Audio Engine?

[Spe3D]

[chagzuki]

Like I've said before, Fruityloops sound fruity, Orion cosmic, Logic logical. . . can't think of any more.

[cold c]

AS for all this, when comparing the MIXING engine, differences are greater than summing. Id love to do a mixing engine test, that would be something interesting

In a previous thread you claimed to already have done extensive research on the subject. Can you just put the research somewhere online so we can view it?

2. Summing engine (alot of recent research ive done)
3. Mix engine (alot of recent research ive done)

Quoted from here http://www.kvr-vst.com/forum/viewtopic. ... 315#318315

[Robert Randolph]

Whoever said it always made an audible difference? Summing engine alone that is... it's the accumulation of errors between mixing engine, summing, and effects that causes the problems. Eliminate or lessen one problem and then focus on the others.

I dont recall anyone saying a better summing bus will make better songs, or recordings or even for that matter that summing is something everyone on earth can hear.

A lot of assumptions being made here, seems like eveyrone is agreeing with me but still trying to find a reason to argue.

-R bert

[Robert Randolph]

AS for all this, when comparing the MIXING engine, differences are greater than summing. Id love to do a mixing engine test, that would be something interesting

In a previous thread you claimed to already have done extensive research on the subject. Can you just put the research somewhere online so we can view it?

2. Summing engine (alot of recent research ive done)
3. Mix engine (alot of recent research ive done)

Quoted from here http://www.kvr-vst.com/forum/viewtopic. ... 315#318315

yes I have, it was not documented formally, but discussed on other forums (n-track board, nuendo board) where no one doubted it after doing the tests. And I first discovered a "quick" way to show it on the tracktion board... tracktion doesnt the rounding thing a bit weird too.

like I keep saying, it's an obious source of errors (noise) but hardly the end all of everything. Alreayd gol and james who seemed to disregard it have admitted to the existance of overflow in 32-bit accumulators... good enough for now. Im working on a "test" for the mixing engine of various hosts... spent about an hour tonight researching my boundaries (file types, fader ranges, included effects etc..) so the test can be performed equally on all platforms and hosts.

[edit] on the nuendo board there's a big thing about how you "cant hear it". That is not the intention, it is the simple acknowledgement that it produces errors which can be accumulated with other factors. For larger projects it can make a significant difference, but i doubt most people here care in the slightest.

-R bert

[pough]

but i doubt most people here care in the slightest.

I care, Robert. <reaches out to pat Robert's hand encouragingly>

[Spe3D]

I think René’s SFZ player looks pretty cool in the tests I have seen – but I would be interested in seeing some tests of hosts rendering at 32-bit float – not too interested in the 16-bit as dither can be so influential to the outcome in some apps if used.

Regards,

Spe3d

:O)

[tony tony chopper]

Alreayd gol and james who seemed to disregard it have admitted to the existance of overflow in 32-bit accumulators

overflow? you wouldn't even know what it means, & no one told about overflow anyway

[Robert Randolph]

Alreayd gol and james who seemed to disregard it have admitted to the existance of overflow in 32-bit accumulators

overflow? you wouldn't even know what it means, & no one told about overflow anyway

Uhhh. not overflow as it would be determined as a runtime error... rounding error. If it were overflowing as such as commonly refered to there'd be a whole 'nother issue with stability not sound
pardon me!

-R bert

[TeeLangSun]

I really want to hear an actual difference in a blind test with several hosts using identical settings.

Seriously, this is the only way to ever get any difinitive answers on this topic. Same settings, same files, same monitors, same listening environment. Otherwise this is just people cheerleading their favorite host.

And even then, how much does it really matter? I've yet to hear any software which sounded so badly that I wouldn't use it. That includes Ejay Ibiza! It sounds great. I don't have a golden ear, but I'd bet that my ear is a little more focused than the average music consumer. A great deal can be wrong with a song and people who don't make music won't notice it no matter how many times they listen to the same song. They will never notice it, even when the singer is off by a mile. In other words, if these hosts do sound different, I don't think it matters. They all sound good enough to make a hit record and people are making excellent quality music on all of them.

[Mr. Slater's Parrot]

Right on the money, TeeLangSun.

Also, isn't it kind of ironic how this sort of discussion takes place on a forum where there is such an affinity for "mangling and destroying beats and sounds"?

Heck, some of us have even been known to put iZotope's Vinyl on a track or two. Speaking just for myself, I like to make sure my hiss, pops, and crackles come through with a lot of clarity and airiness on the high end -- with plenty of punch on the bottom too, of course.

[Spe3D]

Right on the money, TeeLangSun.

Also, isn't it kind of ironic how this sort of discussion takes place on a forum where there is such an affinity for "mangling and destroying beats and sounds"?

Heck, some of us have even been known to put iZotope's Vinyl on a track or two. Speaking just for myself, I like to make sure my hiss, pops, and crackles come through with a lot of clarity and airiness on the high end -- with plenty of punch on the bottom too, of course.

Starting quality is very important – even more so when you enter into extreme sound alteration, wise use of sample rates and bit depths play an important part in controlling unwanted artefacts, and various other horrors.

Regards,

Spe3d

[Mr. Slater's Parrot]

Starting quality is very important – even more so when you enter into extreme sound alteration, wise use of sample rates and bit depths play an important part in controlling unwanted artefacts, and various other horrors.

And my point is that 32-bit floating point provides more than enough "starting quality" for tracks that have plugins that drop the bit depth down to 12, for example -- no disagreement there, right?

[Spe3D]

32-bit float is more than adequate for me.

Regards,

Spe3d

[Mr. Slater's Parrot]

32-bit float is more than adequate for me.

It is for me, too (at least for now, anyway ).

So for me, that narrows down the discussion as follows. If one believes or assumes (or simply accepts for the sake of argument) that 32-bit floating point has sufficient precision for the purposes at hand -- IOW, such that it isn't the bottleneck or weak link in the system -- and assuming that various hosts use 32-bit floating point for the internal representation of sample values, then are there differences in how the various hosts produce the resultant 32-bit values for the mix/summation/output (or whatever you want to call it)?

Leaving things in the 32-bit realm eliminates the perfectly valid variable of dithering algorithms.

So what else is there that might be different?

There could be bugs in the software, I suppose.

There could be behind-the-scenes, builtin effects that change the sample values, perhaps.

There could be differences in default levels that are set.

There could be differences in the approach taken to the handling of panning.

Anything else???

If host A sounds better than host B because host B has bugs and host A doesn't, then we are on to something. If so, someone should show the bugs!

If host A sounds better than host B because host A is applying an effect and host B isn't -- or vice versa -- then that changes the playing field. And it would be nice to know what these effects are and which hosts are or are not using them -- presumably, behind the scenes. I'm not aware of any host that does such behind-the-scenes effects processing. I think most users wouldn't want such a thing, in any case.

If host A sounds better than host B because of differences in default levels or the handling of panning, that doesn't seem like such a deal-breaker since one can readily set the levels and panning to taste in a host (certainly for all hosts that I've used).

So, what else is there? Maybe I've missed something?

OTOH, if you say that host A *sounds* better than host B because the "workflow" and "look and feel" of host A are more conducive to making good-sounding music than are those of host B -- well, that's just a completely different discussion. Mixing (so to speak ) the latter in with talk about "audio engines" is just an invitation for disagreement, IMHO.