Probability distribution is an integral of probability density function, these are not the same. Both are defined positive and negative side. The second in most practically cases is symmetric. Time domain responses are neither symmetric nor defined at positive and negative sides. Therefore convolution should be performed differently in every case. Perhaps, the issue is also about reusing the same functions, but it seems we need several. Among them we can also distinguish between 'cyclic convolution', where one of participating input is periodic (e.g. eye diagram) and non-cyclic (BER or other PDFs) -----Original Message----- From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Mirmak, Michael Sent: Wednesday, October 07, 2009 12:41 PM To: Muranyi, Arpad; IBIS-ATM Subject: [ibis-macro] Re: AMI support for PDF convolution? Sorry, D=distribution or density, as you point out. - MM -----Original Message----- From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad Sent: Wednesday, October 07, 2009 11:38 AM To: IBIS-ATM Subject: [ibis-macro] Re: AMI support for PDF convolution? Mike, All the inputs and outputs are in the time domain, according to the AMI specification, however, inside the DLL's functions you may convert to any other domain if you like, if the mathematical rules allow you to do so (periodicity, LTI, etc...). I am not sure what PDF stands for in your question, but if you meant Peak Distortion F(?) analysis, you are basically on the right half of the flow diagrams (note on the bottom of that half the box labeled "Stateye type analysis". If you are referring to something like Probability Density Function, than you are most likely talking about what I labeled "post process results" or inside GetWave functions on the right side which is up to the model maker. I am not sure I answered your question, but a little clarification would be helpful. Arpad ====================================================== -----Original Message----- From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Mirmak, Michael Sent: Wednesday, October 07, 2009 12:33 PM To: IBIS-ATM Subject: [ibis-macro] AMI support for PDF convolution? All, Based on my understanding, AMI supports time-domain convolution. But does it support PDF convolution as well? - MM --------------------------------------------------------------------- IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/ IBIS Macro reflector: //www.freelists.org/list/ibis-macro To unsubscribe send an email: To: ibis-macro-request@xxxxxxxxxxxxx Subject: unsubscribe --------------------------------------------------------------------- IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/ IBIS Macro reflector: //www.freelists.org/list/ibis-macro To unsubscribe send an email: To: ibis-macro-request@xxxxxxxxxxxxx Subject: unsubscribe --------------------------------------------------------------------- IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/ IBIS Macro reflector: //www.freelists.org/list/ibis-macro To unsubscribe send an email: To: ibis-macro-request@xxxxxxxxxxxxx Subject: unsubscribe --------------------------------------------------------------------- IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/ IBIS Macro reflector: //www.freelists.org/list/ibis-macro To unsubscribe send an email: To: ibis-macro-request@xxxxxxxxxxxxx Subject: unsubscribe