Mixes sound a bit different then using daw

I guess I noticed this for a while using mixbus but I think it’s a situation that can happen in any setup. When I export a mix, I notice that When I import that audio file back in, things do appear slightly different. I guess now I’m thinking it could be the digital to analog conversion, for Anyone notice this as well. I wouldn’t say it’s a issue but more of something to be aware of.

There is no digital to analog conversion between exporting a mix from Ardour (Or Mixbus) and just reimporting. That is entirely in the digital domain. And the mixes will sound identical in any DAW, if they didn’t there is an issue. More likely your issues comes down to either psychoacoustics, or just your mind playing tricks on you.

Psychoacoutics because it will sound different if not listened to at the exact same level, and you could for instance normalize on export which might change overall level. Your mind playing tricks is a bit harder to define but exceedingly common and is part of why I teach my students to make sure they sleep on a final mix before exporting as when they come back the next day it will sound different as their ears are less tired, and this is before exporting.

   Seablade

More to the point: if you are using mixbus, which intentionally not a neutral sound, then importing back into mixbus will never the sound same. If the export was done in mixbus, you will have “twice the mixbus” effect; if the export was done from ardour, you have the “mixbus effect” layered on top.

Mixbus does not aspire or claim to provide a bit-for-bit signal path - it is precisely the subtle effect it has on the sound that makes so many people like it, just like traditional mixing consoles.

1 Like

I understand, but I think I’m on to something, because I’m using the same equipment I used for my mix, after mixbus I exported at the same setting as my session 32bit float and 48khz into presonus studio one. Only thing I did was trim the output by 3 DB to give me some more headroom but I noticed the vocals were a lot more sibalance and harsh then when I was in the mixbus session. And I’m not sure why that was. But I will look into again when I have a chance. But I always kinda had this situation with my last project. I was mixing and when I export I would notice that certain things would need to be readjusted after so I would go back in and just move things up a lil here and there.

When does the digital to analog conversion take place. Is it when I record vocals through my soundcard.

A-to-D conversion takes place when you record a signal that originates in the analog realm (either via a microphone, or from an electric/electronic instrument) and flows through your audio interface.

D-to-A conversion happens when playing back audio through your audio interface to an analog destination (typically loudspeakers/monitors or headphones).

You cannot compare Mixbus with any other DAW. Mixbus deliberately alters the sound without you doing anything. You could get the same results (with quite a lot of work) in another DAW, but out of the box, you should not expect them to sound the same.

You should never continue working on a project by exporting and then reimporting.

I understand. But I was doing a test master and I exported the mix into another daw because I don’t master in mixbus or in my same mix session

I should have been more precise. First of all, by default mixbus will not alter the signal. It will if you play with various settings/parameters, If you use mixbus to “color” your mix (e.g. the tape saturation in the master bus, the compressors etc.), then the result is not comparable to either mixbus out of the box, or any other DAW out of the box.

If you do stem exports from within mixbus, or have left mixbus in its default configuration, it’s reasonable to reimport them into mixbus.

If you have used mixbus to color the sound, then export via the master bus, and reimport into the same instance of mixbus, you will double 'the mixbus effect".

If you have used mixbus to color the sound, then do a stem export, and reimport into another DAW, you will lose a significant part of the color.

I understand mixbus can change a mix, but I export my mix into a different daw so it should sound the same. However others have this same issue and they are using different daws that are not mixbus or ardour. Not sure what’s going on but I’ll try to see if I can figure out why.

Most people who discuss this stuff online don’t know what they are talking about. I would be very wary of getting sucked into this sort of thing.

2 Likes

So after some testing I can verify that it’s generally the same, I forgot about one song not having a deesser on the vocals, I was just doing a test master to see if my mix was in the right direction. however I did notice that some tracks were slightly different when doing a offline bounce vs real-time. So I’ll look more into that. Like the frequency response was slightly different in some areas. But generally is the same, meters may show slight differences in short term or long term LUFS but I can’t tell.
Maybe there is a slight difference when summing a mix session to a stereo file I don’t know. I will make sure I pay more attention moving forward.

That’s quite a level change. Even (or particularly) in the case of small level adjustments (0.5dB or so) you may not perceive it as a change in volume, but instead might interpret it as a change in frequency response owing to the complex way human hearing interprets sound. (For example, it used to be common-place to find a “loudness” switch on some ‘hi-fi’ amplifiers, which was intended to provide a low and high end boost for listening at lower levels, restoring the ‘natural’ frequency balance). Typically this also manifests in the general interpretation of a small increase in output level sounding “better”.
So unless you are comparing like for like, then, unfortunately you cannot always trust what you hear (and part of the skill of a mixing / recording engineer is developing an intuitive sense of how this works - you need to learn your complete listening environment, the DAW, the signal path, your monitors and your room).
You should also make sure you have no sample rate conversion happening (even if not when exporting from Ardour / Mixbus, also when importing into another DAW) as this can noticeably affect the audio in some cases (and might be viewed as analogous to D/A < - > A/D conversion.)

1 Like

I just saw a fabulous takedown video of MQA (the format increasingly used on Tidal), and the producer of that video mentioned that for testing DACs, he uses test gear to level match to within 0.1dB. That’s a fairly good reference.

That is pretty widely considered the appropriate level matching to ensure that any differences heard are not due to level differences.

For those not familiar with dB conversions, the math to direct ratio is:
0.1=20log(V2/V1)
0.005=log(v2/v1)
1.0116=V2/V1
V2=V1*1.0116

So the level of what you are comparing needs to be within 1.1% of the reference (measured as voltage).
So if you are using consumer level (CD standard) 2V RMS max output level, you would have to match the two maximum level signals within 23mV. If you were using professional +4dBu reference levels you would need to match reference level signals within 14mV, or if using something like 20dBu high level signals you would need to match within 90mV.
And the equipment needs to be stable enough to maintain that match (not really an issue if comparing two digital files using the same DAC and output amplifier, but could be an issue if comparing two different analog outputs).

Suffice to say that proper comparison takes a high level of care and attention to detail.