Hi suhas

>HHEd -A -D -T 1 -H hmm14/macros.txt -H hmm14/hmmdefs.txt -M hmm15 tree.hed triphones1 

shouldn't you be executing this command?:

$HHEd -A -D -T 1 -H hmm12/macros -H hmm12/hmmdefs -M hmm13 tree.hed triphones1

> error [+2663] chkTreeobject:TB only valid for 1 mix daigonal covar models 

The HTK book says:

2663 Invalid types to tree cluster
     Tree clustering will only work for single Gaussian diagonal covariance untied models of similar topology.

Have you deviated from the tutorial?  If yes, how?

thanks,

Ken

I had the same problem with this command

error [+2663] chkTreeobject:TB only valid for 1 mix daigonal covar models

Does this error  want to say that i cannot work with a mixture of Gaussian in the definition of the model HMM

because  I worked with a number of Gaussian =4 .And I think it is the only point where I deviated of this tutorial.

Have you checked the HTK email archives?

Ken

hello to all

i think that you can use TC command instead of TB , in your tree.hed file, because TB can handle only single gaussian models but TC is not so, i had the same problem. you can find some text about it in pages 171-172 from htkbook

good luck 

hello to all

i think that you can use TC command instead of TB , in your tree.hed file, because TB can handle only single gaussian models but TC is not so, i had the same problem. you can find some text about it in pages 171-172 from htkbook

good luck