The difference of audio engines

[digitalmessiah]

this topic pops up every now and then. especially back when people were using Reason 2.5 and comparing it with other hosts...that v2.5 sounds 'inferior'.

btw, of all the hosts that i've used/tried, Orion Plat sounds very punchy and loud. i don't know what Rich did to that thing but it sounds bloody loud.

[aallvor]

this topic pops up every now and then. especially back when people were using Reason 2.5 and comparing it with other hosts...that v2.5 sounds 'inferior'.

btw, of all the hosts that i've used/tried, Orion Plat sounds very punchy and loud. i don't know what Rich did to that thing but it sounds bloody loud.

Yeah, I think loudness is the main difference between apps, with one exception; Live. Something happens to the bottom end when 5/6+ tracks are being played back at the same time. Problem is (for Live owners), that Ableton is, and always have been, completely silent on this subject, as if all the complaints on the forum is just crazy talk..

[anselmoso]

Although i find it a bit funny that he achieves better result in 2 different flavours of Cubase.
Are'nt the workflow pretty much the same?
I only have experience with SX but i find it hard to believe that there is so much diffrence between SE and SL.

that's not actually true. In my experience there's also a "difference in the result" between SE1 (which I own) ans SL3.
Could be explained by the new audio-engine that came up with SX2 - maybe "better workflow" - I don't know....

I still think: Steinberg has been working on music applications for a very long time now - so their audio-engine is of course different from other applications in terms of the programmed code. There must be a difference in the way the audio-engine is programmed. The don't all use "the same C++ code" for their audio-engines....do they?

Anyway - I was just talking about my own experience - and my "best results" were made with Cubase SL3 until today - which kind of sucks, because I have a heart for the "little players" like Ext and Tracktion - but for some reason (although I've worked with Tracktion and Ext for several weeks and months) I always get good results in Cubase in a shorter work period.

Right now - to me - Steinberg Cubase is like the epic fight between Windows and Linux. I love Linux. I use it every day at work. But for some reason, even though I've also used it several years at home - today I'm using Windows full time on my PC at home. Even though I was a "Microsoft and Bill Gates hater" and I claim to "hate all monopols" - somehow it just works for me, I payed 80 Euros and have a working system and good applications - it's just "easier" to work with.

I think my "thing with Cubase" can be explained in a several way. I "hate the dongle" - I "hate the distance to the user" (Steinberg not caring for each single user like it seems here in the Ext forum for example) - but somehow this thing just "seems to work" for me.......even though I kind of hate it:)

Like a woman;)

[stefancrs]

You might percieve a difference, but in all tests we've done so far (see that other thread) we've found that there is no difference.

[arke]

Audio engine:

Code: Select all

AudioStream result;
foreach(AudioStream a in mixertracks)
{
  result += a;
}

Its not any different, except maybe programming language but that ends up executing the same machine level code anyway.

[Ozpeter]

To test one program against another, take a file, apply one or more VST effects to it using specific presets, make sure that you are not messing with levels at all (have regard to panning laws for instance) and having done that with each system, invert the result from one against the result from the other. If they are different, the result of the inversion will be the difference. If the same, the result will be silence.

Biggest difference between the various systems will lie in their built in effects. But then you cannot test other than subjectively.

[stefancrs]

Only works with time-invariant VST effects btw, so no effects that includes any kind of oscillator for instance, like a chorus or a phaser for instance.

[aallvor]

Excuse my ignorance, but for the sound to be identical between hosts, wouldn't the summing algoritms have to be identical

[tony tony chopper]

For one, the colour of Sonor's splash screen is dominantly green. Cubase's splash screen is dominantly white / grey. It is not the same. Even the background colours are different.

Use FL Studio. We carefully chose our UI color as charcoal, because it's what gives the best sound (and this was proven by scientists in white suits). Our splash screen also features yellow, orange, and blue, to color the sound equally.

Excuse my ignorance, but for the sound to be identical between hosts, wouldn't the summing algoritms have to be identical

Again, use FL Studio. While other sequencers use popular algo's such as 'Input=Output+Rand(3)' or 'Input=Output+AgeOfTheCaptain', we *patented*, after years of research, the 'Input=Output' algo. Our grandfathers have been in the audio summing business since 1933.

[LIMITAPROACHINGINFINITY]

Excuse my ignorance, but for the sound to be identical between hosts, wouldn't the summing algoritms have to be identical

No.. Just the results need to be identicle. Its like this

algorithm 1:
result = 1 + 1 + 1 -1

algo 2:
result = 2


The algorithm was not the same.. But can YOU tell the difference between the results?

[LIMITAPROACHINGINFINITY]

Again, use FL Studio. While other sequencers use popular algo's such as 'Input=Output+Rand(3)' or 'Input=Output+AgeOfTheCaptain', we *patented*, after years of research, the 'Input=Output' algo. Our grandfathers have been in the audio summing business since 1933.

ROFL!

I mean.. when you think about it.. What are they gonna do? Add extra signal for no reason!

[tony tony chopper]

More seriously, people don't want to know. They already know, and want to be confirmed. For them, an expensive & old app obviously has a better audio engine, because it can't be otherwise.

It all depends what you call 'audio engine', that's all. If for you the 'audio engine' is what takes the input to your soundcard, then all hosts are the same. Your VSTi *will* *sound* *the* *same* in all hosts (and yes you'll find people telling you otherwise). Only difference will be the output volume, but there's no loss (only a theorical one), only volume change. You (as a normal human) may not hear the volume difference as a flat volume difference, but tweak a knob, and it'll be the same.
This is because all of today's sequencers (except pro tools that still mixes in integer) mix audio in at least 32bit floating point.

Now if you call 'audio engine' the tools built-in the sequencer, like sampler, audio clip resampling, equalizing.. there ARE obvious differences in quality out there. Resampling algo's can give extremely different results, little things like volume ramping are implemented differently. But you can still use plugins to replace most of these.

And if you call 'audio engine' the UI color & splash screen, there's no hope for you.

[camsr]

Maybe one of the hosts is doing an unnoticed PDC? The plugin delay will SURELY change the sound.

[stefancrs]

Yes, but then you're comparing a feature vs the lack of the same feature. Comparing features between hosts is one thing, and something everyone should do when they decide what host to use. Comparing the hosts actual audio quality however I find quite unnecessary, since they all sound the same (apart from when using specific features not supported by the host you compare with or vice versa).

[davor]

Can any of you guys confirm that this experience could be influenced by the different audio-engines of these sequencers?

It could be.