Do different DAWs sound different?

[Michael A.D.]

Here's another one:

 

[Starship Krupa]

4 hours ago, Mark Morgon-Shaw said:

RME has an internal Loopback so he uses that

That is NOT the same as physical loopback, where you run cables from 2 of your interface's outputs to 2 of your interface's inputs. Internal loopback bypasses a bunch of system processes that potentially have an effect on the audio. Since a DAW must use those processes, they should be part of the signal path.

I discovered this when trying to do null tests of the effect of using different playback modes. ASIO vs. WASAPI Exclusive vs. WASAPI Shared vs. WDM vs. MME. Some of those are resampled through the Windows audio mixer subsystem and some are not and I was getting results that nulled perfectly, which I don't believe is possible. I can definitely hear the difference between modes that resample through the Windows mixer and modes that don't (or at least I was able to before my tinnitus flared up). I had misunderstood the routing. The interface's loopback must be picking it up before the player sends it to the Windows mixer, therefore, it's not as thorough a test as I want.

As you see, I'm following the scientific tradition of redesigning my test until the results confirm what I already believed. Only then may I rest.

[Starship Krupa]

10 hours ago, jayeyehaich64 said:

 

 

That's Premium jibber-jabber to you, son!

[Gswitz]

I remember doing a flac test.

I bounced to 24 bit 48.

I exported as flac.

Imported. 

Null test bounce. 

Normalize the result of the null test bounce.  

There was clearly one bit of info lost in the flac. 

I think that wave does not store the most significant bit because it presumes it is 1. I think flac does store it. Therefore, flac minus wav loses 1 bit of sound. 

Edited Tuesday at 02:59 PM by Gswitz

[Xoo]

12 minutes ago, Starship Krupa said:

That is NOT the same as physical loopback, where you run cables from 2 of your interface's outputs to 2 of your interface's inputs. Internal loopback bypasses a bunch of system processes that potentially have an effect on the audio. Since a DAW must use those processes, they should be part of the signal path.

Nope - it appears as an audio driver so behaves the same as physical cables (but gauranteed to be bit perfect - no jitter etc).  Assuming using a driver mode in the DAW which doesn't go through Windows mixer (ASIO, for example).

[ptheisen]

39 minutes ago, Starship Krupa said:

As you see, I'm following the scientific tradition of redesigning my test until the results confirm what I already believed. Only then may I rest.

LOL At least you're being honest and up front about it! 😉

[Starship Krupa]

1 minute ago, Xoo said:

Nope - it appears as an audio driver so behaves the same as physical cables (but gauranteed to be bit perfect - no jitter etc).  Assuming using a driver mode in the DAW which doesn't go through Windows mixer (ASIO, for example).

The tests I was running when I came to the conclusion that virtual loopback might not be the tool for the job were tests of different driver modes. When an ASIO playback nulled perfectly with playbacks using MME, WDM, and WASAPI Shared, I got kinda suspicious.

I still want to do my comparative tests of playback using different driver models, but next time I'm also going to include comparative tests of playback using the same driver model using different DAW's. I'm already pretty confident that there's no perceptible difference when it comes to rendering engines, but not yet as confident about their playback engines. Not so much these days, with the rapid increase in computing power, but say, 10 years ago, there were solid reasons to come up with clever ways to reduce the amount of CPU power that playback uses.

Test methodology is tough. Rick Beato "proved" a few years ago that there's no perceptible difference between lossless audio files and MP3's. He did this by setting up comparative listening tests offered up by a web server and having some of his interns, presumably a group of people who would be capable of hearing any actual differences, choose which version of the song under test they thought was the high bit rate, lossless, whatever. Since it was done via web server, he also opened it to the general public.

Whaddaya know, the tests "revealed" that not even trained listeners could reliably tell the difference until you really cranked the MP3 compression up.

That's what he thinks he demonstrated. I took a look at the setup and decided that what he had actually proved was that when you play them back through a friggin' web browser, which then goes through the OS' mixer, most people, including trained listeners, can't tell the difference.

What he didn't test is whether they could perceive a difference when the different files were played back through Music Bee or AIMP or JRiver or some other media player capable of using ASIO or WASAPI Exclusive set to "event driven."

In other words, his tests proved that the web browser's native player, compounded by sending its output through the Windows system mixer for a few rounds of resampling might be capable of masking the sonic differences between these file formats.

I pointed this out in a reply to the YouTube video in question, but of course nobody paid any attention. The Great Beato had spoken.

But for something that is actually within reach of my ability to set it up, I'm not going to be satisfied with "appeal to authority" arguments or hand-waving. I'll find my answers when I or someone I know completes null tests of multiple DAWs' playback engines. NOT rendering engines. I ran my own tests on those, and it seemed that while they didn't null perfectly, there was no way my ears would be able to pick out any differences.

Until then, I don't know whether their playback engines sound alike. I seem to be able to hear differences,  but I'm not going to trust subjectivity when the perception could change based on what color theme the DAW is using.

I'm not trying to prove a theory, I'm trying to answer a yes or no question. It's odd that some people push back so hard on even asking the question.

[Terry Kelley]

I give up.

Yes they all sound different but only the discriminating mind can tell. Especially with beer.

Edited yesterday at 01:51 AM by Terry Kelley

[57Gregy]

Sometimes I'm glad my hearing is so bad. I couldn't tell a bad-sounding DAW from a good one if you offered me a million bucks.

[Starship Krupa]

15 hours ago, Michael A.D. said:

Here's another one:

 

That is a great test of what their rendering engines do when importing a finished song, then exporting it without doing any processing or panning. Which is what I would expect, having done similar tests, albeit with a soft synth and 50L/50R panning. And not all of my renders nulled with each other.

What it doesn't test is how the DAW deals with mixing tracks together, panning tracks, using plug-ins, summing, and, most importantly, what playback does vs. what rendering does.

How about importing stems and then rendering? That would test the summing process. How about importing stems and panning the stereo tracks 50/50? That would test panning. How about putting the same compressor using the same preset on one of the tracks? That would test handling plug-ins.

Also, weirdly, given that he's testing what they all sound  like, he spends zero time listening to what playback sounds like. For all we know, FL Studio could be putting a slight oomph into the bass on playback but not on rendering. There's no law that says they can't. When we try to test it objectively, like Whytse and this guy did, we'll get the same results they did.

To put it metaphorically, concluding that two DAW's "sound the same" after importing a single audio file and rendering and successfully null testing is somewhat like concluding that two automobiles "drive the same" after driving from one city to another and arriving at the same place both times.

Please understand that I'm NOT saying that "different DAW's sound different." All I'm saying is that it's possible. Trust me, I would love it if thorough, objective tests proved otherwise. There are enough things to weigh when selecting a DAW without introducing sound quality questions.

I'm perfectly fine with the idea of my own perceptions of Mick's Craft's playback sounding different from CbB's having been a matter of my brain's reaction to the differences in UI. I would find that even more interesting, really.

[Starship Krupa]

At last!! Playback engine tests!!

But not really last, Whytse admits that there are still further tests he can run, AND Luna didn't null.

Major props to him for paying attention to the comments and not stopping the moment he got the results he was expecting.

(I really wish he had bypassed the plug-ins with the dithering noise when he was doing the playback tests)

I'd really love for him, or someone, to do similar tests on Mixbus, which straight up claims to sound different. And FL Studio, the one whose bass supposedly sounds "better." I still want to check Mick's Craft vs. Sew Gnar.

Regarding FL Studio, if I'm producing EDM, do I want a DAW whose playback has a bass emphasis? Seems like that would result in mixdowns with attenuated bass, but what do I know? Maybe they also apply that bass bump in the render.😄

[jayeyehaich64]

I believe Bapu did a series of tests with regards to Mixbus back a bit, you should be able to dig those up if you desire.

Edited 19 hours ago by jayeyehaich64

[Gswitz]

Mix bus is cool and keeps improving. The export measures loudness on ebu, which i think is helpful if you want the tracks of your record to be similarly loud. 

But in cakewalk, I listen while measuring loudness and I often find things to adjust. In other words, listening to the mix has value. 

My brother pointed out to me that I don't own any microphones with a low enough noise floor To test my interface specs, even if I had a silent chamber to put them in. 

I color lots of stuff. Tubes. Silk on the pre amp. 

I'm totally fine that I cannot hear the difference of two high quality interfaces. I can hear a good preamp from bad. I loved the sound of a focusrite that I had for a bit. It sounded fantastic. 

The rme sounds transparent to me now, but that's because I've had it so long. Idk how long. 15 years? 

I don't replicate all my steps mix to mix even when using templates. If you use templates, try skipping it once in s while. Doing things manually gets you thinking/ listening again. 

 

Edited 18 hours ago by Gswitz

[Starship Krupa]

3 hours ago, jayeyehaich64 said:

I believe Bapu did a series of tests with regards to Mixbus back a bit, you should be able to dig those up if you desire.

Bapu did? How did I miss that? Was it on the old forum or within the past 7 years?

Do you remember what conclusion(s) he came to?

I see that Harrison's pitch is that the sound is "inspired by Harrison's legendary analog console sound & workflow." "Inspired by?"

Like they were thinking about  the hardware product line while designing the DAW?

The first DAW that ever had a console or mixer view was "inspired by" legendary analog consoles. Because why reinvent the wheel?

[Terry Kelley]

People think boosted high end, cut low end, 50db dynamic range and at best 50db s/n ratio followed by cut in high end and boosted low end  (aka vinyl) sounds good. It’s a terrible medium but many like what it does and the things necessary to cram it on plastic.  A good cassette is better than that. But we were at the limit of what we could do. Tape is good but it has its own limits. Hello digital.

A DAW might very well add some artifacts or character to audio but people will either like it or not. I’ve done my own tests on a couple of DAWs and can’t measure a difference but maybe just the visual look of the DAW (e.g. Harrison) has a psycho-acoustic effect.  

But at some point wondering is a time sink. Lord I love to listen to the discussions but I can’t give anyone any facts. I’ve done a lot of high fidelity loudspeaker designs but when asked I say “Buy what you like.” 

Nothing else matters.
 

Edited 3 hours ago by Terry Kelley

[Shane_B.]

A very very long time ago there was a thread floating around about the good saturation some cheap interfaces produce. To this day the best sounding interface, to my ears, was my M-Audio Fast Track Ultra. I still have it, but there's no drivers available for it past XP iirc. And the rubber on the knobs has turned to a sticky goo from age.

Clean isn't always better. In fact, my favorite 'clean' old country recordings from the 50's and 60's have a lot of saturation on the vocals and I personally really like that sound. Tape heads and phono cartridges both introduce compression during recording and playback. 

In my world, I try to emulate the mechanical recording methods in the digital realm. Not an easy task, especially when you are not held back by the limitations of tape and analog gear and have the world at your fingertips IRT compression, saturation, etc. etc.. But it sure is fun trying.

[Bapu]

10 hours ago, jayeyehaich64 said:

I believe Bapu did a series of tests with regards to Mixbus back a bit, you should be able to dig those up if you desire.

 

6 hours ago, Starship Krupa said:

Bapu did? How did I miss that? Was it on the old forum or within the past 7 years?

Do you remember what conclusion(s) he came to?

I see that Harrison's pitch is that the sound is "inspired by Harrison's legendary analog console sound & workflow." "Inspired by?"

Like they were thinking about  the hardware product line while designing the DAW?

The first DAW that ever had a console or mixer view was "inspired by" legendary analog consoles. Because why reinvent the wheel?

On the old forum.

IMO the Mixbus mix was "better" but purely subjective. If I did it again today, who knows what I would think/conclude.

[Amberwolf]

9 minutes ago, Shane_B. said:

A very very long time ago there was a thread floating around about the good saturation some cheap interfaces produce. To this day the best sounding interface, to my ears, was my M-Audio Fast Track Ultra.

I still like the "sound" I get from my 1990s GadgetLabs Wave8*24....but I can't use it in my laptop cuz it's a PCI card and external rack box cabled to it with a shielded DB25.  Last driver I know of (user-created) is from around 2014, and works on win10 IIRC. (at least win7). 

 

 

 

 

9 minutes ago, Shane_B. said:

I still have it, but there's no drivers available for it past XP iirc.

You might see if the Avid drivers for their versions of th M-Audio stuff work on it (probably not, but I'm still running my Avid version in Win10). 

 

 

[Gswitz]

3 hours ago, Shane_B. said:

To this day the best sounding interface, to my ears, was my M-Audio Fast Track Ultra. I still have it, but there's no drivers available for it past XP iirc. And the rubber on the knobs has turned to a sticky goo from age. 

I had one of these interfaces and used it for a while. I was struggling at the time to be able to get recordings on the computer without dropouts. 

For years I recorded on a hard discuss.cakewalk.com recorder... akai dps12... then tascam 2488. The spidf from the tascam worked to sync with the m-audio. I would mix in cakewalk. 

I still record by setting and forgetting. I often record for 3 hours without interruption. 

When i mix in cakewalk, I sometimes mix all songs in a single project. Sometimes i create a project for each song after the main recording. 

[OutrageProductions]

11 minutes ago, Gswitz said:

For years I recorded on a hard discuss.cakewalk.com recorder...

Well... there's your problem... gotta be really difficult to record audio directly to a Web Portal for a Forum. LOL!!!