of suggestion.
 
I am reminded of the hilarious episode of The IT Crowd where Jen breaks the internet.