When making your vocals sounds warm, you can amplify your fundamental, use transformer or tube saturation, and expand low frequencies. Then you can introduce some more creative techniques like parallel upward compression only on the lows, or maybe doubling the vocal with a formant shifted version.
Although most of you are probably aware of this, equalizing is still the easiest way to make your vocals warm. Find the fundamental of your vocal which is usually between 100Hz and 300Hz depending on the key of the song and pitch of the vocal performance.
Once you find this fundamental, create a band and center it on that frequency, boost that note, and it’ll cause a warmer vocal.
Transformer and Tube Saturation or distortion is typically warmer, due to the second-order harmonic they generate. This harmonic is one order higher than the fundamental note we just boosted - it’ll strengthen the lowest note of the vocal, the second-lowest note, and amplify the lows causing warmth.
This is why tubes are typically thought of as warm, the same for transformers.
For this trick, you’ll need a linear phase EQ and an upward compressor - first make an auxiliary send or bus and set the level to unity, then insert your EQ on the aux track. Use a low-pass filter to isolate the lows - then introduce upward compression.
This will amplify the quietest parts of the low-frequency range, adding a lot of warmth and body to the vocal. Again, make sure this EQ is linear phase.
If you want warm vocals and need to compress your vocals, use a compressor with an internal side-chain feature, and cut out the lows from causing your compressor to trigger. This means only the high frequencies will cause the compressor to trigger, resulting in less compression overall.
More importantly, your low frequencies will be able to pass through more often than not, resulting in warm vocals.
This next trick is a little strange but has a cool sound - duplicate your vocal and on the duplicated track insert a formant shifting plugin. MAutoPitch is a good free option for this, with which you can reduce the formant settings making the vocal sound lower and warmer.
The blend in the formant shifted vocal in with the original.
If you want a vocal that’s both warm and unique sounding, use a subharmonic generator to create a harmonic below your vocal recording’s fundamental frequency. This will have an unrealistic effect, somewhat like the formant shifting trick, but it can create a vocal with powerful low frequencies nonetheless.
LoAir by waves is a good option since it lets us align the phase of the original and generated signals.
With a dynamic EQ or multi-band dynamic processor we can expand the low frequencies of our vocal to make it sound warmer. Similar to our original EQ, let’s find the fundamental and expand that whenever the vocal is sung - automatic settings on the EQ make this easy.
On the compressor, use set the threshold as you see fit.
Sometimes a vocal may be lacking warmth because its lows are being covered up or masked by your instrumental - if that’s the case, attenuate some frequencies on the lows, like your vocal’s fundamental. For a more advanced version of this, use a dynamic band instead.
For an even more advanced version of this use the vocal as an external side chain, to trigger this dynamic band whenever singing occurs.
Last up let's take that original EQ we used in the first section, amplify that band a bit more, and then use a compressor - in turn emphasizing this frequency range, causing the compressor to work harder on those frequencies. This will achieve a cool tone and a warmer vocal.
If needed, use a de-emphasis EQ after the vocal, to balance this range back out.