Seperate lossy stems worse in mixdown than lossy conversion of studio mix?

Hi guys.

I've been experimenting with a few stems and found the mixdown to be a bit off. Now, without even CD quality stems to work with I expected this, but the end result just doesn't sound right.

It tends to lack a little treble. But mix is within range and at -1 dB on one. So no obvious sign of distortion.

I've done no adjusting of the levels so I'm leaving it up to the stems to be levelled as they are. Which may be off the studio mix. I only like to adjust when they are too loud when all combined.

If the source is 128Kbps then obviously the mixdown isn't going to be good quality either way. However, for separate tracks, where the frequencies are somewhat reduced, I thought this may work in favour of the stems. Even if there are cross ranges over frequencies.

But, after listening, it sounds like the original in 128Kbps is still better. I'm only testing with 80's and 90's stuff which would have been processed differently in the studio. And a DAW would mix tracks in 64-bit float if it makes any difference. Since the sources are still 16-bit integer.

>And a DAW would mix tracks in 64-bit float if it makes any difference.
No

>Since the sources are still 16-bit integer.
But psycho-lossy codecs like mp3 converts to floating point

>Which may be off the studio mix.
Just use EQ & FX.

Quote:

---

Originally Posted by junh1024

But psycho-lossy codecs like mp3 converts to floating point

---

Good point. I used to convert to WAV 16 bit.

Quote:

---

Which may be off the studio mix.
Just use EQ & FX.

---

I could, but I thought it would be trying to things back in that aren't there any more. Since the original is lost.

Stems often are not mastered, like the final track because it would take more time to do each track. Maybe that's why you hear a difference.

Quote:

---

Originally Posted by djtayz

Stems often are not mastered, like the final track because it would take more time to do each track. Maybe that's why you hear a difference.

---

Yes that would likely account for it as well. I've noticed most stems do have the FX added so rarely ever dry. But given stems are a mix down that would come as no surprise. I prefer them wet even if it can make it hard to cut at breaks with a reverb tail. Some stems, like RB ones, do tend to be 6dB over. Other ones can be too low.

At the end of the day, while mixing and mastering it, you need to remix and remaster it. :-)