Hi all,
mixing with Pulsar mixer with different BUSSes or with VST mixer (ASIO BUS 1 only) what's the difference in term of quality (probabily a stupid question!)?
An other question : It's better FFT EQ or not ? Sorry for bad english.
Bye
Pulsar mixer VS Qbase VST mixer....
Hi DP,
To my ears, mixing in SFP sounds better than mixing in Cubase SX. Especially the stereo image of the mix sounds more open. Because of this, I find it easier to get a mix where all the sounds have their own place.
But you should really try this yourself! I would advise to use at least 8 stereo tracks (the more tracks, the more the difference becomes evident). Don't use any plug-ins, just put some audio tracks in cubase and assign them to the same bus, save your project, assign each track to a different bus and save your project with another name. Now open both the projects and listen if you can hear any difference between them.
FFT EQ's are not necessarily better than other digital EQ's. They're mostly more flexible than traditional EQ's, but sound a bit clinical. Probably because they have no 'sound' on their own. They're also quite CPU hungry and are not very 'real-time' because FFT transformations are buffering audio, which means that it will add some latency to your signal.
I think FFT EQ's are most usefull for sound design and applications that need a 'surgical' EQ with many bands.
cheers,
vincent
To my ears, mixing in SFP sounds better than mixing in Cubase SX. Especially the stereo image of the mix sounds more open. Because of this, I find it easier to get a mix where all the sounds have their own place.
But you should really try this yourself! I would advise to use at least 8 stereo tracks (the more tracks, the more the difference becomes evident). Don't use any plug-ins, just put some audio tracks in cubase and assign them to the same bus, save your project, assign each track to a different bus and save your project with another name. Now open both the projects and listen if you can hear any difference between them.
FFT EQ's are not necessarily better than other digital EQ's. They're mostly more flexible than traditional EQ's, but sound a bit clinical. Probably because they have no 'sound' on their own. They're also quite CPU hungry and are not very 'real-time' because FFT transformations are buffering audio, which means that it will add some latency to your signal.
I think FFT EQ's are most usefull for sound design and applications that need a 'surgical' EQ with many bands.
cheers,
vincent
...and thank you for bringing this up 
, because I'm curious how many people over here are using SFP mixers because they think it sounds better.
I like to lurk on the Nuendo forums and from time to time there are very long and heated discussions about digital summing bus differences. One side thinks that it's just all math and that mathematics can't be fooled: 1 + 1 is always 2. So the difference must be in the head of people. And there's the other side that can't explain why there's a difference, but they think that there definately *is* a difference between different applications. I think I'm a believer of the latter
What do you all here at Z think?
cheers,
vincent
, because I'm curious how many people over here are using SFP mixers because they think it sounds better.I like to lurk on the Nuendo forums and from time to time there are very long and heated discussions about digital summing bus differences. One side thinks that it's just all math and that mathematics can't be fooled: 1 + 1 is always 2. So the difference must be in the head of people. And there's the other side that can't explain why there's a difference, but they think that there definately *is* a difference between different applications. I think I'm a believer of the latter
What do you all here at Z think?
cheers,
vincent
hi Vincent,
it is all math and all math IS fooled.
1+1=2 is only true in integer systems, but a DAW isn't integer at all, so the precision of math it what makes the difference.
Imho one of the great advantages of our beloved Sharc DSPs is that they use a specialized and dedicated math library.
That's why they are called signal processors, opposed to native CPU math which is a general purpose thing.
Of course this doesn't exclude someone writes some smart code also dedicated to this kind of processing for Pentium/AMD/PPC floating point units. But the art of assembler coding on this level isn't very widespread today.
If you have a good test setup I'd suggest the following 'song' for mixdown:
4 stereo tracks with exactly the same pad sound slightly reverbed
4 stereo tracks with the same sharp brass attack.
Any irregularities should be clearly noticable by phasing and distortion.
At least I'd guess the SFP/Cubase result will sound significantly different.
cheers, Tom
it is all math and all math IS fooled.
1+1=2 is only true in integer systems, but a DAW isn't integer at all, so the precision of math it what makes the difference.
Imho one of the great advantages of our beloved Sharc DSPs is that they use a specialized and dedicated math library.
That's why they are called signal processors, opposed to native CPU math which is a general purpose thing.
Of course this doesn't exclude someone writes some smart code also dedicated to this kind of processing for Pentium/AMD/PPC floating point units. But the art of assembler coding on this level isn't very widespread today.
If you have a good test setup I'd suggest the following 'song' for mixdown:
4 stereo tracks with exactly the same pad sound slightly reverbed
4 stereo tracks with the same sharp brass attack.
Any irregularities should be clearly noticable by phasing and distortion.
At least I'd guess the SFP/Cubase result will sound significantly different.
cheers, Tom
well i can't say anything about math processing or why one if better than the other...
I can only say that i have just finished and mastering a dance track using for the first time SFP mixer and Pulsar's plugins instead of Cubase mixer and many dx plugins (also good plugs like Waves etc..) and i can say that it sound 10 times better than using only Cubase..
For the FIRST time i have made something which sounds like a professional product: great kick, great bass, great dynamic...
So i can only say: i can't tell you why, but i can let you hear the difference...
Gab
I can only say that i have just finished and mastering a dance track using for the first time SFP mixer and Pulsar's plugins instead of Cubase mixer and many dx plugins (also good plugs like Waves etc..) and i can say that it sound 10 times better than using only Cubase..
For the FIRST time i have made something which sounds like a professional product: great kick, great bass, great dynamic...
So i can only say: i can't tell you why, but i can let you hear the difference...
Gab
Thanks for all your input. Nice test suggestion Tom, I'll try the test later this week.
Does anybody know for a fact that all math inside SFP is floating point? I know that the Sharc can do both float and integer math and I somehow seem to remember something about integer math being used in Pulsar. But that info is back from DuPont mailing list days and as some people might remember, that's a long time a go. So I'm not sure about this anymore.
cheers,
vincent
Does anybody know for a fact that all math inside SFP is floating point? I know that the Sharc can do both float and integer math and I somehow seem to remember something about integer math being used in Pulsar. But that info is back from DuPont mailing list days and as some people might remember, that's a long time a go. So I'm not sure about this anymore.
cheers,
vincent
-
enumeratus
- Posts: 21
- Joined: Mon May 06, 2002 4:00 pm
- Location: Rheinland/Germany
imho that's the point, DSPs have an optimized math, that fits the operation.Actually they are floating point DSPs
that are also great integer machines.
Basic precision is 32 bit, but can be made much higher easily.
Many processing algorithms are dealing with trigonometric functions which may result in very small or extremely large numbers.
I simplyfied somewhat by calling it '... not integer at all'.
At least even very small errors in these calculations can have a significant effect on sound quality since the human ear (if not hardrocked or clubbed to much) has a 120 db operation range.
cheers, tom
I also prefer to mix in pulsar, but I'm surprised noone has mentioned the problems that SFP suffers from.
Most devices, when split across more than one dsp, will cause phase problems with stereo signals. The exception to this being the mixers which now have phase compensation switches (but this also eats more dsp). I also suspect that when a plugin is loaded on a compensated mixer that plugin may cause the phase problems even tho the mixer is insuring its own internal signal path is correct.
Even after that's said & done i too prefer to mix in SFP for some reason but have learned to make use of these:
http://www.planetz.com/forums/viewtopic ... forum=16&3
Most devices, when split across more than one dsp, will cause phase problems with stereo signals. The exception to this being the mixers which now have phase compensation switches (but this also eats more dsp). I also suspect that when a plugin is loaded on a compensated mixer that plugin may cause the phase problems even tho the mixer is insuring its own internal signal path is correct.
Even after that's said & done i too prefer to mix in SFP for some reason but have learned to make use of these:
http://www.planetz.com/forums/viewtopic ... forum=16&3
Hello to all!
This topic is very interesting to me! You mention things that I have never thought of (and maybe never would think of)!
I did not do the tests you suggested but I take as granted that mixing in SFP is of better sound quality. Now, I have mainly 2 naive questions to ask (forgive my naivenesss but I am not yet a pulsar geek!):
1) How can I know if a particular module/device is spread on 2 or more DSPs, thus causing phase problems?
2) I saw the other thread that valis pointed at. At0mic says that there is a 2-5 samples latency between 2 DSPs. Is that enough for causing phasing problems?
Thank you for this topic!
This topic is very interesting to me! You mention things that I have never thought of (and maybe never would think of)!
I did not do the tests you suggested but I take as granted that mixing in SFP is of better sound quality. Now, I have mainly 2 naive questions to ask (forgive my naivenesss but I am not yet a pulsar geek!):
1) How can I know if a particular module/device is spread on 2 or more DSPs, thus causing phase problems?
2) I saw the other thread that valis pointed at. At0mic says that there is a 2-5 samples latency between 2 DSPs. Is that enough for causing phasing problems?
Thank you for this topic!
hi rodos,
best thing would be to apply that 'sample delay' plugin to 2 signals parallel on your mixer. Depending on the kind of signal you'll be surprised how much the sound is influenced by even small delays.
The human ear is very sensitive for the attack part of a signal and the influence increases with frequency.
Female vocal 's' or high brass sounds can be really unpleasant if something goes wrong (either with phase or later with aliasing on the converter).
Imho the stereo difference isn't such a big problem, but signal which are distributed over several channels may get blurred if an 'out of phase' processing occurs.
The phase fidelity is also the most important (and often neglected) property of a good amp and speaker.
cheers, Tom
best thing would be to apply that 'sample delay' plugin to 2 signals parallel on your mixer. Depending on the kind of signal you'll be surprised how much the sound is influenced by even small delays.
The human ear is very sensitive for the attack part of a signal and the influence increases with frequency.
Female vocal 's' or high brass sounds can be really unpleasant if something goes wrong (either with phase or later with aliasing on the converter).
Imho the stereo difference isn't such a big problem, but signal which are distributed over several channels may get blurred if an 'out of phase' processing occurs.
The phase fidelity is also the most important (and often neglected) property of a good amp and speaker.
cheers, Tom
Check the other thread for the methods that are described for tracking down the phase problems...On 2003-02-07 08:39, rodos1979 wrote:
Hello to all!
1) How can I know if a particular module/device is spread on 2 or more DSPs, thus causing phase problems?
2) I saw the other thread that valis pointed at. At0mic says that there is a 2-5 samples latency between 2 DSPs. Is that enough for causing phasing problems?
Thank you for this topic!
Although it might be more convenient to mix in the native sequencer environment, I still prefer to use the sequencer like I would a tape deck and nothing more (apart from automation). To my ears, keeping as much as possible in the Pulsar environment sounds better at the end of the day.On 2003-02-08 12:04, braincell wrote:
Don't the EQ and the effects sound better in Cubase?