What Are Stems? (Professional Audio Engineering Guide)
- Scott Hannon
- 2 days ago
- 4 min read

šļø The Classical Definition of Stems
Stems are submixes of grouped tracks exported as audio files that can be recombined later to recreate the mix.
In the traditional film and music world, stems were typically grouped like:
Drums
Bass
Music
Vocals
FX
š Think of stems as mini-mixes inside the full mix.
Key rule (classical definition):
When all stems are imported into a new session at unity gain (0 dB) and aligned at the same start time, they should reconstruct the approved mix.
Thatās the gold standard.
š§ Why Stems Exist (Historically and Practically)
You nailed the big reason.
Before universal plug-in compatibility, studios needed a way to:
Move projects between different consoles
Work across different DAWs
Deliver to film/TV post
Allow remixing without giving raw multitracks
Protect proprietary sounds
The big problem:
Different studios did NOT share:
The same outboard gear
The same console
The same plug-ins
The same automation systems
So the solution was:
ā Print the sound
ā Commit the processing
ā Deliver audio that recreates the mix anywhere
š Stems vs Multitracks (Critical Distinction)
Many people confuse these.
Multitracks (raw tracks)
Example:
Kick
Snare top
Snare bottom
Hi-hat
Tom 1
Tom 2
etc.
š These allow a full remix from scratch
Stems (submixes)
Example:
Drum stem (all drums summed)
Guitar stem (all guitars summed)
Vocal stem (all vocals summed)
š These allow controlled remixing
Quick rule of thumb:
Multitracks = ingredients
Stems = prepared dishes
šļø How Engineers Actually Print Stems
This is where things get interesting ā and messy in the real world.
There is NO universal standard, and practices vary widely between:
Music mixers
Film mixers
Mastering engineers
EDM producers
Hip-hop producers
Live engineers

Letās break down the common approaches.
š¹ Type 1 ā Fully Processed Stems (Most Common in Music)
These include:
EQ
Compression
saturation
inserts
bus processing (sometimes)
Goal:
Recreate the mix exactly.
Example:
Drum Stem includes:
Channel EQs
Channel compression
Drum bus compression
Parallel compression (sometimes printed in)
Saturation
Console emulation
ā Pros
Mix translates exactly
Safe for clients
Safe across studios
No missing plug-ins problem
ā ļø Cons
Less flexible
Harder to remix radically
Processing is ābaked inā
š¹ Type 2 ā Dry Stems (More Common for Remixers)
These are printed with minimal or no processing.
Usually includes:
basic editing
clip gain
maybe corrective EQ
Goal:
Maximum flexibility for the receiving mixer.
ā Pros
Full creative control
Easy to reshape
Preferred for remixes
ā ļø Cons
Does NOT recreate original mix
Can sound flat
Requires more work downstream
š¹ Type 3 ā Hybrid Stems (Very Professional Approach)
This is what many high-end mixers (and smart engineers š) do.
They deliver:
A) Wet stems (processed)
AND
B) Dry stems (optional)
Why?
Because different clients need different things.
For example:
Film dub stage ā wants wet
Remixer ā wants dry
Mastering ā wants wet
Archive ā wants both
š¹ Type 4 ā FX Printed Separately
This is VERY common in pro workflows.
Instead of baking reverb/delay into the stem, engineers export:
Dry vocal stem
Vocal FX stem
Delay throw stem
Reverb return stem
š¤ Example Vocal Delivery
Option A ā Fully Wet
Lead Vocal Stem (with verb/delay printed)
Option B ā Split FX (preferred in many pro rooms)
Lead Vocal Dry
Vocal Reverb
Vocal Delay
Vocal Throws
ā Why split FX?
Because the receiving mixer can:
Adjust space for film
Fold to mono safely
Change ambience
Fit into surround mixes
Manage dialogue clarity
Film and TV people LOVE separated FX.
š¹ Type 5 ā Bus Processing Printed vs Not Printed
This is one of the biggest debates between engineers.
Scenario A ā Bus processing included
Drum stem includes:
drum bus comp
saturation
tape
console glue
š recreates mix exactly
Scenario B ā Bus processing excluded
Engineer prints:
individual drum processing
but NOT the drum bus glue
š gives flexibility
Professional compromise
Many mixers deliver:
Stem with bus processing
AND
āNo bus compā version
šļø Master Bus Processing ā The Big Question
This is where people REALLY mess things up.
Rule (generally accepted):
š« Do NOT print master limiter on stems
š« Be careful with mix bus compression
š« Avoid clipping the stems
Best practice
Deliver:
Stems without master limiter
Mix print with limiter
Optional: stems through mix bus comp (clearly labeled)
š¦ Professional Stem Delivery Specs
Most pro deliveries follow these rules:
Technical
All files start at 00:00:00
Same sample rate
Same bit depth
No normalization
No clipping
Peak headroom preserved
Typical Format
24-bit WAV
48kHz (film)
44.1kHz (music)
Interleaved stereo (or mono where appropriate)
šļø Real-World Example Stem Package
Music Delivery
01 Drums Stem
02 Bass Stem
03 Guitars Stem
04 Keys Stem
05 Lead Vocal Stem
06 BG Vocal Stem
07 FX Stem
08 Print Mix
09 Instrumental
10 TV Mix
Film/TV Delivery (more granular
DX Stem (dialogue)
MX Stem (music)
FX Stem (effects)
VO Stem
Music Submixes
Reverb Returns
Printmaster
šØ Common Stem Mistakes (Youāll Appreciate These)
Scott ā these happen ALL the time.
ā Stems donāt null with mix
ā Hidden mix bus processing
ā Time misalignment
ā Missing automation
ā Printed twice through limiter
ā Parallel buses missing
ā FX returns forgotten
ā Phase issues between stems
š§Ŗ The Professional Null Test
The gold-standard check:
Import all stems
Set faders to unity
Route to stereo bus
Flip polarity vs mix
š Should nearly cancel.
If not ā something is wrong.
šÆ When to Choose Each Stem Style
Choose WET when:
Client wants mix recreation
Mastering delivery
Film delivery
Archival
Label delivery
Choose DRY when:
Remixers
Sample packs
Producer swaps
Sound design
Collab production
Choose HYBRID when:
You want to look like a pro š
You donāt know the downstream use
High-budget clients
Sync licensing
š„ Pro Tip for Your EchoCraft Workflow
Given your producer + engineer + collaborator world:
Best default delivery:
ā Processed stems
ā FX separated
ā No master limiter
ā Headroom preserved
ā Optional dry pack

Comments