4 DAW's, 4 renders, 4 results

[Starship Krupa]

14 hours ago, Bruno de Souza Lino said:

It's that famous thing we've all done at least once: taking a mix then tweaking, say, a snare to perfection only to discover the EQ was bypassed the whole time

At least once?? 😂

Yes, I've tweaked compression settings to perfection using bypassed compressors. About once a month. I expect the frequency to increase as I get further past 60.

14 hours ago, Bruno de Souza Lino said:

or you were tweaking a different track and being sure your actually heard a change at what you thought you were tweaking.

Ah, but the thing about this scenario is that tweaking the EQ and compression on one track will affect the sound of another track. Right? Make some cuts or compress over here, and then suddenly the track over there starts popping out in the mix. It can even happen with things like delay and reverb where the track we're working on gets moved back in the mix. In the Jenga game of mixing, what affects one track usually affects the others.

I stumbled across that one, and while I already "knew" it intellectually due to studying masking and how to reduce collisions, the experience of tweaking the EQ on one track and hearing the results on another track was the revelation moment. 🎆

Okay, here are the revised files.

Edited May 3, 2023 by Starship Krupa

[Starship Krupa]

Revised the package of files again.

I listened to them and Thwarp C.FLAC sounded as if it were lower in volume than the others, so I opened the DAW project to check. All of the track levels were set to what my instructions call for....except the fader for the Master bus, which was set to -3. No idea how it got that way. I never deliberately touched it.

Thwarps Beta 3

[Peter Morgan]

On 4/30/2023 at 11:03 PM, Starship Krupa said:

Definitely a consideration. I chose arp patches that produced the same notes when they are triggered. I may have got that wrong, though. It's all a big experiment.

How about making two passes at each DAW and then compare pass 1 and 2 of each DAW to be sure they are the same as each other.  If they aren't you aren't really comparing the DAWs so much as the takes.

[Starship Krupa]

Okay, I took my files, and using a tool recommended by @John Nelson, HOFA 4U+ Blind Test, I was able to do a blind listening test.

Drawback to the tool: it's a plug-in, so you wind up listening through a DAW's playback engine. At least I used a DAW that didn't automatically copy and convert imported audio files (I know Cakewalk can be set up like that).

Results: I couldn't reliably pick a favorite.

Conclusions, lessons learned:

For the It Is Carved In Stone: All DAW's Sound Alike crew you can add it to the "some guy on a forum tested it" file, I suppose.

For me, who is kinda agnostic on the matter (I believe that it is possible for two different DAW's to sound different), my conclusions are that first, it is really hard to come up with objective ways to test this. You can use sine waves, square waves, impulses, whatever.

But it's also really hard to get two DAW's to sound alike. Which suggests that in practice, given similar projects produced using different DAW's, they will inevitably sound different in ways that have nothing to do with how the programmers chose to implement Fourier's theorems. This way and that, they nudge you in certain directions.

So another conclusion is that if you sit down and compose and mix an ITB electronica project, doing it with FL Studio vs. say, Ableton Live! will produce sonically different results. Same with a band recording. Even the way the meters look and respond will influence decisions you make about setting levels and so forth.

If you're in doubt, it's easy enough to just try it: download the free or trial versions of the software you're considering and mess around with the kind of project you usually do and pay attention to how they sound. If one stands out, there ya go. It's your ears that need to be pleased, even if it's ultimately down to a placebo effect. Hit records and audiophile material have been created in a dozen different DAW's. Some people with very talented ears claim to be able to hear a difference; Ray Charles famously chose Cakewalk SONAR out of all the ones he listened to. Mr. Charles certainly wouldn't have cared what the program looked like.

Another lesson (not so much learned as confirmed) is that when doing a final mixdown and render of any project that includes randomized elements like arpeggiators, glitch FX, even (or maybe especially) reverb, multiple renders will be different. So do what I've been doing since day one: render once to lossless, convert as needed. If you do a render for each format you wish to distribute, they will all differ from each other, even if it's just subtly. But I don't think anyone wants the MA4 to sound different from the FLAC which sounds different from the MP3.

For the conversion task, I like to use MediaHuman Audio Converter. I'm also investigating AuI ConverteR 48x44.

And for heaven's sake, although nobody has ever paid attention to me on this, when you're listening to your rendered files, use a music player that can at least use WASAPI Exclusive, preferably ASIO. Music Bee, AIMP, Foobar, etc. Windows' internal mixer is known to have a negative effect on audio that's sent through it, and it's best to bypass it (as we do in our DAW's).

[Lord Tim]

As one of the "if they don't null they're broken" crowd, I both stand by that... *and* agree with what you've come up with.

Like I said earlier, if you have a repeatable source (imported or recorded audio) and just mix in mono with that, any changes will be practically inaudible between DAWs. Add panning and you're at the mercy of pan laws, which can get fairly close and you can compensate a bit by raising or lowering levels, etc. I'd say that if you got them close, the results would be tiny in difference.

Then you have effects. Again, unless we talk about upsampling, most effects effects will work almost identically across each DAW, other than free flowing things like modulation or algorithmic reverbs, etc. But they would sound different in the same DAW on each pass anyway. Does anyone use all of the same effects in each DAW? Possible I guess but there's a better chance someone will use a stock effect somewhere that's not present in the competition.

Synth rendering, especially with synths that have those kind of built in effects (like, most of them) will be the same as what I just mentioned.

So I think both statements can be correct.

That said, I don't think anyone would be trying to 100% match settings from DAW to DAW anyway, they'd just be actually mixing, and letting their ears guide them.

And therein lies the real smoking gun.

I personally don't think any DAW has it's own sound ("I mix in ProTools because the high end is more glossy, Bitwig sounds more analogue", etc) and if you just mix as you ordinarily would, you'll naturally be compensating for the differences in pan laws, etc. But in this case, it's now a *different* mix, and a different mix won't null, even if underlying math of the DAW *will* null with other DAWs.

The quality of DAW engines these days is exceptional, and even entry level audio interfaces are pretty damn good if not great. The real differences, sonically, between them all is ultimately negligible, taking out the variables above. Workflow is the real decider at the end of the day. If that suits you better in one DAW over another, you'll make different or maybe better choices with your mix. THAT will have a huge impact compared to any possible math differences in engines.

And on top of that, let's not count out speakers and environment too. Slap some panels over that back reflecting wall or do something to cut down on your room standing waves and I bet your mixes suddenly sound SUBSTANTIALLY better, no matter which DAW you use! Far more than a different pan law or how a compressor reacts in a FX bin.

Ultimately I choose Cakewalk because the workflow lets me get to the goal the fastest. I have no doubt I could get similar results from any of the other DAWs but it's less enjoyable if you're fighting it all along the way. A happy mixer dude is going to make better mixes (but I can't guarantee I'll null against myself if I don't have coffee on hand 😒)

Great work on the tests, Erik - the last lot of stuff especially was super interesting

 

[azslow3]

10 minutes ago, Lord Tim said:

That said, I don't think anyone would be trying to 100% match settings from DAW to DAW anyway...

As you could conclude from my previous post, there are such people... At least I (the author of ReaCWP) and  the author of AATranslator have tried.

Every single tiny parameter and property, including "trivial" (f.e. fader position, timeline position), are not easy to match even with dedicated and analytical effort. In some cases that is at least possible (f.e. fader position), but in most it is not (f.e. an automation of fader, if it has any curves and used interpolations inside DAWs do not match).

And so

22 minutes ago, Lord Tim said:

... they'd just be actually mixing, and letting their ears guide them.

that is the only reasonable way to go 😏

[OutrageProductions]

In all my years of experience there have been two unassailable assets regarding production of a record. 

And none of it (after the recording phase anyway) has anything to do with the actual gear. All things being relevant and subjective, it doesn't depend on the recording media, or the console, or the room, or the outboard gear, or the DAW, or the plugins, etc.

It all depends, ironically, on the mixing engineer and the mastering engineer, and their ears.

[Glenn Stanton]

if you want a more flexible "listening" app for your 2-track (or multi-track if you're doing that), then an audio editor like SoundForge, Acoustica, etc can be used in lieu of a media player which may (or not) have "stuff" enabled to make things "sound better" (even if you think you disabled it). the advantage is, if you find something which is nagging at you, those apps let you do some surgical tweaks (which you can go back and adjust in your mix later) to see if there are things you might want to change before the mastering step (or even after if you dare 🙂 )

Edited May 8, 2023 by Glenn Stanton