> > What is interesting (?) to me is that I did the chdman "-update" and the
> chdman
> > verify specifically gave me good messages on the MD5 and SHA1 verification.
>
> Well...doesn't have to be interesting at all. If the rawimage was different to
> the original one in MAME, you get an error-free chdman output, since it simply
> calculates the checksums of the 'wrong' rawimage.
>
> ClrMamePro recalculates the sha1/md5 values of the raw image (by decompressing
> the chd) and compares it with a) the stored one in the chd header and b) the one
> in the used datfile/mame rominformation.
> And that seem to fail...so find a correct image.
> The raw file has to have the mentioned sha1 and md5 values. Yours have different
> ones, so it's not the one which mame wants.
>
> Again: chdman only compares the values with the one in the chdheader, not
> against the ones which mame wants.
>
> Roman Scherzer
> ClrMamePro
>
Thank you Roman. Your feedback led me to more investigation and I found the problem to be a "user error" (as you probably suspected). When I originally did the "-createhd" I did not use the [cylinders headers sectors] parameters. This led to an acceptable .chd to be created in chdman, but not equal to what MAME expects (and CMP checks for). I went through the procedure again (using the proper parameters) and it passed CMP.

Sorry about the false alert. I'll try to do more investigation on my own before posting next time.

Joe