Bitcoin Forum

hello there

i have my own bitcoin address and public key and a WIF recognized as invalid

this txt file was old and recovered from my old HDD

i assume that the wif has more than 19 missing characters at known positions


 is there is any chances that i can recover it ??

any guiding help will be appreciated :)

19 characters gone/lost already? Sorry, you will not be able to recover the WIF private key.

Alright , thank you i thought that too

what is the maximum WIF missing characters that can be recovered  ?

19 missing characters is a big chunk of key space. You're looking at having to brute-force around 40% (give or take a few percent, depending on exact locations) of the key. Searching this ~100 bit space is pretty much hopeless.

Sadly, as Charles-Tim said, you will not be able to recover it.

Sorry for your loss, take better care of your private key next time.


There's no simple answer to this one, it depends on their location. If they're randomly distributed, then I would hazard a guess that the upper limit is somewhere around 10 missing characters (would still take a lot of compute power and time though).

are you saying the hard drive failed and you could only get part of the text file that's how you lost the 19 characters because otherwise why are there 19 missing character?

Pretty bad you have 19 missing characters from your WIF key it takes years decades before you can able to recover it and there is no maximum number of missing characters that you can recover 1 to 3 missing characters can be easily recoverable but above 3 it takes a few months or years.

Why not try to recover it there is an easy tool to use for recovering base58 encoded strings such as WIF keys.

Check this link The FinderOuter, a bitcoin recovery tool (https://bitcointalk.org/index.php?topic=5214021.0)

Are you confusing WIFs with mnemonics?
In a mnemonic more than 3 missing words starts becoming very hard to recover but in a Base58 WIF it is still very easy to recover keys missing 5-6 characters and anything higher than that will start taking hours and exponentially become harder.

First at all: 19 is a lot.
How did it happen? What do you have? Text file? How are you sure the rest of characters are correctly recovered? Do you miss beginning or end?
If there would be less characters, but maybe randomly distributed, I would say my WifSolver could help: https://github.com/PawelGorny/WifSolver
Because you have many characters missing, if you lost the beginning you may use much faster version, build for GPU: https://github.com/PawelGorny/WifSolverCuda
That program would not work if you have your characters missing at the end. Then you should try to convert problem into simple brute-force tool like BitCrack.
There is also 4th option. If you know public key (address has outgoing transaction) we may use Kangaroo, normal or modified for solving WIFs (https://github.com/PawelGorny/Kangaroo)
Let me know if you need hep with any of tools, we will try to find the best approach. But honestly speaking, as you sad you are sure 19 characters are missing, the only chance to restore it in reasonable time is to have public key and attack with Kangaroo.

In a German coin forum a user has proven to be able to recover up to 20 missing characters from the end of a compressed WIF key in a rather short amount of time, but he always also needed the public key of the private key for fast success and somehow it's also restricted to P2PKH legacy addresses (he and his business friends seem to have a big GPU farm at disposal and some self-made software tools for such number or recovery crunching, no particular details regarding software tools and amount of GPUs used were disclosed). Uncompressed keys should work, too, but his claims were verified with compressed keys only.

He offered that as a recovery service. I had many doubts, especially regarding compute time. Multiple challenges have been setup for this guy starting from 18-20 missing characters at the end of the WIF key which he solved in a shockingly low amount of time.
I was really impressed.

So if @Turing0x's case meets the constraints and he's desperate enough to give it a try. Disclaimer: I don't know the user from the German coin forum personally nor can I say anything how trustworthy he and his service offering actually in a real recovery case is when real coins are at stake. All I can say: multiple challenges with 18-20 missing characters from the end of a WIF key have been perfectly recovered. No errors in any of his solutions for the challenges.

If @Turing0x is interested, I can try to contact that user and assist to establish communication. I have no business relation with that user.

He just used one of BSGS or Kangaroo programs. By default they work on specified range of private keys, which is easy to find having lost end of WIF. As I wrote in my post there is Kangaroo version modified by me which allows search in more complicated cases, when missing characters are in the middle or in the beginning.
The fact if key is compressed or not, does not matter.

This is contradictory which raises suspicions. If you have the public key then you don't need the address regardless of its type, since any address type can be derived from the public key.

Or lack of knowledge how to verify if given public key creates desired 3… address etc.

On the other hand I do not think there are many people who created privkey backup for a single address other than legacy, as they are usually created already in “seed era”.

I don't know why a public address would be needed at all and I know that you can derive any address type from the public key. But it's not know what kind of software they use/have modified/tweaked as their recovery tool. As I'm not active in this private key finding game, I don't know the algorithms involved and why this and that might be needed.
My knowledge is limited to some cornerstones, like (I might be wrong should I missunderstood some Bitcoin details):
* a private key and the assoc. public key likely have a unique and singular relation
* there are likely ~2⁹⁶ public addresses for every public key, simply due to the fact that the hash space is reduced by 96 bits when you take RIPEMD160(SHA256(public key))

Anyway, I have no idea why that GPU farm recovery tool would have to fiddle with a public address at all. The main playground with missing characters aka bits of a private key and its public key should be the elliptic curve stuff. For a known public key any derived address type is known and fixed, there's no ambiguity on the path public key --> public address types.
Only reason that comes up for me is that a recovery client might not know the public key as you usually don't have to deal with it. The recovery constraints for this case here imply address reuse as the public key must have been exposed on the blockchain by a previously spent UTXO from that address.

Only assuming? There's a chance that it's not a WIF but a prvKey in different format.
See if there's a similar format on the prvKey formats listed in this thread: [overview] Recover Bitcoin from any old storage format (https://bitcointalk.org/index.php?topic=4959742.0)

Thank you really for helping out :)

Now i have another thing despite from the wif damaged

i got the dumped wallet and it contains a secret

can i recover my wallet throght the secret ?

I'm going to assume the wallet is a Bitcoin Core wallet.

It is only possible if the wallet is not encrypted, or if the wallet is encrypted but you remember the password. If the wallet is encrypted it's probably more (theoretically) faster to just brute-force the WIF.

You need to be more specific without disclosing sensitive information of your wallet like private keys and such.

What kind of wallet is this?
What kind of dump of the wallet do you have?

If it is a Bitcoin Core HD wallet and you see as "secret" the private key that is used as seed for the HD wallet, then I think it should be possible to recreate your wallet's keys with that.
If you have a Bitcoin Core wallet dump with lots of private keys in the dump then you can import those private keys e.g. in an Electrum wallet (in my opinion the import is easier with Electrum than with Bitcoin Core itself) and see if you find any of your coins.

It's quite some time ago that I did some experiments to recover a Bitcoin Core HD wallet by using the dumped HD key from the donor wallet dump, but I don't remember the results unfortunately, sorry. Only I remember is that I had some trouble, but that might have been due to lack of knowledge with proper command usage for Bitcoin Core console or bitcoin-cli.

to be more specific with you it's not a wallet dump , my few wallets was lost in and the HDD is damaged , however  those are some of the backups in text format i made before in my 2nd HDD .

the format in the txt files follows this :

"addr": " My Bitcoin Address "
"compressed": false,
"encrypted_privkey": " Privatekey "
"hexsec": " dfgdfgfdffhfg" ex.
"label": "",
"pubkey": Pubkey ex: "0000000046556641541"
"reserve": 0,
"sec"
 "secret": " Contains a secret hash

All those Values are given so i have the enc_privkey , hexsec , pubkey , sec ( which is a full privatekey but it refers to another empty wallet address ) , and finally the secret .


Any Help will be appreciated :)

yes you 're right , but the wif brute forcing will take too much time due to it's more than 19 missing characters at known positions .

i have tried BSGS and Pollard Kangaroo with no success , it takes like forever

i have tried also Wif-Solver Cuda and Calculated the stride & Range which is too big and after running it for few hours it still at 0.000 % .

so i ran out of ideas :(

Could you write what are positions of missing characters? At the end, at the beginning, in the middle? Is it one group or there are several groups?

If it is one group and as you seem to have public key (is it indeed correct?), it should be doable with Kangaroo in a very reasonable time.

Ok , i will give you an exact example of the positions of missing character


this is just a WIF as an example

5HpHagT65TZzG1PH3CSu63k8DbpvD8s5ip4nEB3kEsreAnchuDf

The Missing Characters ( Replaced By * ) Are as Follows :

5HpHagT65TZzG1P*********************EB3kEsreAnchuDf

I Have converted my public key to an address and the output is the Address of my wallet , ( My wallet has no outgoing Transactions ) , so i got the public key from the info of the wallet that i found earlier in the txt file .


I have some questions and i hope that you can give me a good explanation or some help if you can


- can i determine what is the range of the bitcoin address to search for it in a specific range , i know that the range could be determined in Privatekeys but i am just wondering if something like this exists .

- is there is some useful methods to try to recover my wallet instead of bitcrack ?

- if the bitcoin address contains 0.054 BTC , will it be a 54 bit  2^54 ? ( sorry if it's a dump question , but i have limited knowledge ) .

- can i split the range that i will be searching for into 2 parts , for example let's say that i will search for a specific range for 7 days , how can i split this search into 2 devices ?

i have seen before that the ckeys of the wallet can be cracked by some software , but i don't remember exactly what it is .

it is not the best example, because you used WIF from private key=1. Normally you calculate range as follows: replace unknown characters by "1", decode, it will be your starting range. Then add 58^unknown characters (in your case 21?), if you add to starting range you will have the end range for search.

Yes, that's what we try to do with Kangaroo

No, it has nothing in common.


doable.

i have replaced the unknown characters by 1 and got the start range , then i replaced again with z and got the end range , is that right ?

how can i add 58^21 into my starting range ?

No, that’s not correct. I mean, it is, technically, but it is not the best way to solve WIF problem. You know that you look for a private keys with a given stride, which is “58^number of known characters to the end”, in your case 58^15.

Thank you very much , i will try to prepare the necessary configurations and will try to solve it .

can the ckeys in the wallet.dat files cracked through the master key , i saw something like this before but i can't remember what it was .

The problem with your WIF is that stride collides with checksum and you may have improperly calculated private keys, tomorrow I will launch my calculator to see how to solve it, usually it may be solved by launching several works with a different starting points. Just confirm how many characters you know l, from the gap to the end.

Keep in mind that this method is only accurate when the missing characters are all at the end since the farther away they get from the end, the bigger the range is going to be compared to the actual range of missing characters. Take the following private key for example (to simplify things I didn't remove checksum):
If the last character is replaced by 1 and then incremented by 1 (...19v1 to ...19v2) the integer values are the following:
The difference between these two numbers is 100% accurate and equal to 1 (we incremented by 1)

Now lets say the missing character is the second from the right, and we repeat the same thing (...191s to 192s):
The difference is now less accurate and equal to 58.

Now lets take a random character in the middle and repeat the same thing again (..UoS1ejn8... to ...UoS2ejn8...):
Now the difference is huge and equal to 3199866632452173458088315935260672 while it was only 1 character and we only incremented the character by 1.

I know, but I have prepared a version of Kangaroo which works with given stride - just to be able to solve WIFs, preferably with missing beginning.
https://bitcointalk.org/index.php?topic=5315607.msg56298967#msg56298967

Assuming you manage to tweak Pollard Kangaroo (don't know if it's possible, looks kinda tricky), here are the expected run times and cost (on-demand, spot instance could be 1/3rd of the price) when using AWS GPU instance p3.8xlarge (4x V100):
MC - missing characters, bits - equivalent missing bits

|	MC	|	bits	|	ETA	|	$	|
|	14	|	82	|	00:11:17	|	2	|
|	15	|	88	|	01:25:53	|	18	|
|	16	|	94	|	10:54:07	|	133	|
|	17	|	100	|	3d 11:01:37	|	1,016	|
|	18	|	105	|	26d 08:18:52	|	7,740	|
|	19	|	111	|	200 days	|	58,942	|
|	20	|	117	|	4 years 68 days	|	448,893	|
|	21	|	123	|	32 years	|	3,418,664	|
|	22	|	129	|	242 years	|	26,035,770	|
|	23	|	135	|	1849 years	|	198,282,544	|
|	24	|	141	|	14083 years	|	1,510,074,752	|

This was using my software (not published), Jean_Luc should be faster (50%?), newer hardware should be faster as well.

well , it seems like i had no luck in my recovery process because my characters is more than 21 and the searching range is so big ,

is there is any other thing i can do , like lattice attack on my outgoing txns , or something .

i have the public key which computes right to my address , and have everything except for the private key .

Any Help will be appreciated , i am out of luck as you see and hope that i can find something .


also i did not ever shared my wallet with anyone , so no one have it except me ( the owner ) .

If it is uncompressed key, it complicates thing, as we cannot calcululate "the path" for additions. If we would have compressed key, we may base on compression flag encoded in WIF, to find what should be the starting points for Kanagaroo ranges. With uncompressed, it is much more complicated, as we do not know how to find stride (and from which key) to calculate keys which would not impact last bytes of decoded WIF.

In your case it depends on the public key  :)

Why public key?

public key -> address -> amount of BTC


if it contains 0.054 BTC --> not doable  :)
if it contains 5.4 BTC --> doable

i have both uncompressed and compressed

how can do it in the compressed public key that is starting with 02 ( even ) or 03 ( odd ) ??

any info. and help needed please :)

Your BTC address, was it created as compressed or uncompressed?

i have both compressed and uncompressed created addresses  .

you are missing 21 characters from your wif which is a lot....
as someone posted above if your address contains small amount of btc than it's simply not worth it (the costs for finding the key are higher than the wallet itself)
if the wallet balance is good maybe a pool can be created for collective search of private key (you will get your share and also the people who searched for your key)
Ofcourse the second method is a little risky (you need to find a trusty person for the pool) but it's still better than nothing.

If you try to recover the key alone chances are 100% that you will not find it even in 10 years from now.
Just my 2 cents, good luck recovering your funds

But do you have partial compressed WIF? As I said before, first you must find the correct starting point and using that start your search (adapting also a stride).