Different hosts produces different sound ?

[kara]

At the risk of sounding stupid I have a question or rather a weird experience

I'm a long time user of Quartz Audio Studio Professional and recently switched to Tracktion.

Now I have the strange experience that when I render the same track to WAV with the 2 hosts that they don't sound the same. To my ears the Traction sound is 'colder'.

Is this possible or should I take an appointment by the ear doctor
Would that mean that some hosts target a specific type music?

Rony

[ceenda]

I had this happen recently as I was playing around in FLStudio and Orion Platinum. When just playing around in Orion, the sound quality seemed alot better. Clearer, crisper somehow.

Anyway, I realised that it's probably the volume that's doing this. When you look at the Mixer for FL Studio, the values are already near the top, and also inserted Generators don't start off outputting at full value. I had to route Generators to one of the Send Channels to get the same 'oomph'.

I've had both running at the same time, doing identical compositions (same tempo, soundfonts etc.) and doing the "eyes-closed test" and I can't prove that they sound different. (but then, I can't render to WAV to analyse)

There _may_ be a substantial difference in the way that the internal Samplers and Soundfont players work between hosts. Just using the same SoundFont or WAV file between 2 hosts doesn't necessarily mean they'll be reproduced the same, or go through the same level of processing when they are finally dumped to WAV. I know alot of packages have different ways of interpolating when they render out tracks.

But I'm curious to see this 'myth' dispelled, or clarified. As sometimes I sense the sound quality is different between hosts, but I can't prove it, so it must be bollox.

[wrench45us]

i understood this was a myth and have read threads saying it just wasn't so
but then after using Project 5 for a year or so, I recently picked up Sonar3PE, and even within two Cakewalk products I hear a substantial differnece.

some of it seems to be volume differnces -- a given mix moved over into Sonar as Rewire doesn't have the same relative volume levels. I don't know why.

but working with Absynth a lot recently and some Brian Clevinger Absynth patches esp. invertebrate rhythms render very different in Sonar than they do in Project 5. Sonar has a lot more 'emphasis' for lack of a better word on mid/upper ranges than P-5.
Th eper track eq's are in neutral, so I don't think it's that. Most other instruments come across with very little difference in overall timbre, (maybe Sonar sounds a touch brighter) but for some reason Absynth is quite different.

That's what my ears tell me anyway.

[ceenda]

Whoops... I owe FL Studio an apology.

I selected the wrong output in the "Audio OUT" settings. It was using another driver and not my ASIO one. The volume is now the same as the other host and I don't have to route it through a send channel any more.

But I'd still like to know what the score is with the sound quality, in particular Sampler quality and Render quality between various hosts.

[saulc12]

Different hosts will process audio differently internally and the sounds of the vsts will have passed through the hosts audio processing at mix down if no where else, because of this most plugins will sound very slightly different from host to host - some will sound very different, this is also the case with different audio drivers, so you may find that a vsi played through the same host on different machines, but with identical sound cards and monitoring might sound different if say one machine was playing back using asio drivers and the other was using direct x. Again this is entirely down to how the data is processed in the mix down of the audio streams.

This is one reason why anybody producing a vst should test in as many hosts as possible.

[tha_man]

There should be absolutely no difference in exported wav file on different soundcards/drivers but same host. It's all done in software anyway. There could be difference when playing that wave through different drivers or soundcards, because then various equalizers and such stuf comes into play (especially when using directx). And D/A converters.

Also there should be no difference when exporting single wave from any host, except maybe in volume. I'm talking about audio tracks here, not when playing sounds with VST sampler! But when mixing multiple tracks together, it depends on host's mixing algorhitm how it will sound.

There are big differences between VST (or internal) samplers though, especially when pitching samples.

Don't know about VST synths though. In theory they should sound the same in all hosts, but they probably don't because of bugs or some features that certain hosts support and others dont. I've heard that there is no official guideline how to make VST host, only how to make a plug-in, so there might be some differences in VST implementation between hosts that could affect the plugin's sound quality.

[nuffink]

Different hosts will process audio differently internally and the sounds of the vsts will have passed through the hosts audio processing at mix down if no where else, because of this most plugins will sound very slightly different from host to host - some will sound very different, this is also the case with different audio drivers, so you may find that a vsi played through the same host on different machines, but with identical sound cards and monitoring might sound different if say one machine was playing back using asio drivers and the other was using direct x. Again this is entirely down to how the data is processed in the mix down of the audio streams.

This is one reason why anybody producing a vst should test in as many hosts as possible.

[saulc12]

There should be absolutely no difference in exported wav file on different soundcards. It's all software anyway. There could be difference when playing that wave through different drivers or soundcards, because then various equalizers and such stuf comes into play (especially when using directx).

Also there should be no difference when exporting single wave from any host, except maybe in volume. I'm talking about audio tracks here, not when playing sounds with VST sampler! But when mixing multiple tracks together, it depends on host's mixing algorhitm how it will sound.

There are big differences between VST (or internal) samplers though, especially when pitching samples.

if one host processes audio internally ie through its on mixer as 32 bit audio while another processes audio as 16 there will be a difference because the 16 bit process will intoduce more aliasing -

As for the different sound cards and drivers, my reference to this was only in relation to playback not to rendered output, although there are some cases where rendered output could also be affected, if for example the soundcard would only support output at a specific playback rate or bit depth and if the host only offered the same setting for rendered output as it did for playback - and anybody who does not believe this shold try taking any vsti they like and playing through different hosts - rendering output, then do a sample by sample comparison (ie subtract one wav from the other in a sample editing package - if there is no difference the result will be zero, if there is a difference the result will be low level noise). Try it.

[saulc12]

Different hosts will process audio differently internally and the sounds of the vsts will have passed through the hosts audio processing at mix down if no where else, because of this most plugins will sound very slightly different from host to host - some will sound very different, this is also the case with different audio drivers, so you may find that a vsi played through the same host on different machines, but with identical sound cards and monitoring might sound different if say one machine was playing back using asio drivers and the other was using direct x. Again this is entirely down to how the data is processed in the mix down of the audio streams.

This is one reason why anybody producing a vst should test in as many hosts as possible.

and as for you 'nuffink'

[nuffink]

and anybody who does not believe this shold try taking any vsti they like and playing through different hosts - rendering output, then do a sample by sample comparison (ie subtract one wav from the other in a sample editing package - if there is no difference the result will be zero, if there is a difference the result will be low level noise). Try it.

We've done just that here at k-v-r. Cubase and Sonar nulled to (if I remember correctly) < -96dB. So please stop talking bollocks.

[valley]

if one host processes audio internally ie through its on mixer as 32 bit audio while another processes audio as 16 there will be a difference because the 16 bit process will intoduce more aliasing

are there *any* hosts in this day and age that mix at 16 bits?

[cron]

if one host processes audio internally ie through its on mixer as 32 bit audio while another processes audio as 16 there will be a difference because the 16 bit process will intoduce more aliasing -

Aliasing artifacts have nothing whatsoever to do with the bit depth of the audio.

[bitcrusher]

are there *any* hosts in this day and age that mix at 16 bits?

While I don't claim to be a DSP theory expert (or novice even lol) but I remember hearing from a DSP engineer that for mixing multiple digital audio signals, you theoretically want an added half-bit (3dB) of headroom for each track mixed. This is theoretical and not of course a real-world figure (you can do more in less bits without clipping or artifacts in a practical sense).

ProTools sounds better to my ears when mixing things than many hosts. It also has a 56-bit stereo accumulator. whether its all hype, I dunno. I trust my ears.

I've also had FL sound better when letting it play in realtime versus rendering. Maybe thats related to the actual quality *increase* when rendering on pitch interpolation. I guess I like my beats a little dirty

[saulc12]

if one host processes audio internally ie through its on mixer as 32 bit audio while another processes audio as 16 there will be a difference because the 16 bit process will intoduce more aliasing -

Aliasing artifacts have nothing whatsoever to do with the bit depth of the audio.

okay then smart arse, if aliasing artefacts are not a prodct of resoulution what are they - and as for nulling at -96db, well I am sorry, but this is not nulling at -100db is it, this low level noise that just happens to have been culled by a noise gate or at least that would be my interpretation, all that I do know is that I test my own vst instruments quite thoroughly in cubase, orion platinum and fl4 and they all produce very subtley different output and they all have different internal processing. So I may be talking bollocks, but at least my bollocks is based on observation and testing and at least I made a valid attempt to answer the original question.

[original flipper]

HI

Valid point; these are my findings -

LOGIC - is initially quite foward and warm (on a PC)but once you settle into its sound you get lulled into a false sense of security - which can have devestating consequences.

REASON - I do like the this app, it has a full bottom end which unfortunately is offset by the top end seemingly lagging behind - PDC?

ORION - new kid on the block (for me) much talked about in semi-pro circles as being more than able to instill that cold sterile harshness which was in vogue a few years ago - I disagree and in fact purchased this last week based on a warm bottom end and full bodied upper mid.

FRUITY - never heard of it, up until recently and must say if its a sparkling top end your after then......look elsewhere.
Only jocking, keep your pantie holes on, yes its all about texture with this one, I love the built in aliasing alogrithms that are strapped across the outputs - everything just sounds kinda valve like.

Tracktion - Mackies flagship host with a boast, or is that the most - who cares its gonna sound just like sx2 sooooon with all that digital debris flying around - but in the meantime you are likely to appreciate the tight highs and loose lows that are a trade mark of what or who I don't know, but lets not forget there is an onboard dynamics proccessor to bring things firmly under control should you get nervous or need a more spatial sound.

Part 2 to follow.

Flipper.