mixing with sfp mixers

dickster · Post by **dickster** » Fri Aug 06, 2004 2:09 am

Through the last few years I have purchased alot of CW cards,ultra16,and many plugins.I am also one of the early tripledat users,but have moved on to cubase sx when CW stopped updating tripledat.Although I love the CW plugins and routing,I often think how much simpler it would be to record and mix entirely in cubase,which also has some great plugins.But I have heard conflicting comments on the quality of mixers in the more popular daws.Some say the summing is not very good,and also that virtual mixers should ideally be 48 bit. One reason being that every time you move a fader you lose bits.Someone once told me that the only virtual mixers they would mix on would be protools HD 48bit mixer or CW STM2448.They said that CW added a few extra bits to 32 for this reason.It is very time consuming to do this type of ab comparison. I am wondering if anyone knows any technical reasons why mixing in SFP will give better results,or if anyone has experimented with this.

at0m · Post by **at0m** » Fri Aug 06, 2004 7:35 am

Check out this active topic.

Music Manic · Post by **Music Manic** » Fri Aug 06, 2004 8:58 am

Think we all need to understand signal protocol.Things such as how Scope deal with maths,then how it's converted to ASIO,then how Cbase deals with math,then how Windows deals with kernel etc.I think Cubase audio engine is terrible.If Creamware updated 3dat and added midi it would be a killer.

dickster · Post by **dickster** » Fri Aug 06, 2004 5:49 pm

I am quite suprised that there has not been any mention of the new tripledat, due to be released in about 2 weeks.Both Ali and Ralf have told me about it. It will be 32 bit,XP compatible ect. Still no sequencer, but so what. I think a STM 2448 will be built in. It will only be able to handle 24 stereo tracks, for now. Personally,I cannot wait. There will also be an upgrade price for 3dat owners!

at0m · Post by **at0m** » Sat Aug 07, 2004 12:04 am

On 2004-08-06 18:49, dickster wrote:
I am quite suprised that there has not been any mention of the new tripledat, due to be released in about 2 weeks.

We didn't know of all that

Looking forward to try it out!

Immanuel · Post by **Immanuel** » Sat Aug 07, 2004 7:11 am

Samplitudes mixer doesn't make you loose any bit depth. It is Floating point.

dickster · Post by **dickster** » Sat Aug 07, 2004 5:46 pm

I am sorry if there was some misunderstanding,but I was not bitching about anything. Maybe you should read again my post. I am just trying to learn as much as I can about virtual mixers. Which method of routing will give me the best results. My suspicions were right,that I am better off mixing and summing in SFP than cubase.All of the links and posts were very helpful. Thankyou all who helped. By the way I would never complain about anything concerning CWA.I have been a creamware user since the early days of tripledat. I beleive I bought the first pulsar in the U.S.and I own 4 cards,an ultra 16,and almost every plugin.I am also a certified pulsar geek and certainly was not"bitching".This was also my first time posting on this forum.

dickster · Post by **dickster** » Sat Aug 07, 2004 6:09 pm

In rereading my original post,all of the negative things about mixing "in the box" I was refering to mostly things I had heard about steinberg products. I simply wanted to know if STM mixers had similar problems with summing and mixing. There are many who think that protools has the same problems.Like I said before I have heard that protools HD and STM mixers dont have these summing problems and loss of bits. Apparently samplitude is also very good.This is why I think some engineers prefer outboard mixers. I think we have a great alternative in STM2448.

Immanuel · Post by **Immanuel** » Sat Aug 07, 2004 10:17 pm

Sorry

You misunderstood me.

The Bitching part was not for you. It was a signature (default stuff at the end of posts), that I created after the Scope4 release, when a lot of people came here to ... Bitch. I just forgot to turn it off, when I posted me reply to you.

dickster · Post by **dickster** » Sun Aug 08, 2004 12:23 am

Immanuel,you mentioned that samplitude mixer was floating point. Are'nt CW mixers also floating point? I am not sure I have understood which is better,or what the difference is between floating point and integer.

Immanuel · Post by **Immanuel** » Sun Aug 08, 2004 12:33 am

you are right.
but in your initial post, you talked about loosing bits, when moving faders.
that does not happen, when you are in floating point.

astroman · Post by **astroman** » Sun Aug 08, 2004 5:55 am

On 2004-08-06 03:09, dickster wrote:
...Someone once told me that the only virtual mixers they would mix on would be protools HD 48bit mixer or CW STM2448.They said that CW added a few extra bits to 32 for this reason...

that's interesting

I once read a detailed description of how the bit depth in ProTools mixing is preserved.

Unfortunately the reference has left my mind, as did the exact procedure, but it all was obvious and clear to understand.

Bottom line was something like if you reduce a track in PTHD by (say) 80db and write the result to disk, still enough bits are retained that you can blow it up to full level next time
If the CWA mixer follow the same scheme, that's a very good reputation.

as I write this it just entered my mind that Samplitude (like TripleDat) uses a non-destructive approach in processing it's data .
The 'trick' written above would be easy to perform, because only the instruction is written to file. So 80 dB down, 80db up would result in 'don't touch it at all'

Btw Magix calls this 'object oriented processing', but it's nothing else than the classic EDL list used in AV processing long before Magix was even founded.

cheers, Tom

Immanuel · Post by **Immanuel** » Sun Aug 08, 2004 6:22 am

You are mixing the terms Tom.

Object oriented refers to the way you cut a file into virtual objects. It has nothing to do with the way Samplitude processes sound.

That doesn't have to say, that you re wrong in your description of the methodology. I however always thought that this was just an advantage of using Floating Point.

astroman · Post by **astroman** » Sun Aug 08, 2004 12:32 pm

hi,

you certainly cannot get from my post how upset I was when I first read that Magix blurb.

Such 'slogans' drive me nuts...
That 'Object oriented' paradigm is leant from programming, where it actually means that a data structure also contains code for processing, both contained in an entity called 'Object'.
Properly applied this allows a much more straight forward programming as you don't have to constantly redefine some interface definitions.
It is (was) important, really important, as without it NO GUI as we know would exist.

I'll leave the historic background, but from then on it stood for a certain leading edge technology.

Magix abused the term to gain a higher regard of their product, nothing else.
They use a totally simple (yet effective) method which is known for decades (usually called 'edit description list') and declare as their new invention.
I can't stand this attitude, whatever qualities the program may have...

But in fact it's exactly that approch to 'collect' processing commands and only work on temporary data until the final stage is reached that gives Samplitude a big quality advantage.

As shown in the example above it allows to significantly optimize the processing chain.

A floating point number may look incredibly precise, but it gets as inaccurate as anything else once you start operating on it.

cheers, Tom

valis · Post by **valis** » Sun Aug 08, 2004 1:58 pm

To be sure Object Oriented was a fad term during the late 80's and early 90's, but its used daily by programmers in the same way that a musician would call notes notes. To me such a generic term sounds more like it would have come out of a programmers's mouth and gone into the early production documents, and stayed that way. Its also perhaps somewhat accurate, as the 'object' not only contains the file and a nondestructive history (akin to an EDL as someone else mentioned, but managed invisibly for you) as well as a lot of other nondestructive effects and parameters:

I would agree that it didn't take a lot of imagination to call something an "Object Editor" (how about PART editor?). However I don't think its a term that was thrown out just because OOP was used to create the GUI. Also Samplitude's object editor is a bit more powerful than an EDL, even in most modern packages: http://www.samplitude.com/de/feat_c.htm

Just to clarify (though I would think its already obvious to everyone) these controls are tied directly to the object/part in the arrange and independant of any track/bus etc. controls for the mixer.

Also note that I'm not a Samplitude user (waiting on robust midi) but rather a Logic user. And I recall how much print time Logic has given to discussing its own 'objective' nature.

<font size=-1>[ This Message was edited by: valis on 2004-08-08 15:00 ]</font>

astroman · Post by **astroman** » Sun Aug 08, 2004 3:15 pm

you're right Valis - there has been an inflationary use of the term 'Object' in the past.

In fact it's no big deal (for me anymore) and refers to a memory almost as old as the dates you mention - just wanted to make an ironic side-kick

only to clarify my GUI statement, which is of course not related to the program at all:
it's a reminescence to the old XEROX PARC system Apple licenced as the foundation of their own GUI developement.

Apple then had to create ObjectPascal (no such languages existed on 'small' systems) and in that context one of the original developers mentioned: we'd never be able to succeed in an event driven user interface without the object oriented programming approach.

cheers, Tom (babbling too much..)

blazesboylan · Post by **blazesboylan** » Mon Aug 09, 2004 2:29 am

If GUI library designers ever bothered using their own libraries then we would not have any more "event driven programming" at all, and much less of the (admittedly somewhat useful) object oriented fad.

Event driven programming may as well be replaced by stochastic methods for all it's worth in any real world situation.

Expect the next big fad to be "aspect oriented programming", which is completely useless and more full of gobbledygook than the catholic realm of "object oriented" programming is today. I can't wait until some wingnut comes up with an aspect-oriented sequencer which simply separates the EDL back out from the WAV file again. Brilliant. Back to square one.

One of these days maybe some other programmers will realize the brilliance behind CreamWare. Data and good ol' state machines are all you need. That's exactly what SFP is -- a musical Turing machine. Simple and 65 years old.

Sorry for the off-topic rant.

Johann

valis · Post by **valis** » Mon Aug 09, 2004 4:25 pm

On 2004-08-08 16:15, astroman wrote:
cheers, Tom (babbling too much..)

Forums are great for babbling too much--I do it with images even!

Shayne White · Post by **Shayne White** » Fri Aug 13, 2004 10:33 pm

On 2004-08-06 18:49, dickster wrote:
I am quite suprised that there has not been any mention of the new tripledat, due to be released in about 2 weeks.Both Ali and Ralf have told me about it. It will be 32 bit,XP compatible ect. Still no sequencer, but so what. I think a STM 2448 will be built in. It will only be able to handle 24 stereo tracks, for now. Personally,I cannot wait. There will also be an upgrade price for 3dat owners!

3Dat?? I thought Cutmaster was supposed to be the upgrade to 3Dat?

Post by **garyb** » Sat Aug 14, 2004 9:15 pm

all coming together very soon....