Reaper vs. Protools don't Sum the same (??)

[Drummerbman]

Before you say...
again?! "this topic has been discussed to death"...and
"summing is summing"...or "who cares... just mix" (which I happily do, thank you)

I did as much forum search field and googling as I thought fair and could actually not come up with too much.

For those of you familiar with the summing shootout on youtube put out by stream records... https://www.youtube.com/watch?v=-0SKwEI_7Wc&t

I did a follow up video on this.
I'm not here to promote the video...
I made the video (and it wasn't so easy lol)so I could have a legitimate discussion with those who would know far more than myself.

https://www.youtube.com/watch?v=VL_tL8vRiG8&t

Watch if this topic interests you at all...
Is it sampling errors?
Computing errors?
Hypothetically: Wouldn't the Eiffel Tower be affected if it was built on mac and cheese?

Not looking for troll and/or "know it all" answers...

Shout out to Justin!
Cheers my reaper dudes

[cyrano]

I don't know about nulling...

But your video is too loud. I guess you're used to high spl, as a drummer, but you're shouting. When you come from the Stream records video, that has a conservative level, you dive for the volume knob. Wherever it is.

[Judders]

Well, we can rule out pan law discrepancy because the levels are too low.

What version of Pro Tools was used? What is the audio engine bit depth of that version? Was there any dithering used? Was the dithered mixer used in Pro Tools?

[karbomusic]

I defer to my sig as my know-it-all remark. Because... I can find -175 dBFS issues in lots of scenarios and it wouldn't matter then either.

[poetnprophet]

I would love to hear a discussion on this, but I don't really think you're drawing a conclusion, or even making a statement about anything. Is one better than the other or not?

[karbomusic]

Originally Posted by poetnprophet

I would love to hear a discussion on this, but I don't really think you're drawing a conclusion, or even making a statement about anything. Is one better than the other or not?

Go find the most annoying sound you can, drop it in a project @ -175 dBFS, report back. I wasn't even going to post in the thread, but this is going to cause even more people to unnecessarily stress over something that is ridiculous - not the OP or their test being ridiculous, but the uncertainty it creates for zero good reason.

[azslow3]

I seems to be a troll... But I have watched the (second) video 2 times and digged the web site. I have not understood a bit. Where are the files? Where is the new comparison ?
The first video is clear. Without going deep into technical details, the man has matched everything correctly and demonstrated the result. But the second one puzzle me, sorry.

"Is the summing the same?". Summing is summing, 1+1=2. For some reason some people believe that "audio summing" in the digital domain is not mathematical summing. From my knowledge it is. Sample from one track + Sample in the next track + ... = summed sample.

Yes, what is going on around is another topic. Sampled signal has finite noise level (-96dB for 16bit, -144dB for 24bit, 32 bit FP format also has just 24bit precision, just for any signal level). Dithering increase the noise level to improve sonic perception. Pan law is not extremely well defined thing, different DAWs can scale the signal differently, even in case they call the preferences the same (I have tried to find REAPER preferences for all possible Cakewalk Sonar pan law preferences, only with partial success). Etc.

Sorry, I am just confused what this thread is about

[bigjoe]

Different Pan Laws...
I worked 10 years on ProTools and i'm using reaper since 4 years... really you can watch this if you are into the "DAW sound" conspiracy

https://youtu.be/CWdiOud20Es

[1111Eugene]

EVERY SINGLE DAW SUM THE SAME
https://www.youtube.com/watch?v=JrDpuToHEFQ

1+1 should be 2
if it's not in some DAW, then it cannot be used

[EvilDragon]

Reaper has options for track mixing bit depth in Project Settings->Advanced. That could also play a factor here.

But also, not all DAWs are the same due to a myriad of factors other than summing.

http://admiralbumblebee.com/music/20...aw-Differ.html

[bigjoe]

I watched now the video and, to me...that seems a lot like de-normalization noise (something that prevent the CPU to work on REALLY small numbers) and probably isn't something which is in the files them selves and has probably to do with the analyzer

[Judders]

There was a period of time when Pro Tools was behind the curve and still using fixed integer summing so measurable differences between, say, Logic, were a thing.

Those were the days when people were using the dithered mixer in Pro Tools:

Originally Posted by Bob Olhsson

...I find mixes made with the dithered mixer can sound less crunchy. I've called people up a number of times and asked them to check if they had the dithered mixer installed and if not, run their mix again with it. I've never been wrong about it not being used although I'm sure there are probably many times I haven't noticed.

https://repforums.prosoundweb.com/in...?topic=24748.0

[cyrano]

Anyone seen a decent comparison between Reaper and Mixbus?

It's the only one I think sounds different. Mind you, it's just an impression. I've never really compared, because of Karbo's sig.

[Judders]

Originally Posted by cyrano

Anyone seen a decent comparison between Reaper and Mixbus?

It's the only one I think sounds different. Mind you, it's just an impression. I've never really compared, because of Karbo's sig.

Bladerunner did. He thought there was a big difference in soundstage between the two, but then once he set pan law the same the difference was negligible, as far as I remember.

They have some modelling on the busses, but nothing you can't do with plugins.

[Drummerbman]

Thank you for all the replies and discussion.
I am sorry if the video or thread confuses some.
I rewrote the details under the video on youtube and hope that clears it up better for everyone.

My video directly responds to the "stream records" shootout video.
It's meaningless without it.
I hoped to cover all my bases on my findings and feelings about it.

The problem I had was when you type in google "protools vs reaper".
This "stream records" video is the number two thing that pops up... And as meaningless, petty and/or fearmongering as it may seem to some...I uncovered a flaw and inaccuracy in the information it was portraying.

If this topic doesn't interest you, it's understandable. For me, it's fun and interesting learning about these details... I'm the kind of engineer that doesn't worry about much but my instincts when it comes time to make music and record, but in my free time I like looking into even the minute things as I feel this train of thought keeps me learning and better at gauging the practical vs the hype.

The real reason I even started looking into this subject (which I have in the past) was my good friend who I got on Reaper almost 6 years ago has this new co-worker... This is an older gentleman who claims he's been a professional engineer all his life & is very knowledgeable about gear, etc ,etc...He told my friend "Just get on pro tools already... it sounds the best... it's got the best summing engine and "conversion""... and THEN continued on about how he could hear differences between protools vs reaper mixdowns and even claimed reaper "always has an 8k spike" on it's mixdown...

Now... lets just throw all that shit out the window and not even talk about it, but alas... this left my friend in great doubt and uneasy, he felt like he was just "schooled" by a "professional". Talk about fear mongering.


So here we are...
and here I am bring it all up...
If I didn't, I would just continue to lurk as I have for many years here..
I feel I know reaper pretty damn well. ;D


Back to the noise I found.
One person mentioned it may just be dither or noise shaping added and that may be exactly it. Can you turn these things off in protools like you can in reaper? I guess we wont know if stream records overlooked this and is the case.

@bigjoe,
I don't think it's the analyzer, I have been able to recreate a "true" null and analyze with the same tool on several other tests without extra signal showing up.

Cheers everyone.

PS. As for the dude who thinks I'm too loud.
"If it's too loud, you're too old" (cranks stereo) - Airheads
"Twist and Shout!"- Beatles
"Shout it.. Shout It, Shout it out loud! - Kiss ;P

[karbomusic]

Originally Posted by Drummerbman

Back to the noise I found.
One person mentioned it may just be dither or noise shaping added and that may be exactly it. Can you turn these things off in protools like you can in reaper? I guess we wont know if stream records overlooked this and is the case.

We can replicate this easily by placing something like Pro-Q on a track and inverting with zero changes on the EQ. So, yes probably denormaling, dither or similar unrelated to "summing". I have no idea why you are asking if such things can be turned off as there is literally no difference in any audible or practical imaginable way. In all honesty, until you followed up, I figured this was click bait.



I'm not sure anyone appreciates just how large -175 dBFS is (likely mentally impossible to even do so) and how this is likely much smaller than trying understand the difference between a gnat flying 10 feet away while standing beside the space shuttle when it's launching...



I don't think it even deserves the question/term of "summing" as it may not even be related tbh or which DAWs are involved.

[jerome_oneil]

Originally Posted by karbomusic

I'm not sure anyone appreciates just how large -175 dBFS is (likely mentally impossible to even do so) and how this is likely much smaller than trying understand the difference between a gnat flying 10 feet away while standing beside the space shuttle when it's launching...

I work with industrial radios all day long. We don't even get down to -175. Pretty much anything below -95 is considered weak, and I'm pretty sure that the radios don't even consider it signal below -120. And that's in RF where the reference is way lower than audio.

-175 is one of those calculus problems that starts out as "As the signal approaches infinity...." It's Dr. Evil asking for one hundred beeeelion dollars....

[karbomusic]

And that's using power right? Or am I missing something which is entirely possible - I ran out of zeros LOL...

[Drummerbman]

Originally Posted by karbomusic

We can replicate this easily by placing something like Pro-Q on a track and inverting with zero changes on the EQ. So, yes probably denormaling, dither or similar unrelated to "summing". I have no idea why you are asking if such things can be turned off as there is literally no difference in any audible or practical imaginable way. In all honesty, until you followed up, I figured this was click bait.

I'm not sure anyone appreciates just how large -175 dBFS is (likely mentally impossible to even do so) and how this is likely much smaller than trying understand the difference between a gnat flying 10 feet away while standing beside the space shuttle when it's launching...



I don't think it even deserves the question/term of "summing" as it may not even be related tbh or which DAWs are involved.

@karbomusic
Thank you.
Your insight (as well as others) posts is exactly why I went through the trouble.

I don't want to have to turn it off for practical reasons. Just wanted to identify the source and learn more. As stated, I know very little about such technical things.
With your explanation I can see very much that it could not have anything to do with summing now and will read more about denormaling.
So in this case, most likely something "unrelated" to summing added a noise to either the reaper or protools track in the stream shoot out.
Perhaps I'll need to make a follow up to my own video when I am ready but alas...

On one hand the stream shoot out was accurate for all intensive purposes stating the two sum the same.
but on the other... was not telling an accurate truth in regards to the 100% null.

Thanks again.

[karbomusic]

Originally Posted by Drummerbman

On one hand the stream shoot out was accurate for all intensive purposes stating the two sum the same.

but on the other... was not telling an accurate truth in regards to the 100% null.

Thanks again.

Something to consider there is in our universe, everything can be reduced to some imperfection, it is the literal nature of our universe. Get rid of this "anomaly", there is nothing preventing someone from then finding something else @ 2000 dB down, then 6000 dB down (proverbially).

That's why there is going to be a point where it is an accurate functional truth because we must draw a sensible line somewhere, otherwise we enter into a realm of technical absurdity.

Xeno's dichotomy paradox comes to mind.

https://en.wikipedia.org/wiki/Zeno%2...hotomy_paradox

And off-topic aside: I found this paradox before knowing it existed in my early 20s - I had gotten my first computer (Atari in 1984) and was programming a graphic that extended a line by 1/2 the remaining distance to the edge of the screen iteratively - I discovered that theoretically it would always be moving forward, always slowing down, yet never ever reaching the edge of the screen. This exercise is kind of like that.

[jerome_oneil]

Originally Posted by Drummerbman

... all intensive purposes ...

* twitch *

"...all intents and purposes..."

[Drummerbman]

Originally Posted by jerome_oneil

* twitch *

"...all intents and purposes..."

A common misspelling... lol
Thanks Mom,
I know I couldn't get away without a digression...
But there's far better ways to make me feel dumb with the subject at hand.

Originally Posted by karbomusic

Something to consider there is in our universe, everything can be reduced to some imperfection, it is the literal nature of our universe. Get rid of this "anomaly", there is nothing preventing someone from then finding something else @ 2000 dB down, then 6000 dB down (proverbially).

That's why there is going to be a point where it is an accurate functional truth because we must draw a sensible line somewhere, otherwise we enter into a realm of technical absurdity.

Xeno's dichotomy paradox comes to mind.

https://en.wikipedia.org/wiki/Zeno%2...hotomy_paradox

And off-topic aside: I found this paradox before knowing it existed in my early 20s - I had gotten my first computer (Atari in 1984) and was programming a graphic that extended a line by 1/2 the remaining distance to the edge of the screen iteratively - I discovered that theoretically it would always be moving forward, always slowing down, yet never ever reaching the edge of the screen. This exercise is kind of like that.

There and back is my journey and zen.

[serr]

So what's the issue?

An old Protools session that you want to recall to pick up where you left off and make a mix alteration? But Protools is long in the past in your studio so you want to try to recall that session in Reaper?

And now it's making you crazy because you want to just make an addition/alteration but keep all other elements exactly the same? And that "exactly the same" part is elusive with a different DAW.

You didn't save (post) mix stems did you? That would be a way to not have to deal with the parts you want to preserve.

If the lengths you'd have to go through to duplicate the session to null against the original are just absurd, keep moving forward. If that mix really has to be dusted off and tweaked... then dial it up again in Reaper. Put up the original mix for reference and get it done. Once you get it back up, get the new parts in, explore a few new tricks you have up your sleeve since the original session... everything will be fine.

Shootouts and nulling competing DAW sessions? Screw that! You upgraded to Reaper for a reason!

[Tomm]

notice how DAW programmers rarely get involved in these discussions... probably just cringe or say "oh how cute they are" and move on to more important things

[alanofoz]

Originally Posted by karbomusic

And off-topic aside: I found this paradox before knowing it existed in my early 20s - I had gotten my first computer (Atari in 1984) and was programming a graphic that extended a line by 1/2 the remaining distance to the edge of the screen iteratively - I discovered that theoretically it would always be moving forward, always slowing down, yet never ever reaching the edge of the screen. This exercise is kind of like that.

When I was in high school (nearly 60 years ago) a maths teacher used an example where he said that if you lined up all the girls on one side of the room and all the boys on the other side and kept halving the distance between them, they'd never actually meet, but they'd get close enough for all practical purposes. That created a sudden improvement in my calculus skills!

[chip mcdonald]

If they're exactly the same you could just bit compare the files without loading them in your DAW.

I'd guess it's dither, or something like a floating difference where PT or Reaper starts with 1 instead of zero, or a LSB difference, what bit a file starts at. A the boundaries of the sample maybe a bit gets tossed on one and not the other.

I speculate that nonsense because let's say that's the case, the result from both systems would be indistinguishable by a human: because in both instances both would be equally accurate in yielding the correct result, you're only introducing a variable between the two that is yielding a difference that doesn't exist alone on either system.

Because the point to the comparison is to attempt to say one is superior to the other, or one is flawed or not. One *could* be shifting a single bit off of either end of a sample and it only matters relative to the other's "starting point".


Far speculation would the difference in order of operations of 2 mono samples sequentially versus reaper's stereo file handling. The original video's procedure should have simply avoided that difference.

[azslow3]

I have too slow inet connection at home, so I can not have a look at files (now, when I understand where they are

IMHO. Theoretically 24bit without dither should give -144dB difference and with dither a bit more. Since dither influence the last bit (is that always true?), it should be no more then -138dB. In case the difference is worse, that is something else.
BTW mixing settings in both DAWs should be checked, if one mix in 32bit and another in 64bit, the result will not be the same.