Sorry for the lack of details.
I am loading .38 special in a lee classic loader (the kind where you hammer the bullets in).
Here is my process:
Take the new Winchester brass I bought, neck size it.
Flare the case mouths.
Prime the cases with a separate Lee Had Primer tool.
Measure the powder in my RCBS 10-10 scale.
Charge the cases with the powder.
Seat the bullets.
Crimp the case mouths.
I loaded 25 rounds with Hornady XTP 125 gr JHP bullets over 4.2 grains of Titegroup.
I also loaded 25 rounds of Speer Plinker lead 158 gr SWC bullets over 3 grains of Trail Boss.
These are both minimum or near minimum loadings.
Both of these rounds have data specifying a COL of 1.455"
When I measure them with my Frankford Arsenal dial calipers, I get readings between 1.420" and 1.442" or so. I have noticed that the measurements are different depending on the placement of the calipers on the cartridge, how hard I turn the wheel tightening the calipers down on the round, etc.
All of these finished rounds look identical to one another, and each one looks visiualy extremely similar or identical to the factory loaded ammo in similar loadings I have.
The case mouth on the bullets seems to be in the right place; on the crimp channel with a teeny tiny bit of the crimp channel still visible.
Could these differences in length simply be due to different brass sizes, slight bullet inconsistencies, and user error in measuring the rounds? Since the bullet seating mechanism is locked down, I do not see how else the COL of one round could be any different from the one loaded immediately before or after it.
Also, a google search and browse through archived THR and TFL forum posts showed a consensus of "as long as the COL is not extremely off, seat the bullets properly in respect to the case mouth on the crimp channel; this is the correct seating depth for any given bullet."