Wet vs Dry Stems: What Should You Send?

When you're preparing stems to send to a mix engineer, one of the most common questions is whether to include your effects or strip everything back to the raw signal. The answer is "it depends" - but there's a sensible default, and understanding why helps you make the right call for your project.

What Wet and Dry Mean

"Dry" means the raw recorded or programmed signal with no effects - no reverb, no delay, no chorus, no time-based processing. Just the signal as it was captured or created.

"Wet" means the signal with effects included. This could mean a vocal with the reverb already baked in, or a guitar with the delay tail included, or a synth with the chorus and flanger already applied.

The Default: Send Dry

In most cases, you should send dry stems. Here's why: a mix engineer's job is to place every element in a three-dimensional space - front to back, left to right, top to bottom. Effects like reverb and delay are the primary tools for doing this. If you pre-bake reverb onto a vocal, you've removed the engineer's ability to make decisions about room size, pre-delay, diffusion, and how that reverb interacts with the reverb on the drums. You've committed to a choice they may not agree with.

A dry signal is flexible. A wet signal is a commitment. In most cases, that commitment belongs at the mixing stage, not before it.

When to Send Wet

There are legitimate reasons to send processed audio:

• The effect is part of the sound - a heavily saturated synth bass, a heavily effected vocal that's part of the artistic identity of the track
• The effect can't be recreated - a hardware reverb, a specific pedal, a plugin that the mix engineer doesn't have
• You have a strong creative vision for a specific sound and want it preserved

In these cases, the best approach is to send both versions: a dry signal and a wet reference. Label them clearly. The mix engineer can use the dry signal to work with and reference your wet version for the creative intent.

Bus Processing and Drum Submixes

There's a middle ground that often makes sense: gentle bus processing. A drum bus that's been run through a compressor with tasteful glue compression, or a bass that's been run through a gentle saturation plugin, can legitimately be sent that way - as long as the processing is subtle and serves the sound rather than limiting the engineer's options.

What you should remove before sending: limiters and heavy compression on individual track buses, any processing that significantly reduces dynamic range, and any heavy parallel compression that's already been printed to the signal.

Label Everything Clearly

Whatever you send, label it. "VOCAL_LEAD_dry.wav" and "VOCAL_LEAD_wet_ref.wav" immediately tells the engineer what they're looking at. Ambiguous filenames - "vocal_final_v3_USE_THIS.wav" - waste time at the start of every session.

What JNP Prefers to Receive

Our preference is dry stems as the default, with wet reference files for anything where the effect is genuinely part of the sound. When clients send a full set of dry stems alongside a rough mix for reference, we have everything we need to make independent mixing decisions while still understanding the creative intent. That's the ideal starting point.

If you're preparing a session to send to JNP and you're not sure how to handle a specific element - a heavily processed vocal, a synth that's been run through hardware, a sample that's already been treated - just ask. We'd rather answer a question before the session starts than deal with a mislabelled wet stem halfway through a mix.