Several years ago I had seen the James Grime (YouTube name: singingbanana) video using four bit based binary deBruijn and 16 cards.

The 16 bit deBruijn he used is:
1111 0000 1001 1010 binary bits. He used a special sequence of 16 playing cards such that the red suits fall where the binary "1"s are located and the black cards fall where the binary "0"s are located. The cards he used were/are easy to remember and had their own pattern. He then repeated those 16 cards in the exact same sequence two more times giving a deck of 48 cards.

After a spectator cuts the deck several times and then deals 4 cards sequentially he then asked which cards are red cards? Based on the spectator(s) answer he knows the pattern and thus the identity of all four cards.

I have developed a ternary (trinary) type of deBruijn that uses three different type of backs of playing cards (Rider back, Maiden back, Mandolin back) that very few people will notice the difference and even if they do they may brush it off as a mistake.

During my experiments I noticed that the mandolin back cards can be recognized at a great distance (versus rider or maiden back cards) because the four corners of mandolin back cards do not have bright white figures.

This means that the 16 bit deBruijn used by James Grime could take advantage of this feature especially for magicians that use a memorized stack because the magician could take the first 16 cards (or any 16 card sequence in their stack) and apply those cards to the debruin sequence 1111 0000 1001 1010
the mem deck cards that fall where the "1"s are would be Rider back cards whereas the cards that fall for the "0"s are would be Mandolin backs.

James could also do this on his deck and then not have to ask who has the red cards because on his deck he would have Mandolin cards for the red back cards and regular Rider back cards for the black cards (or vice versa). James would not need a cheat sheet because his scheme directly ties to the patterns.

The mem deck magician would have a cheat sheet for the 16 patterns and their corresponding cards and could easily be on the back of the card case or on an ad card. With only 16 patterns the magician could probably memorize those too.

Also the magician can easily see this pattern from 30 ft away assuming the four cards are held up so the magician can see their backs. I like to use the ruse of having each spectator hold the card with finger and thumb face side against their forehead and concentrate.

One drawback to the multiple back design approach (especially with three duplicate 16 card patterns) is the cost to construct such a deck:
Three rider back decks at $2 each equals about six bucks. Three mandolin back decks at five to $8 each equals let's say about $20. Therefore it'll cost about $26 to construct the 16-bit based deBruijn deck.