I did notice that. I tried the test on just the kick and got the same results as you. I think you may be on to something here. I hadn't considered this before because I assumed that since I am not panning anything that the panning law did not come into play. But maybe it does. This is great. I hope that this is what is going on. If nothing else, it is most excellent that we are learning more about all this. I think it is really cool either way.
 
Thank you so much for taking the time to conduct your own investigation.

Yea its pretty cool . 

If you have dithering turned on in either DAW or both, even if it's the same kind of dithering, then the test is invalid. Dithering introduces random noise. Random anything is a big no-no when trying to see if two signals will null in a test. To compare the summing of two DAW's you have to make sure EVERYTHING is the same with no tolerances or random factors. Even if you think the dithering is doing nothing because of your sample rate, it's not a scientific test until you turn all dithering completely off. 

But it does nulls with Dithering turned on. I had tested other tracks as well which nulled despite dithering. So its not something to be worried of in this case. 

So I just completed a null test comparing Sonar and Studio One.
 
Original tracks from VST's in Sonar with no processing at all in either DAW.  Exported with no dithering in either Sonar or Studio One and I'm getting complete nulls from individual tracks and from the combined final mix.  I agree dithering needs to be off on both.
 
That was a very cool process to see.  Here are my files if you'd like to check as well.
 
https://www.dropbox.com/sh/a287g58shf5jiny/AAA-_1QyxcFy-L90SUm7tvioa?dl=0
 

To: sonarman1,
 
I found out how to change the pan law in Studio One. You have to use a plugin...ha ha ha. It is called Dual Pan. I placed it on the "Main" and set it to the same as SONAR which is 0dB center.
 
I also found out how to remove the dither on Studio One (which is enabled by default.) You have to go to Studio One>Options>Advanced>Audio or something like that. Uncheck the Dithering option.
 
So I did all that. I removed dithering from both SONAR and Studio One. I changed the Pan Laws to be identical. I ran the test again, test 3, and they still will not null. Man, I was so hoping that they would. I want so badly to put this to rest. But no matter how hard I try, they will not null. I even double-checked the Mix FX is not on.
 
This is maddening. The only conclusion that I can draw from the hard science results is there is indeed a difference in summing between the two. There just is. After the panning law adjustment, I felt that things got a little closer, but still will not null, proving that they are not identical.
 
To others: I am sure Jeff is a nice guy, but to simply dismiss this out of hand is now beyond ridiculous. I am doing the procedures correctly. I have checked pan law, dithering, Mix FX, no plugins, no panning. What else is there? I have shown my documentation. I have shared my results. I have provided the test files. If anyone still wants to dismiss this, please provide your evidence as I have done. Taking ones word for it is not an option. Show the work and the science as I have done. I can't come up with any other conclusion. 
 
But like I said, I will still use Studio One. I accept the difference as irrefutable (unless someone can prove otherwise, show me how my tests are flawed, if they are at all) and will move onward and forward. 
 
This has been a great enlightening and learning exercise. 

doncolga
So I just completed a null test comparing Sonar and Studio One.
 
Original tracks from VST's in Sonar with no processing at all in either DAW.  Exported with no dithering in either Sonar or Studio One and I'm getting complete nulls from individual tracks and from the combined final mix.  I agree dithering needs to be off on both.
 
That was a very cool process to see.

That would be another twist in all this, if Studio One can null with SONAR using VST's (I assume you mean VSTi's) but not with audio.
 
Can you show your test and show that they are nulling?

Sylvan

doncolga
So I just completed a null test comparing Sonar and Studio One.
 
Original tracks from VST's in Sonar with no processing at all in either DAW.  Exported with no dithering in either Sonar or Studio One and I'm getting complete nulls from individual tracks and from the combined final mix.  I agree dithering needs to be off on both.
 
That was a very cool process to see.

That would be another twist in all this, if Studio One can null with SONAR using VST's (I assume you mean VSTi's) but not with audio.
 
Can you show your test and show that they are nulling?

I don't have a screen capture app on my machine, but here are my files:
 
https://www.dropbox.com/sh/a287g58shf5jiny/AAA-_1QyxcFy-L90SUm7tvioa?dl=0
 
I feel a little silly asking, but just wanted to double check that you exported all totally from the very beginning again on your test with the dithering off?

doncolga

Sylvan

doncolga
So I just completed a null test comparing Sonar and Studio One.
 
Original tracks from VST's in Sonar with no processing at all in either DAW.  Exported with no dithering in either Sonar or Studio One and I'm getting complete nulls from individual tracks and from the combined final mix.  I agree dithering needs to be off on both.
 
That was a very cool process to see.

That would be another twist in all this, if Studio One can null with SONAR using VST's (I assume you mean VSTi's) but not with audio.
 
Can you show your test and show that they are nulling?

I don't have a screen capture app on my machine, but here are my files:
 
https://www.dropbox.com/sh/a287g58shf5jiny/AAA-_1QyxcFy-L90SUm7tvioa?dl=0
 
I feel a little silly asking, but just wanted to double check that you exported all totally from the very beginning again on your test with the dithering off?

At this point, nothing is silly to ask...ha ha ha. Yes, I did. I made sure they are sample-accurate as far as that goes. I can happily say that Studio One and SONAR are exporting from the same spot to the same spot with spot-on accuracy by default. And yes, I had dithering turned off on both. Thank you for asking.
 
I am starting to wonder a little bout Cubase. If I demoed that, I would run the same tests and see what happens. Out of curiosity I may do that. But I am so short on time now and have gotten a bit behind with client work because of having to prove my case...ha ha ha.

Sylvan: I haven´t tested but I think the plugin you mentioned (Dual Pan) does not change the S1 pan law if you just put it on the master. If you move the knobs then yes.

As an electronic engineer I agree with all the science to check / comparison. And that's good to review in looking at any two DAWs for peace of mind. However I'm not sold on the idea that SO Pro would be a product that would sound inferior in a direct comparison.

Otherwise that surely would have come up in the forum long ago. 😊 Please don't think I'm knocking this effort. I'm following to see how it turns out...

I am a nice guy!   The only thing here at my end is I do get complete nulls even with 32 tracks of instrumentation and vocals etc.. Plus live rooms as well.  The final mix being a very complex wave.  I see this comparison as leaning much more on the side of these things actually cancelling out meaning that DAW summing engines are in fact pretty similar.   The lower end of your result is not doing a good null.  The top end seems to be much quieter indicating it is frequency dependent. 
 
In my test I used Studio One as the playback machine.  Have you tried doing the nulls in Studio One instead. Although you will know more about your own DAW though which is good.  You could also do something without knowing it in Studio One.  Use the Mix Tool plugin and select the Invert Phase option.  Have you checked track types e.g. stereo/mono, panning laws, any sign of some processing you do not know is happening.  Any form of saturation. e.g. Mixbus has permanent saturation and for this test you have to ensure it is all removed providing the cleanest signal path.  Also line up the audio by eye seriously zoomed in as well.  Don't just rely on the timing of the rendering.  In that regard DAW's can be different. 
 
I find with virtual instruments the results are a little different rendering out what a virtual instrument might actually be doing at any point in time.  It is one situation where there it is more likely to be not a perfect null.  In synthesisers for example there are so many things that are freewheeling e.g. modulation sources.  You will never get them to be in perfect alignment in two different DAW's at two points of time.  Audio is more likely to be fixed and never change.  What some plugins might do though could also be similar to the virtual instrument case.  Others would be totally static e.g. an EQ.  But we are not using plug-ins in this case anyway. 
 
I get a different sound when I sum all the stereo stems coming out of Studio One in my Yamaha digital mixer compared to all that being done ITB.  I swear the image is wider.  It has nothing to do with the sound or bottom end etc, it is just a stereo thing.  It sounds like something nice is happening inside the digital mixer but it should not in fact.  I have been digging around to see what it could be and not found anything yet.  I started to think I was imagining it.  Then the engineer who engineered the last album by James Taylor said exactly the same thing in an audio magazine article.  That entire album is digital end to end and he mixed it on a Yamaha digital mixer because he felt the image sounded different.  (It is a fine example of yes you can get the sweetest most analog sounding result if you are skilled enough all in the digital world.  It really shows that those who think you can only get this sound via tape are wrong!) I need to do a null test on that one.  I do think this thread highlights the power of a null test actually. 
 
If you have a really strong idea about how something should sound, I don't think the summing engine is going to get in the way of achieving it.  I will have to find the tracks again in my library but I might when I can get to it try it again and see.  I have only got Producer 8.5 though. Can anyone advise me where the Sonar pan law settings are in 8.5. That would help. 

My results with original files, using Reaper...
 
For SONAR (only)
With Reaper (panning law 0.0).
 
TEST 1 : REAPER-01 to SONAR-01, NULL achieved! 
Note: before rendering it is clipping +1.8 ! After rendering (so with digital clipping) the result is NULL.
I am sure Sonar master also was +1.8, you just have not shown that in the video...
 
TEST 2: REAPER-03 to SONAR-03, NULL achieved!
There is no clipping. But there is a trick, in EXACT volume settings:
Kick  0.00dB
Overhead -13.00dB
Snare -7.97dB
Note that the last one is critical! -7.98 (and obviously -8) there is no perfect match.
 
In general, the accuracy of parameters in DAWs is 3-4 digits. I do not remember exact numbers for Sonar, can check if you want, something in the region 10-14bit resolution.
NOTE 1: that has nothing to do with the audio engine accuracy.
NOTE 2: -8.21dB and -12.2dB are both written using 2 meaningful digits, while -20.0dB has only 1(!).
 
For null testing, it is better to mix entering precise values with keyboard (and matching the same way, that is how I have found exactly -7.97dB)