"All DAW's sound alike" and other wisdom of the Internet

[Starship Krupa]

17 hours ago, Tezza said:

Mixcraft sounds "grainy" which is good for dirtier sounds, Cakewalk sounds "analog", deeper and  thicker, good for making VST instruments more "live". Harrison mixbus takes that same sound to another level. Cubase sounds "clinical" can be very clean and revealing, which is good for mixing. Protools sounds "boxy", like the high and low end have had cuts automatically applied, I don't really use protools much though, I don't find it good for midi.

That's interesting, Tezza, my impression of Mixcraft vs. Cakewalk is just the same as yours, and I've heard it expressed elsewhere, too. Mixcraft, at least its playback, does (or did) seem to have an edge to it that became especially apparent once I got my hands on CbB. Cakewalk just immediately sounded....buttery by comparison. I'm now interested to try out the others you mention and see if my impressions are similar.

But as I mentioned in my original post, what the audio engine does on playback is different than what it does on render. I'm thinking of all the options we get when it's time to export or mixdown or render or whatever your favorite term is for creating that .WAV file at the end of the project. I've verified this by testing between Mixcraft and Cakewalk. With the very first project that I imported via raw stems from Mixcraft to Cakewalk, using Windows' own Resource Monitor, I observed that when Cakewalk plays back a project, it always streams its audio from the disk, while Mixcraft somehow doesn't.

It's true, on the same dataset, about 20 tracks of drums, guitar, vocals, bass, there was zero disk activity on the Mixcraft.exe process. Which suggests that they're somehow loading the whole project into memory and playing it from there, which also explains why dropouts are about as rare as women at a King Crimson show. When I got that first build of CbB I was astonished at the whole concept of the audio engine being this thing that could stop, and then I'd have to restart it, like pulling the cord on a lawnmower. With Mixcraft, if the audio engine stops, that's a crash, and you see if you can get it to happen again so you can file a bug report.? Thank heaven that Noel and the crew have gotten it not to do that so much.

But it may be that the stability has a cost, which is that they might be doing some compression finagling to get all the project audio into memory for playback, and fidelity may be suffering as a result. But as I say, playback and rendering are two different operations.

15 hours ago, Starise said:

Lately my answer has been YES they do sound different "if" you take into account additions besides simply the audio engine.

I think Cakewalk has a more overt up front sound. Not harsh, but not soft either...leaning towards detail . Don't forget the Pro Channel adds something to the sound in it.

By contrast I hear Studio One as being slightly softer. Not that you can't make it sound identical to Cakewalk with a few tweaks and plugins. Mixes I make in it usually don't need as much top end rolled off.

Mixcraft was a very "in your face"  DAW. Like many DAWs all channels are initially at 0db but for some reason it came off as hotter than most to me. I would say almost harsh. The new Mixcraft Pro 9 seems a little better.

Ableton seems closer to Studio One. My theory is they lit some of the mixing buss up with something to "smooth" things out in that same way as Studio One. A secret sauce of some kind. Something buried in the code.

Nothing in any of them is overt enough to make me not use it because the differences are so small you could easily overcome them with a few tweaks. I do tend to generally mix more "up front' when using Cakewalk.

Yes, you touch on another aspect, the look, feel, workflow, and available tools naturally leading the users of them in a certain direction. I typically turn that Console Emulator on for all channels except for my Mix output bus, and I wouldn't seek out and use a console emulation plug-in to use in that role in another DAW. I use it because it's right there in the ProChannel, so why not? Similar with the QuadCurve EQ. It's knocked MEqualizer out of my top spot, which is a feat. If QuadCurve EQ were not such a top notch (pun intended) processor I wouldn't use it, but its presence (there I go again) will contribute to Cakewalk productions having a "sound."

What I really wanted to vent about and get opinions on and headchecked about was what I consider the technical impossibility. I think it's probably true that they all record audio just the same, I would trust any of them to capture audio from my interface, but what I do with it after that? How is it possible that they would all handle it the same? Even, as @marled says, I agree, if they're using the same libraries, which I have considered when pondering this, libraries are a big deal with signal processing, the libraries are like using a certain company's tires on your automobile. There's still the rest of the car to consider.

And why would I bring this up here and now, @ensconced? ? Well, because I like this forum, and the people who contribute, and I respect your opinions, and I've now been around here long enough for people to respond to a question I post, and I wanted to toss out my opinion and see what you all had to say. Your memes imply that this topic has been discussed to death here on this forum, but I missed those discussions. Seriously, all I've seen on this is people on other forums saying as an aside "of course all DAW's sound alike, that's been proven" and I don't know what "proof" they're referring to, it seems unlikely past the point of impossibility. Is that opinion universal? Happy to see it's not. Do you have a take on it (other than you think it's already been discussed too much?)?

In the end, I do absolutely agree with @mettelus that the question itself is so, so not worth losing a minute's sleep over. Don't get me wrong on that. My rant had more to do about the idea of it than putting anything into practice. The perception of a sonic character is merely one thing to consider, along with ease of use, reliability, etc. and I am sure that any of us in this thread can make a decent-sounding mix using any DAW with sufficient familiarity (just as long as I'm not the one who has to use Reason).

[Starship Krupa]

14 hours ago, Michael A.D. said:

I can listen to something I did last night and think it was great and tomorrow it just sounds so-so.

I recently had a horrible experience with this....well, I thought it was this, and then it turned out not to be this. I have a system in my bedroom with a 1/8" plug next to my bed running to an old Yamaha A/V receiver on my dresser, going into an older pair of Infinity bookshelf speakers. After I've wrapped it up at the workstation, I do love to plug the phone in, fire up VLC, browse on my home network over to my Mixdowns folder and listen to the fruits of my labor, lay back on the pillow and do a bit of referencing, check out the soundstage in the dark. It's usually a fun moment, my mixes translate well to that system.

But this one time, I plugged my phone in, and everything just sounded like hell. And all those imposter syndrome "ha ha, how did you ever think you could be a musician or recordist or mix engineer, you suck harder than a Shop Vac!" voices were lined up in a mocking chorus. Oh man. How could I think such wimpy low end and weird phase-y rolled off highs were even acceptable? How did I miss that? What is UP with this?

I was kind of in the dumps for part of a day about this. Then I went into my bedroom and decided to check on this situation. I plugged the phone in and played my favorite track from the last Dave Tipper album as a test and....hmm....one of the most amazing mixes I've heard in my life also sounded anemic. So I put on my glasses and looked, and somehow, probably me fumbling around in the dark, the processing button (which for some reason affects all inputs, not just the video ones) had gotten pressed and pressed again, to the point where the Yamaha thought that it was a 5.1 system decoding ProLogic. ?‍♂️

[Wibbles]

Have any of you tried comparing DAWs using random double-blind testing?

Everything else can be put down to confirmation bias.

[Starship Krupa]

30 minutes ago, Wibbles said:

Have any of you tried comparing DAWs using random double-blind testing?

Everything else can be put down to confirmation bias.

I certainly haven't. How would you even design such a test?

I suppose it could, if one is looking to put things down. If a person is convinced that "they all sound the same" and without benefit of controlled testing tries a few DAW's (we'll give them similar program material in their own studio), and comes to the conclusion that yes, indeed they do, their findings would have the taint of "confirmation bias" as well.

Perceived differences in sound between DAW's can only be attributed to "confirmation bias" if one assumes they are designed or claimed to sound exactly similar in the first place, and there is no expert or authority that I know of that says they are, just random forum hearsay. I don't know of any DAW manufacturer who has made that statement, to the contrary, MAGIX, who make 3 of them, made explicit claims otherwise. And they made the claim about one revision of their own products in relation to a previous one. That is to say, the people who make these things say that they build them to sound different from the next company's offering. Qualitatively "better," in their collective opinion.

I have seen laypeople making the claim that all DAW's sound alike on discussion forums, accompanied with the statement "it has been proven." I am perplexed by the former claim and curious about the latter. If some people believe that this (to me, unlikely) claim has been proven, I want to know how. I'm curious why this is such a common belief among DAW users. I've already said why it doesn't seem very likely to me, I would love to hear a credible explanation to the contrary (it would make life so much simpler).

I question whether DAW's are designed with that goal in the first place. I think that it's probable that the people who program Cakewalk have no interest in making sure their product sounds like REAPER and vice versa. I doubt that the programmers at Avid regularly give an A/B listen to Logic Pro to make sure that Pro Tools is still conforming to Universal DAW Soundalikeness. However, they might very well sit down with it to make sure that by their standards Pro Tools still sounds "as good" or "better."

My contention is that too many unrelated people would have had to go the same way on too many independent decisions for it to be even possible for 2 DAW's to sound exactly similar. And by "exactly similar," I mean in actual use, mixing and panning, using plug-ins, and rendering out a mix from all that.

[paulo]

8 hours ago, StudioNSFW said:

There is another DAW besides Cakewalk?

According to Bapu there is, but I think he's just paid to say that.

[Wibbles]

1 hour ago, Starship Krupa said:

How would you even design such a test?

Designing the test would be trivial, but the results would be of negligible value.

[Starship Krupa]

2 minutes ago, Wibbles said:

Designing the test would be trivial, but the results would be of negligible value.

To clarify, I'm not talking about subjective tests, I'm talking about bit-for-bit and audio nulling tests, things that you can see on test equipment and analysis plug-ins. That's the kind of proof I'm interested in, otherwise, I'm with you, listening tests are subjective. And then, who cares, Studio One has flat graphics, so I think it sounds precise, and Cakewalk has rich 3-D graphics, so I think it sounds round and full, and Ableton has bright high tech graphics, so I think it sounds "in your face," and Mixcraft's graphics are kinda jaggy, so it sounds "grainy." ?

I've designed boutique guitar amps, I'm no stranger to silly corksniffery, ceramic capacitors vs. film vs. silver mica when it's a shunt going to ground? I was actually bummed when I had to admit to myself that ceramic caps actually do sound harsh and grainy when used as couplers. Ugh. At least I can see it on my oscilloscope.

I had been thinking that designing the test was not trivial, but you're right, it ain't all that.

How about we start by giving our DAW's three audio files. 30 seconds each. A 440Hz sine wave, a 1KHz sine wave and a 30Hz sine wave. All 44.1/24. We import them to 3 tracks, pan the 30Hz center, the others at 35L and 35R (uh-oh, pan laws), send them all to a stereo bus, then render the output of the bus to a WAV file. Then compare the resulting waveforms. Using an audio editor or spectrum analyzer or whatever analysis tools.

If we really wanted to get gnarly, we could start tossing in square waves and see how their mix engines handled those.

Thanks, Wibbles, I might try this and see what happens.

[smallstonefan]

I was very fortune to just spend 7 hours with Andrew Scheps this week for a special Mix with the Masters virtual class and he mentioned that he thought Ableton might have a limiter on it's master bus. Not an effect that you can see, but that it's there.

That would affect the outcome.

Now I've got to go take photos and box stuff up because he inspired me to get rid of my analog mix bus chain. True stroy!

Edited August 1, 2020 by James Foxall

[craigb]

1 hour ago, Wibbles said:

Designing the test would be trivial, but the results would be of negligible value.

So, our governments will be racing out to waste millions on testing soon, ya? ?

[Bapu]

22 hours ago, Michael A.D. said:

I can listen to something I did last night and think it was great and tomorrow it just sounds so-so.   I guess that would apply not to just my creations, but to any music - my state of mind has more to do with my reaction than any purely technical aspect.

Yup, been saying that for years.

[Bapu]

1 hour ago, James Foxall said:

Now I've got to go take photos and box stuff up because he inspired me to get rid of my analog mix bus chain. True stroy!

Wow after all that studying and purchasing and setup.

Andrew is a big proponent of mixing ITB.

[Bapu]

@Starship Krupa I believe what has been considered proven in the way past is that when all things are equal (pan law etc.) between the then tested DAWs on pure audio (i.e. no use of any FXs) they always null tested out.

Now Andrew Scheps' observation (if true) could shape a null test between Ableton and any other DAW (purpoted to not have any inbred processing). 

I personally believe that Mixbus/MIxbus32C would not null to any of my other DAWs because they advertise that you are by definition getting a Harrison console, which is different for Mixbus vs Mixbus 32C, right in the DAW.

[Bapu]

3 hours ago, paulo said:

According to Bapu there is, but I think he's just paid to say that.

Nah, I was not paid to say that. I truly paid to say that. ? 

@paulo Whenever you'd like I can list all the DAWs I own and which one is my primary. ? 

[paulo]

25 minutes ago, Bapu said:

Whenever you'd like I can list all the DAWs I own

 

[Starship Krupa]

4 minutes ago, Bapu said:

I believe what has been considered proven in the way past is that when all things are equal (pan law etc.) between the then tested DAWs on pure audio (i.e. no use of any FXs) they always null tested out.

Ah! Thanks, Ed, this was exactly what I was looking for. So someone supposedly did, back in the mists of time, do something similar to the test that I outlined.

I'd really, really like to know who, when, and how, y'know? And what DAW's they tested, and if they published the results on a website or what.

I was just poking around the www to see what had been written about this and found a fascinating section in the Ableton Live! manual that goes into great detail about exactly what operations do and don't affect audio quality (their words). They have a suite of 473 tests that they subject the product to with each build to make sure that any changes haven't accidentally degraded some aspect of the measurable audio performance.:

https://www.ableton.com/en/manual/audio-fact-sheet/

In the course of describing this, they do say that they made improvements after v.7, so they are in effect saying that v.8 sounded different from v.7, no way around that. You can't be "better" without being "different."

Also, I've been keeping this in my back pocket, but here is a site that shows the visual results of having different DAW's perform the simple task of downsampling a swept sinewave file from 96K to 44.1K. We're all pleased to see that SONAR has had nothing but that one ghostly harmonic from day 1, while some others were fireworks shows at first but cleaned up their acts, likely as a result of looking bad on this site. Take a look at earlier versions of Sequoia!

https://src.infinitewave.ca/

I am sorry to say that I downloaded their file package and ran the tests on Mixcraft 8 and they were not 100% linear. I submitted the results back to Infinitewave, but they've never published them. Maybe I botched the submission somehow, but I was pretty fastidious. I also told the devs at Acoustica about my findings and they were silent. No reply at all. I could play those converted sweeps up to 20K and using Voxengo's SPAN, see (and hear) the harmonics starting to kick off around 10K. And they were up high, but I could hear them affecting the audible range with my over-50 post-punk hearing.

So therein might be some explanation for Mixcraft's edginess, or maybe not. I could only break their conversion algorithms in certain directions, in others they were clean as a whistle. Would this indicate any non-linearity in the main audio engine? I haven't a clue.

[smallstonefan]

1 hour ago, Bapu said:

Wow after all that studying and purchasing and setup.

Andrew is a big proponent of mixing ITB.

It was really cool to be able to ask him questions. He is using an Apollo x8p I think it was and said they were great! He says the conversion is fhe biggest problem and staying in the box prevents the double degradation you get going out and back in. It makes sense and the guy has really done his homework.

I made a copy of my current mux and set up his 2bus and bypassed my analog chain. I had to tweak a few things but wow was it better. I figure if it is good enough for him and Brauer than having less variables to deal with is good for me.

I've enjoyed the summing experience and playing with some really good gear. But havi g to constantly check gain staging is a pain. Andrew also said everyone you have to adjust your gain out fir summing you are effecting your mix because you hit the analog slightly differently. The science makes sense to me.

Now, what to buy with the money I get from selling gear? Maybe a sitar?

[ensconced]

[ensconced]

14 hours ago, Bapu said:

@Starship Krupa I believe what has been considered proven in the way past is that when all things are equal (pan law etc.) between the then tested DAWs on pure audio (i.e. no use of any FXs) they always null tested out.

Now Andrew Scheps' observation (if true) could shape a null test between Ableton and any other DAW (purpoted to not have any inbred processing). 

I personally believe that Mixbus/MIxbus32C would not null to any of my other DAWs because they advertise that you are by definition getting a Harrison console, which is different for Mixbus vs Mixbus 32C, right in the DAW.

Didn't you (or someone) do a test on the old forums that showed at that time that if none of the EQ's Comps, Sat etc on each channel/Bus of Mixbus were used that they actually did null? My memory may be failing me but . . .

[Edit] Perhaps it was Jeff Evans, I found this comment from dear old Sharke (God bless is soul, he was a fun chap) and mettelus but as yet haven't found what he is referring to, it may be what I am recalling (but I grow weary of the search)

Quote

But as Jeff Evans pointed out recently, tracks summed and exported from a "transparent" DAW will null perfectly with tracks summed and exported from Mixbus provided you don't engage any of its effects. So it's not really doing anything to the sound on its own. The Harrison sound comes instead from the Harrison effects - the inbuilt EQ and compression, the tape drive, and the various add-on effects you can purchase from Harrison. I read somewhere that the tape drive isn't even that sophisticated - it just adds some third order harmonics and engages a slight rolloff.

and this by mettelus

Quote

But seriously, Jeff Evans has posted about comparisons between 4 DAWs using the same stems and the results were identical. He has posted that a few times now with the conditions of the test he used, stems only, no FX, and only L/C/R for pans IIRC.

 

Edited August 2, 2020 by ensconced

[mettelus]

Jeff made reference to that several times in the old forums and discussed it in detail at least once, but I cannot readily find it. This is one of the references he made that calls out the DAWs he used (Logic, ProTools, SONAR, and Studio One).  There is also a lengthy thread called "Comparing SONAR to Studio One" that he chimed in a few times on. This topic gets resurrected off and on, along with others.

[ensconced]

Yeah I came across those and quite a few more on my search, what I recall was more to do with Mixbus, and I remember the end of it all was kinda like a unanimous 'Well there you have it, no need to wonder about that any more' type of thing. But I have spent enough time looking, it could be a discussion inside a discussion inside a discussion.

Anyway doesn't really matter, as long as no one falls off and gets hurt . . .