In my experience there are actually 3 "generations" of the M50.
M50 (approx. 2007 to approx. 2010): Bassy and V-shaped. Very good.
M50 (approx. 2010 and forwards): Less bassy and V-shaped. Also very good
M50X: Even less bass and very different sound signature. I don't like it.
IMO: try to to find a non-X M50 in decent condition used.
Interesting.
It's been too long since I used the original M50. And never actually compared them side by side. Most of the plots that I've seen of the two headphones make the M50X look more bassy and darkly tilted though. Some examples would be these pairs of plots from Tyll at Inner Fidelity, and also plots from Headphonedotcom.
INNER FIDELITY
M50:
https://www.stereophile.com/images/ifmeasure/AudioTechnicaATHM50B2012.pdf
M50X:
https://www.stereophile.com/images/ifmeasure/AudioTechnicaATHM50x.pdf
HEADPHONEDOTCOM
M50:
https://github.com/jaakkopasanen/AutoEq/tree/master/results/headphonecom/headphonecom_harman_over-ear_2018/Audio-Technica ATH-M50
M50X:
https://github.com/jaakkopasanen/AutoEq/tree/master/results/headphonecom/headphonecom_harman_over-ear_2018/Audio-Technica ATH-M50x
The brightest looking measurement I have seen for the M50X is Oratory's.
https://github.com/jaakkopasanen/AutoEq/tree/master/results/oratory1990/harman_over-ear_2018/Audio-Technica ATH-M50x
^ This is with diffuse field compensation (top solid red curve), and raw (dashed red curve) vs. the DF curve in light gray. And it makes the M50x look like it follows the DF curve in the mids. And is even
brighter than the DF curve in some spots in the treble. Which would suggest it is well to the brighter side of a neutral response there. The brightness in the treble and mids is
somewhat balanced out by the rather sizable shelf in the lower frequencies though. The dip in between the bass and mids is somewhat problematic though.
I've always sort of suspected that this was either a defective unit. Or possibly an older M50 that might've been mislabeled as the X version. Here are apparently three different units though of the M50X measured on Crinacle's GRAS rig. And they all have different responses in the bass, ranging from low to more bassy. The raw response curves are the solid black curves on these...
Crin's single measurement on the Ears rig looks bassier though...
https://github.com/jaakkopasanen/AutoEq/tree/master/results/crinacle/ears-711_harman_over-ear_2018/Audio-Technica ATH-M50x
Some more plots of the left and right channels of the M50X by Rtings...
https://www.rtings.com/headphones/1-5/graph#295/7917
https://www.rtings.com/headphones/1-5/graph#295/7918
Rtings takes their bass measurements in the ears of their staff btw. While they do the midrange and treble on a simulator.
The condition of the pads could be a factor in some of these disparate M50X measurements. Some headphones will start to sound bassier, for example, as the pads become more worn and compressed, and the user's ears begin to get closer to the drivers. That is the sort of thing that could potentially show up in the measurements as well.
It is probably difficult to achieve a reliable consistent seal on the M50X though, because its earpads are so small for an over-ear HP. So that's probably a more likely explanation for why some of the measurements in the bass, and also people's perceptions of the HP's tonal balance can be somewhat all over the map.
The unevenness in its response can also make it seem brighter with some content, and bassier with other kinds of content, depending on the spectral/timbral balance of the recording. Without EQ, it can be a somewhat chameleonic sounding headphone in that respect.
The seal on this headphone is probably important though to achieving adequate bass.