The sound of other DAWS versus Tracktion....

[Aleatoric]

I was reading on some other forum about DAWS actually sounding different....like Logic sounding different than Cubase SX. It has to do with how some engine is written or something. How does Tracktion compare sound wise to the other DAWS out there?

[inklling]

I swear that I can hear a difference between Tracktion and the Cubase 5.X I was using before. I think Tracktion sounds a lot more clear and each instrument is more defined. Is this an illusion creasted by how clear the work flow is? My finished tracks do sound very different now...

[valley]

these threads come up from time to time, and are usually pointless.

In my world, most serious hosts (Tracktion included) work at 32bit internally and basically will produce the same output given exactly the same input and parameters.

In the world of others, and I'm sure a few f them will be along soon to tell me I'm stupid, there is a clear audible difference.

It is worth noting though that some people claim to be able to hear the difference between FL3 and FLStudio despite the fact the Gol has gone on record as saying he never touched the audio summing code.

Is this an illusion creasted by how clear the work flow is? My finished tracks do sound very different now...

Thta's a pretty likely guess. The screen looks bright and open, there is less clutter, and you *hear* that as a part of your sound because that is the mood that you are currently in.

[AndrewSimon]

I think the answer is in the built in effects that tend to color and sound different from program to program.
If you use the same effects in both programs the problem will probably disappear.
It is also possible that the interface has an effect on you and the way you play and mix.
I know it sounds silly but I always play better on guitars that have dots on the fingerboard as oppose to inlays. There is just something about it that changes the way I play.

[P.T.]

I tend to think that some sequencers have some hidden EQ applied to the output.

When I started in with computer music I got a copy of magix audio studio.(ver5 I think).

I then got guitar tracks and imported some of the waves I had made in magix.

The sound seemed a bit muffled in GT. When I later got the demo of Sonar I loaded those same waves and they sounded the same as in Guitar Tracks ( also made by Cakewalk).

Since I believe Sonar to be a 1st class sequencer, I tend to think that the magix audio studio must have had some high freq EQ to brighten up the sound.

[IIRs]

I am convinced that I heard an improvement when I moved from Cubase VST to Tracktion. Possibly my old version of Cubase does not have a fully 32bit audio engine?

The differences I heard were subtle but cumulative: more detail in reverb tails for example, better stereo imaging & generally more "depth" to the sound.

I am not convinced that there would be any difference compared to another 32 bit app however..

[ericj23]

i think the difference between apps is not because of summing engines

merely that duplicate the exact settings (particulalry to do with panning laws) between them is almost impossible

all 32 bit sequencers are capable of sounding great
- unless i use them !

[Mr. Tingle]

This is a great topic, I hope something concrete shows up. I mean is there only one way to write audio summing code? Maths is maths, right?

I'd love to think that Jules approached the whole thing from a different angle and has done something different(better) that no-one else thought of, kind of like beginners mind.

[ZooTooK]

I havn't coded any DSP stuff in 15 years, but I think it's making things too simple just to say that it's math and 32 bits and therefore all DAWs are equal. Being in the digital domain let you have great control and using 32 bit make things easier but never the less, you can do it in several ways.

[valley]

I tend to think that some sequencers have some hidden EQ applied to the output.

No host *should* be applying hidden eq to a sound in any form.

If you take a wav file, load it into your sequencer, set everything at unity, disable dither, and export it, you should end up with something that is as near as damnit a perfect replica of the original.

Any hosts that are applying hidden, uncontrollable processes to a sound are doing a bad thing.

[valley]

I havn't coded any DSP stuff in 15 years, but I think it's making things too simple just to say that it's math and 32 bits and therefore all DAWs are equal. Being in the digital domain let you have great control and using 32 bit make things easier but never the less, you can do it in several ways.

Why would you though? Summing is as simple as s0+s1+...sn

Why would you seek to make it more complicated?

The only different approach I can think of other than altering bit depth, that may make some level of difference is switching between integer and floating point math. However, since all plugins are floating point, whatever advantages integer has as a summing format would be lost by converting to and from floating point anyway.

I'd be interested to hear what alternatives you know of, and in what way that would affect the overall quality/accuracy.

[Karbon L. Forms]

Does Jules write all his own DSP stuff then? If so I'm impressed! Or is there some use of 3rd party libraries?

[christianmusicmaker]

I tend to think that some sequencers have some hidden EQ applied to the output.

No host *should* be applying hidden eq to a sound in any form.

If you take a wav file, load it into your sequencer, set everything at unity, disable dither, and export it, you should end up with something that is as near as damnit a perfect replica of the original.

Any hosts that are applying hidden, uncontrollable processes to a sound are doing a bad thing.

Agreed that would be strange.

[ttauri]

If you take a wav file, load it into your sequencer, set everything at unity, disable dither, and export it, you should end up with something that is as near as damnit a perfect replica of the original.

I had some stem mixes lying around from the last project, so I gave it a try, and indeed at unity, a phase reversed file from Logic or Tracktion nulled the file from the other.

OK all's well and good. But then I tried a second experiment: I bounced down with all the stems' volumes set -3.1 db (reason for that value has to do with the increments available on Logic's mixer fader). After doing that and trying the null test, they don't null. Tracktion's problem? Logic's problem? My problem? I dunno. But a simple change produces different results.

Peece,
T. Tauri

[valley]

But is the difference distortion, or is it merely that they don't quite agree on what constitutes as -3.1db?

My guess is probably the latter, since most scalling is simply a case of value * scale

It's hard to get that bit wrong.

what is easier to get wrong though is the bit that takes a value in a non-linear scale, such as dB, and attempts to convert it to a linear form. Rounding errors there could easily give slightly different results. This doesn't signify a difference in sound though, because if you were mixing by ear, and were caspable of hearing the difference betweem 3.1dB and 3.11dB, you'd naturally compensate for it.