[Help Thread] Mirror Cube Discussion

[Cristobal]

I have started recently to (try to) solve the mirror cube. In that purpose I have watched and collected some algorithms from YT. For now, I managed to finish a few solves with these algs, but it is often a hit or miss:

When it comes to place the corners, then it becomes rather confusing and the alg (the 3rd in the picture) does not work every time; so I would like to know whether something is wrong.

 

[AlphaSheep]

That alg is called Niklas and it cycles 3 corners. When you get to that step, there are 4 possible situations:

1. All 4 corners are in the correct spot: skip the step
2. Just 1 corner is in the correct spot, the remaining three corners need to cycled counter-clockwise: Perform the alg once to place the corners.
3. Just 1 corner is in the correct spot, the remaining three corners need to cycled clockwise: Perform the alg twice
4. No corners are in the correct spot, do the alg once to place one corner, then rotate and do it once or twice more to place the remaining corners.

In the long term, you're better off thinking about how your normal 3x3 algs apply to mirror cube than wasting too much time practising this method.

 

[Cristobal]

Thanks.

It's because the beginner's method I use did not work that I came to these algs.

Here is my personal memento for beginner's method (on the left side) whose the algs did not seem to work on the mirror cube.

 

[AlphaSheep]

Thanks.

It's because the beginner's method I use did not work that I came to these algs.

Here is my personal memento for beginner's method (on the left side) whose the algs did not seem to work on the mirror cube.

These beginner algs are much better than the ones in your first post, because they do OLL first, then PLL, which is easier to recognise than PLL first.

A mirror cube is just a normal 3x3 with heights instead of colours, so all algs that work on a normal 3x3 also work on a mirror cube. There are no exceptions. The most likely reason they aren't working for you is that you're mistaking where pieces need to go.

I don't know which side you're solving first, but I prefer to start with my cross on the thickest side. That way, the last layer is on the thinnest side, so recognising OLL cases is much easier. For each piece, look where the thinnest side is pointing, and match that with the yellow stickers on your cheat sheet.

With PLL, the concept is the same. For the corners, you're looking for headlights (two corners where the sides of the corners on the same face line up. Place these at the back and then perform the A-perm. If there are headlights on all sides, then the corners are solved, and if there are none (on each side, the two corners are different thicknesses), then you need to o the E-perm. Edges are a little trickier to recognise. I find it best to move the U layer until the corners are solved, then look at whether each edge can be solved by a U, U2 or U'. If all four edges are unsolved then if all four edges need a U2 to solve them, then it's an H-perm. If the edges need Us and U's to be solved, then it's a Z perm. You can use a similar idea to distinguish between the Ua-perm and Ub-perm.

 

[Cristobal]

OK. Intuitively I recognise the OLL patterns, so I use the sune, but then it get confusing when permuting corner and edge. I will carefully study your instructions. Thanks again.

 

[SenorJuan]

I personally think "Edges then Corners" (like Cristobal's first post) is an OK way to solve the LL of Mirrorblocks. The edge-solving step only has 13 (?) cases, and you can AUF if needed to help with recognition. Having the edges solved makes it more obvious what needs doing to the corners. Full L4C does involve plenty of algs, but thankfully it can be simplified in various 2-step ways.

 

[Cristobal]

I personally think "Edges then Corners" (like Cristobal's first post) is an OK way to solve the LL of Mirrorblocks. The edge-solving step only has 13 (?) cases, and you can AUF if needed to help with recognition. Having the edges solved makes it more obvious what needs doing to the corners. Full L4C does involve plenty of algs, but thankfully it can be simplified in various 2-step ways.

Could you develop please.

 

[Sue Doenim]

Could you develop please.

AUF stands for adjust U face, so just turn the top layer. You could do L4C in 2 looks by doing permute all with Niklas or F (R U R' U')3 F' then solve CO using one of these algs.

 

[Cristobal]

AUF stands for adjust U face, so just turn the top layer. You could do L4C in 2 looks by doing permute all with Niklas or F (R U R' U')3 F'

then solve CO using one of these algs.

That's a lot of algs...
https://www.speedsolving.com/wiki/index.php/Corner_Orientation

Aren't these algs simple enough?
L4C with a beginner method
Another approach is a beginner's method of last layer, permutation first:
U R U' L' U R' U' L - permute corners except front-right-up corner
R' D' R D - repeat - rotates corners. Must rotate only up face to switch which corner is rotated
https://www.speedsolving.com/wiki/index.php/L4C

 

[SenorJuan]

Regarding the L4C: obviously there's the "Permute the Orient", which is essentially a fleshed-out version of your first-post example. The step-3 algorithm you posted has a inverse which cycles in the opposite direction, you'll find a combination of these two can sort out any permutation case. And Susie's previous post links to Orientation algs ...
I think an intuitive beginners choice is:

F D' F' R' D' R ( U or U2 ) R' D R F D F' ( U' etc)

Note that you don't need to learn the inverse of this alg, just rotate the cube: y2 z

The two-look simplifications of L4C are mentioned on the Wiki page:

https://www.speedsolving.com/wiki/index.php/L4C and
https://www.speedsolving.com/wiki/index.php/L3C

... they're OK for the 36 awkward cases where one corner is in the right place but not oriented, and the remaining 3 corners need cycling. If you know all 18 3-corner cycle algorithms ( 9 and their inverses) you can use a combination of two of them to (1) solve one corner; (2) solve the remaining 3 corners with one of nine 3-corner cycles:
Simple example: On a solved cube, do:

L' U R U' L U R' U' y2 L' U R U' L U R' U'

then solve with:

U R U' L' U R' U' L y2 U R U' L' U R' U' L

the first step solves the Back-Right corner whilst disturbing the F-L corner, second step will be one of the 9 counter-clockwise L3C cycles.
There are other variations: but the first step has to move the in-position corner, whilst solving another corner. I'm guessing this is too advanced for you at the moment.

I really like Mirrorblocks, but I'm no expert, so please read other people's replies.
And you have to try it 'unsighted', solving by tactile means alone.

 

[Cristobal]

OK, so I finally gather easy enough algs in a quite simple method that suits me. It is likely not the fastest (neither the smartest), but it requires only 5-7 algs to learn and the progression of the solve is rather linear.

The idea is to get one correct corner and use A-Perm to place the other ones in the correct spot. Sometimes, A-Perm cannot be performed, so you just have to repeat the alg to get an A-Perm case (3 corners that can rotate clockwise or counter-clockwise). Sometimes, you get lucky and can skip the A-Perm step (or get a 2 look OLL, if you know them).

As for the 90 degrees centres rotation, sometimes only opposite centres to rotate occur, so you just have to do the alg, which subsequently will create intermediary centres to rotate on other sides until you get adjacent sides with centre to rotate.

Here is the memento I created to store in my smartphone:

 

[AlphaSheep]

To turn the front and left centres 90 degrees:
M' U M E M' U' M E'
and to turn the front and back centres 90 degrees:
M' U M E2 M' U' M E2
Notice the only difference is the quarter turn of the E slice for turning adjacent centres, and the half turn for turning opposite centres.

 

[Cristobal]

Thanks, I'll look into it.

 

[Cristobal]

To turn the front and left centres 90 degrees:
M' U M E M' U' M E'
and to turn the front and back centres 90 degrees:
M' U M E2 M' U' M E2
Notice the only difference is the quarter turn of the E slice for turning adjacent centres, and the half turn for turning opposite centres.

Those algs are brilliant! Thanks again.

 

[Cristobal]

I have updated my memento:

 

[Xxcookid123xX]

So a few days ago i went to this discarded amazon items dump where everything was 1$ and i managed to find a fisher mirror cube in there, and now i'm trying to solve it.

Unfortunately I cannot find a single tutorial on this cube, and the ones i do find aren't any help at all.
I've been trying to solve it manually by logically moving pieces, but i always get stuck where the final layer has flipped pieces:

Sorry for the weird way the photo is taken, i had to take the picture with a laptop (front camera)

So how can I fix this? And if it's not fixable, does anyone know any tutorial to solve it normally?

 

[GenTheSnail]

With fisher cubes, sometimes if you put in a middle edge flipped (which you won't be able to tell until later) you'll get an impossible LL case. However, it doesn't look like that's happened here.
Adjust the U face so that the two flipped edges are in front of you and to the right. Do F R U R' U' F'. I don't know how the corners will turn out, but by doing combinations of R U R' U R U2 R' U2 R' U' R U' R' U2 R (which twists both corners on the right side) it should get solved.
The pieces might also need to be permuted; if they do, give us another picture (your laptop camera is fine) later and we can figure it out.

 

[MusiCuber]

Replying @GenTheThief
I got this cube as well... there is no pll on it.
And if you should get that 1 edge flipped case... You can just take 1 edge of your f2l pairs and flip it that will solve the impossibility.

 

[parsa419]

Hello everyone who is reading and willing to answer my question. I have a very rare case on my cube (mirror) that i have no idea how to do. This is what it kind of looks like:
O = placed correctly
Q=not placed correctly
*view from last layer
O - O - O
Q - O - Q
O - O - O

If you could just visualise it. I need to swap the 2 Qs without breaking the bottom. Ty if you answer.

 

[Thom S.]

In theory this is not possible. It is very possible that you didn't actually buy a real mirror cube, but instead an inequal cube in which this is possible as there are identical pieces