I found this video interesting, the guy is basically saying that what is touted as "tube warmth" is more a result of the transformers used than the tubes. Is this generally considered to be true or is he talking BS? I really know little about tubes or transformers. 
 

only your ears, with your gear, should be the judge.
 
 
tubes offer distortion,
when applied correctly,
can add harmonics to the sound that some folks really dig.
 
transformers,
is where the buck stops.
 

I have an old Carvin transistor amp from the early 70's that used an output transformer and it sounded pretty awesome. So there may be some validity to his thesis. 

it's not really the same animal we are talking about here....
 
every amp has a output transformer....
 
we're talking recording gear here,
not tube guitar amps.

It isn't an either or.  Both tubes and transformers affect the sound.  There is something special about tubes, either live or for recording.  And transformers, too, the way they round and soften a sound passing through them.  And something funny, most of the tube stuff, esp. the older stuff, is full of transformers too, so you get a double-heapful helping of goodness
 
I'll say one thing, solid state transformer gear can sound just as smooth as tube stuff and usually costs less, too.
 
@

Yes... Tubes are only part of the equation.  Since tube amps always need a transformer on the output, that poor transformer is often overlooked as contributing to the overall sound.
 
Just as saturating the plates affect the sound, so too saturating the magnetic core affects the sound.  Wire diameter of the windings,  number of turns in the coils, size of the core, all play a part in sculpting the sound whether it's a guitar amp or some other piece of "studio gear" with tubes.
 
Just as sound quality can change from one set of tubes to the next...even with the same model number tube.... swap out the transformer in a repair job with another similarly rated transformer because the factory direct transformer is either not available or too costly, and you will very likely notice a difference in tone, quality and overall performance. The amp that was once crisp might become a bit spongy or vice versa. Someone familiar with the amp's sound will notice a slight difference in that amp.

I laughed out loud at that video, not because it was BS but because it was funny. Tubes for warmth - just grab one with your fingers if you don't believe they get warm!
 
My problem is with the word "warmth". Where does that word come from? Like so many audio metaphors, it's borrowed from a visual reference. When a photographer says a print needs to be "warmer", he means it needs more red, red being the closest visible color to infrared, which is heat.
 
It means longer wavelengths, which should by analogous to bass frequencies in audio. But when we talk about the effect tubes and transformers have on audio, we are talking about harmonic distortion - adding high frequencies. This is why "warmth" has never made sense to me in this context.
 
But I do understand what people mean when they use the term. In fact, adding harmonics to low frequencies actually makes those low frequencies more perceptible. If you have Ozone, you can easily hear this effect by creating a band below 120 Hz in the Exciter module (Tube mode) and start turning it up. You'll hear the bass guitar, kick drum and floor tom start to sound deeper and fuller, even though you're not adding more bass to the mix. Waves' RBass and Maxxbass plugins work on the same principle.
 
I can see how people might confuse tubes and transformers, because they usually go together. Tube gear nearly always has transformers. Solid-state gear generally does not (power transformers don't count), unless they've been added specifically for tonal reasons, such as in a microphone preamp or inside a condenser microphone. So if the tube gear sounds richer, which component is responsible, the tubes or the transformers?
 
The answer is "both". Half the harmonic goodness you get out of a tube guitar amplifier is from the high-gain triode stage at the front, the other half from the push-pull circuit at the output that incorporates a transformer. Some of that latter half is from the pentodes' nonlinearity and some is from transformer saturation. The triode gives you odd-order harmonics, while the transformer gives you even-order harmonics. Together, you get the whole set. The same thing a sawtooth waveform gives you in a synthesizer, btw, which is why that waveform is so popular in subtractive synthesis.
 
It's silly to argue whether tubes or transformers are more important. Odd-order harmonics are easier to hear, so they have a more profound impact. But transformers, in addition to supplying even-order harmonics, also act as low-pass filters, which softens the effect of the tubes that precede them. 

 
In my limited view of things, I associate warmth with soft clipping, even-order harmonics (related), and less high-end (but that is perhaps less high-end odd-order harmonics).
 
I have this purely intuitive feeling that a transformer on the output would tend to even things out, round things over, ie. generate more 'warmth' from all the electromagnetic fields interacting.
 
 
 
 
 

EVERY device which sinks, sources, or passes current will impart an effect onto the audio. Most of these effects are considered non-linearities, and a lot of engineers spent a lot of time trying to minimize them.
 
Tubes are active devices, while transformers are passive devices. That alone suggests that whatever personality they impart on the signal is going to be different. And they are<G>...
 
Tubes can - they don't have to - generate harmonics. If they are biased as Class A/B or B they can also create some non-musically related components, the result of a nasty spike that can occur when the signal crosses the bias point.
 
Transformers can - they don't have to - attenuate lower frequency energy, and when they saturate (again they don't have to) the start to slow rise times and compress the dynamic range a little.
 
I am all for as many tools as possible to scuplt the sound, and I am a big fan of understanding why they do whatever they do, but I'm also pragmatic enough to realize that if I spent all my time making measurements I wouldn't have time left to make music<G>!

Not only is it good to know the fact that one is active and the other passive and also the idea that both can be used in many capacities but transformers do exist on their own, but also the fact that there are different ways of using these components.
 
There what might be called Hi Fi mode where valves and transformers are set up for minimum distortion. eg Williamson class A valve power amplifier.  (Classic situation of valves and transformers set up for the most perfect sound or valve RIAA preamps)
 
Then there is distortion mode and lots of it and people such as Randall Smith from Mesa Boogie fame will know a lot about how to put valves and transformers into those non linearities to make guitars sound better.
 
Then of course there are some modes midway between the two for Mic preamps and compressors say that are basically clean but verging on some slight controlled crunch.
 
All interesting uses for them both but have an idea what mode you are into perhaps. And why you are passing signals through them.
 
Makes you realise how great digital is because it has the ability to totally bypass them and let you hear how things actually sound. Better in most cases. And not only that but digital is emulating the sound of those things rather well too.
 
In the very early days when the signal path was though valve preamps and transformers into valve reel to reel recorders and then back into more transformers and valves then into our record groove it could be said there is a lot of accumulation there of analog sound. No wonder we got used to what we heard then.
 
The mistake might be to make it the bar that everything has to aspire to and pass. It is not necessarily right. We are hearing things differently now and we are better for it. Signal chains with and without valves and transformers are just different, not so much one better than the other.
 
Here is a case in point. When a class A valve amplifier drives a Quad electrostatic loud speaker there are two transformers involved. The output transformer of the amp and the step up transformer inside the Quad electrostatic. I was lucky to hear one day a modification where the valve amp was made transformerless and also the step up transformer in the Quad was also bypassed so the valve amp output stage was directly driving the plates of the Quad speaker! I will never forget that sound. Breathless was a word that comes to mind. Showed up the fact that transformers are actually not much good at all in that situation