There is evidence of at least three printers (outside of the 1996 one). The order can also be inferred. Here is a hypothetical timeline assuming the analysis in my report is accurate.
- A Konica Minolta printer used to make photocopies of true, original prototypes (frontside) and 9 cards were used to make copies of the backside (probably fitting on one sheet in a 3x3 orientation).
These trace back to at least April 2024
If beta are recent fabrications, they would have been made potentially starting around this time.
- An inital set of betas were printed (HQ). The printer used added no dots (Epson? Could be a home printer). The backside has traces of the Konica Minolta dots, perhaps a single back sheet from the previous batch was used to photocopy/scan more backs. Notably, it seems that beta Dodrio (I believe it was) uses the same backside as the alphas (a leftover from the first batch?). Also alpha Magneton and at least one alpha Raichu appear to lack dots and share scanning straitions on the front side similar to the HQ beta backs, evidence that the batches are not entirely separate.
HQ beta, I believe, were also the copies that have the earliest public appearance at Card Party (June 14).
At this point subsequent beta-like variants (delta, etc) would have been made
- LQ beta, beta-like variants, more alphas (trainers, popular Pokemon) were printed from the Xerox printer (June 29 based on the copy in the OP). The beta fronts are likely not photocopies but instead printed from the source file. Many of the alphas made are photocopies of photocopies which is why they have double dot patterns - this also implies the person who did it would have access to all those initial photocopies at that point in time. Likewise, the back is also a copy of a copy, which is why it has double dot pattern and why it’s so washed out.
This story is not demonstrated as fact but it does align with all available evidence.
If we want to stray a bit further into speculation, we can make a guess on why three different printers were used. First, a photocopy of all 100+ real prototype fronts were needed, which would be easier to do with a large office-sized photocopy machine like a Konica Minolta. Plus one would assume the quality would be better than what you could do it home. Japan also has far more publicly available resources that are more widely used than we do in the west (such as self-service photocopiers across convenience stores) which should be kept in mind.
Then with betas, a first “test” batch may have been printed from a small, home inkjet printer. Presumably the betas were printed straight from a file so it can be done in a single print job. The backs were a single scan which is also easy to do.
Finally, there are at least 8 LQ beta sets → 1200 cards → 133 front sheets and 133 back sheets. So again, it makes sense to go back to a larger printer to run this job instead of doing it from a home printer. I can’t remember if we’ve seen multiple dates on the dots of these but it’s possible the person went back to the same printer on multiple different days to produce them. EDIT: just checked and looks like some have dots as late as November
Hope that helps.