Another thread about summing in scope

[Mike Goodwin]

OK so I bet this post is not going to make me very popular around here but I feel that it is important.

So I wanted to try and find out if there was indeed a different result from summing in the scope environment as opposed to summing with Ableton Live my native summing buss. To cut to the point I did a very straight forward phase cancellation test session and the result shows that there is absolute no difference at all. The test is very easy to reproduce so before you flame me or get upset please just try it yourself. It can be done in about 10 min in Ableton if you know your way around. And if you do not use Ableton Live I'm sure it can't be that hard in other host programs as well.

This is what I did. I loaded up seven stereo tracks with 24 bit wave files. (all warping is turned off) Sent each channel to a different channel on a Dynamic mixer in scope. I then created an audio channel in live with the input set to receive the summed signal from the Dynamic mixer. I then recorded the summed signal.

The next step was to get a Ableton Live summed signal with the exactly same sample offset. What I did was send each channel to the master buss in live. Then simply send the master buss in live to one stereo channel on the scope Dynamic mixer. Then as before, re-record the summed signal into a different audio track in Ableton Live.

So at this point I have two recordings with the exactly same sample offset. One summed with Lives summing buss and one summed with the Scope Dynamic Mixer summing buss. I then phase inverted one of the summed signals and the result it complete silence. This leads me to believe that there is no benefit to summing in scope. This test has nothing to do with my "feelings". Please let it be know thatI was hoping that there would be a difference. But from what I can tell there is none. As I said please try it out for yourself and see if you come up with a different result. I have done this same simple test with between Pro Tools LE and Live as well and found the exactly same result.

If there is enough interest about this I am more than willing to share the Live session with anyone that cares to see it for themselves.

All the best

Mike

[Jah Servant]

to really test this you need to not run your live mix through scope at all, you'd have to manually shift the files though to get them to line up. Just shift 1 sample at a time with one inverted to check for cancellation. Whether there's a difference or not doesn't matter much to me, I like the sound of scope when it comes to the eq's and compressors, I don't really care if it sounds the same with the tracks running straight through or not. As a matter of fact, ideally, the sound should be the same when there is no processing being done, otherwise scope is always coloring our sound, that's just my opinion. It should be clean and only being changed when we process it, as such, some people like vinco and some don't because it colors the sound somewhat. If this was the case some would complain that the mixer's are coloring the sound, which I haven't heard as yet.

[nightscope]

This leads me to believe that there is no benefit to summing in scope

Well, if all you're ever going to do is stream a bunch of tracks with no processing then your tests have, indeed, shown that Scope sums those tracks accurately and consistently. Which is great. That's all I'd ever ask a summing environment to do. Provide totally accurate reproduction of recorded material. I personally don't want an environment to sprinkle any fairy dust over my mixes until I decide it shall happen. There's a whole kit and caboodle of fairy dust at hand for me in the Scope environment that is unavailable in any native host.

Now, if you want to process those tracks that's another matter. Obviously, this is where personal taste and a sense of sonic aesthetics come into play. There is summing, one thing, and then there's mixing using whatever tools are available and are preferable to the user, another.

ns

[Mike Goodwin]

I think that you guys are missing my point. Many people feel that there is a more accurate, more musical result from summing in Scope due to the way it is crunching numbers. I am not saying that the mixers in Scope should colour the mix.

To: Jah Servant
I am not sure how I would avoid going through the scope enviroment and have a sample accurate recording to work with. As far as can see off the top of my head it would open up way to much room for debate, converter and so on. Anyone have any ideas how how one could do this?

[valis]

I think he's saying: Sum internally at Live's master bus via recording back from there to an internal track or do an offline bounce. Since you'll probably be generating 24bit renders in Live from a 32bit floating point summing bus, and 32bit int summing in Scope there has been some discussion about that as well over the years. Make sure you've phase-locked dynamixer as well when doing the scope summing.

In any case nightscope is correct, in theory the only differences in digital summing engines should be via pan law and differences in bit depth in the main 'sum'. There just isn't any other math to perform. Most of the comments about improved sound revolve around 'mixing' in Scope's synths, effects and processing tools. There are some native tools now that compare quite favorably as well (Sonalksis & Sonnox for me) but that doesn't discount what Scope offers.

There was also some thought as to how slight phase shifting between dsps may affect mixes done in Scope, and how that innacuracy may actually wind up being perceived as having more 'depth' (same with midi data timing variations). How you feel about randomly introduced variables in your mixing is a matter of opinion though, this isn't 'Science' hence the endless discussion and poring over details.

[Mike Goodwin]

I think he's saying: Sum internally at Live's master bus via recording back from there to an internal track or do an offline bounce. Since you'll probably be generating 24bit renders in Live from a 32bit floating point summing bus, and 32bit int summing in Scope there has been some discussion about that as well over the years. Make sure you've phase-locked dynamixer as well when doing the scope summing.

Live works in 32bit audio recording and 64bit internal summing. As for doing a bounce I will give that a try and see if I can get the samples to line up.

[garyb]

use whatever works for you.

i trust my ears and my clients ears. when all i've done is seperate tracks into Scope with no changes of level, people say "what have you done? wow! it sounds better already!". this may be from a number of factors that fall into the concept of "summing". i would believe that Live sounds great and that raw tracks might not show ANY difference, but i also believe(because of experience) that complex mixes using a number of eqs and compressors sounds significantly better in Scope. since the exact same processing is not available in both sequencer and Scope, there is NO POSSIBLE truly scientific experiment to PROVE whether or why Scope sounds better. too bad.

ultimately, your ears and mixing sensibilities will have to be your guide.

[nightscope]

the chosen ones...

There are only ONE CHOSEN ONES (RIP JOE)

You could stream this lot through a sewer and it'll still come up fresh as daisies. OOOHYAH!!

http://uk.youtube.com/watch?v=16u0wwCfo ... re=related

Having a Clash Day.

ns

[siriusbliss]

All hosts and Scope SHOULD/WILL sum to themselves by design, so this is a somewhat old school test of an otherwise redundant process IMO.

What you need to 'hear' is TRANSPARENCY in the EQ staging, and how well various effects/compression, etc. get transferred all the way down the line.

We wouldn't tolerate our hosts coloring the sound at the back end of the process, and we wouldn't tolerate Scope coloring the sound either.

So, I say it comes down to transparency.

Greg

[Mike Goodwin]

there is NO POSSIBLE truly scientific experiment to PROVE whether or why Scope sounds better. too bad.

Well here we will just have to agree to dissagree. This is math.

[garyb]

when you can only truly compare finished mixes and when those mixes don't have the same processing, there's no scientific study possible, only subjective opinion. what do you like?

[Mike Goodwin]

All hosts and Scope SHOULD/WILL sum to themselves by design, so this is a somewhat old school test of an otherwise redundant process IMO.

What you need to 'hear' is TRANSPARENCY in the EQ staging, and how well various effects/compression, etc. get transferred all the way down the line.

We wouldn't tolerate our hosts coloring the sound at the back end of the process, and we wouldn't tolerate Scope coloring the sound either.

So, I say it comes down to transparency.

Greg

I am not sure what you are getting at when you say "sum to themselves". Could you lay that one out for me? Yes I would have to agree when it comes to transparency.

[Mike Goodwin]

when you can only truly compare finished mixes and when those mixes don't have the same processing, there's no scientific study possible, only subjective opinion. what do you like?

I am not talking about the processing in the Scope domain. Even you said that yourself and your clients can hear a difference simply by summing though scope. "when all i've done is seperate tracks into Scope with no changes of level, people say "what have you done? wow! it sounds better already!". It is this that I am talking about. I am saying that there are indeed people that feel that to be true. And I am saying that phase testing can prove this right or wrong. That is all I am getting at.

[valis]

Live works in 32bit audio recording and 64bit internal summing. As for doing a bounce I will give that a try and see if I can get the samples to line up.

Live works at 64bit summing at individual summing points, the processing in Live still occurs at 32bits. When summing across multiple points (or using effects) there is a small amount of degredation as the 64bit values are truncated to 32bits, but this isn't anything I'd stay away worrying about.

You're correct though that you can enable 24bit or 32bit files to record to in Live. What you can't do is monitor 32bit INT via ASIO, it will be truncated to 24bits, which again isn't really a huge concern. However in sending data to Scope to determine if it's truly 'bit transparent' this will affect results, even if the effect is negligable still. Differences will be larger if you start changing gains...

Debating summing math has been done extensively between sequencers already, any differences you detect (barring gain changes or pan-law differences) in Scope are largely going to come down to truncation at various interfaces, such as the ASIO drivers again (which only has 16bit INT, 24bit INt and 32bit which I believe is FLT *not* 32bit INT). Again the differences will come from 'mixing' using effects, eq, gain changes and so on. There's no way to mathematically correllate these differences, even accounting for gain changes (getting faders to output exactly the same values) would be a tedious affair.

There's actually a full section in the Live manual under Chapter 26 - the "Live 7 Audio Fact Sheet" - where they break down how to minimize in using Live's capabilities, which it seems you've read. Perhaps you're asking for something similar for Scope?

[siriusbliss]

All hosts and Scope SHOULD/WILL sum to themselves by design, so this is a somewhat old school test of an otherwise redundant process IMO.

What you need to 'hear' is TRANSPARENCY in the EQ staging, and how well various effects/compression, etc. get transferred all the way down the line.

We wouldn't tolerate our hosts coloring the sound at the back end of the process, and we wouldn't tolerate Scope coloring the sound either.

So, I say it comes down to transparency.

Greg

I am not sure what you are getting at when you say "sum to themselves". Could you lay that one out for me? Yes I would have to agree when it comes to transparency.

If I were to do the same test using Samplitude, I would also get phase-cancelled summing null because your source and destination are the same - through Scope. Therefore Scope isn't really effecting anything, which is good.

The other test would be to do a sum test between two different hosts to see what one host captures when recording, and what the other host sees after processing through Scope. This also should sum, but may reveal subtle differences in gain structure/EQ, etc.

Nowadays it shouldn't matter, since most hosts SHOULD sum.

So it comes down to how transparent Scope is to be able to pass data through it's environment.

Sometimes sounding 'better' just means that you're 'hearing' things in a better room, better monitors, etc. that were not apparent before. I don't know - it all depends.

I HAVE had the experience of people sending me files recorded in some other host, and I bring them into Scope/Samplitude, and literally just burn a CD, and the client says it sounds 'better' - so who knows?

Greg

[Mike Goodwin]

Live works in 32bit audio recording and 64bit internal summing. As for doing a bounce I will give that a try and see if I can get the samples to line up.

Live works at 64bit summing at individual summing points, the processing in Live still occurs at 32bits. When summing across multiple points (or using effects) there is a small amount of degredation as the 64bit values are truncated to 32bits, but this isn't anything I'd stay away worrying about.

You're correct though that you can enable 24bit or 32bit files to record to in Live. What you can't do is monitor 32bit INT via ASIO, it will be truncated to 24bits, which again isn't really a huge concern. However in sending data to Scope to determine if it's truly 'bit transparent' this will affect results, even if the effect is negligable still. Differences will be larger if you start changing gains...

Debating summing math has been done extensively between sequencers already, any differences you detect (barring gain changes or pan-law differences) in Scope are largely going to come down to truncation at various interfaces, such as the ASIO drivers again (which only has 16bit INT, 24bit INt and 32bit which I believe is FLT *not* 32bit INT). Again the differences will come from 'mixing' using effects, eq, gain changes and so on. There's no way to mathematically correllate these differences, even accounting for gain changes (getting faders to output exactly the same values) would be a tedious affair.

There's actually a full section in the Live manual under Chapter 26 - the "Live 7 Audio Fact Sheet" - where they break down how to minimize in using Live's capabilities, which it seems you've read. Perhaps you're asking for something similar for Scope?

Well to be honest it seems to me that you understand this topic better than I do. I can see what you are getting at though. It is not though summing where one might find the difference. This being said I am not sure that the software that I own in the Scope environment stacks up well against the software that I own in the native world. I am not saying that this is the case against all the software in the Scope world as I have only started to look around. It also makes little sense for me to start automating the scope mixer when automation within my host is so simple and well integrated.
When you mix though Scope do you do all your volume and pan automations with the scope mixers in stead of your native host? If so do you find that you gain depth in your mixes?

Thanks for your well informed feedback.

[garyb]

Mike, i understand what you're getting at, but it's not going to be proved conclusively. either you think the card helps you or not. if not, sell it. personally, i have too many successes to doubt.

[Mike Goodwin]

Mike, i understand what you're getting at, but it's not going to be proved conclusively. either you think the card helps you or not. if not, sell it. personally, i have too many successes to doubt.

Garyb
Hey don't get me wrong I like my Scope platform just fine. It enables me to do things that would otherwise be a huge pain in the ass. I like it for routing and my main use for it is my trusty Minimax and Pro12 as well as several other synths. I am an electronic music producer and not a guy recording bands. I was just trying to get to the bottom of the "hocus pocus" surrounding the "it just sounds better" statements that to be completely honest piss me off after a while. Even my very modest scope setup runs about $1950 after taxes here in Canada. That is about the same price as a used Yamaha 01V96VCM. One might hope that mixes that go though the scope summing engine would sound unarguably better. There is much more to the system though and the value of that can't be debated as it is the only thing that I have ever heard of that can do what it can. I do not plan on giving it up any time soon. I also look forward to what 5 has to offer when ever it comes out.

[garyb]

sure Mike!

i never meant to imply that you were unhappy or a malcontent.

i'm just saying that it's difficult to impossible to prove that Scope sounds better, scientifically. i will say this, ALL my clients report improved sound when they get a Scope card. ALL of my studio clients think that i have the best sound out of the home studios that they have been to, so there is some anecdotal evidence anyway.

i always call the difference "summing" for the benefit of clients. probably, "summing" alone isn't the difference, never the less, my best sounding mixes are always from the Scope environment and for this, i MUST trust my ears...

[Mike Goodwin]

What software are you using? I have the synth and sampler pack and that is it. I do not have any add on mixing tools. I might conceder buying some down the road. All this being said It is hard to know how these things stack up against the top shelf native stuff these days. Waves, PSP, Sonalksis and so on. It is up to me to listen and decided for myself in the end.