Which matters more, DAC or transport?

Thanks alot for the input. I wish I had all the equipment I'm interested in to do an actual listening test. It's really the only way.

The idea of splurging on a DA then buying a cheap transport, is foolish.

well said/written!
Indeed the Transport is more important than the DAC. The DAC is "nothing" without a good signal feed into it AND the axiom garbage in, garbage out holds in spades here.
Think about it: what happens if you have a cost-no-bar DAC, say the dCS DAC or the Esoteric DAC, and you feed it with a BestBuy $40 DVD player used as a transport. What sort of sound can you expect to get??
Will the DAC fill-in for the crappy transport?? If it's lost at the output of the transport, will the DAC bring it back??
No and no!
It's the transport/source.

Based on my own experience both obviously matter but I would say transport 70% and the DAC 30%. There is nothing a DAC can do to compensate a poor performing transport. It all depends on a superior signal.

Yes, you need a good jitter free signal but getting one from a transport or other source and cable or via reclocking right at the DAC is not rocket science these days and no longer must cost a fortune either IMHO.

Assuming a good signal can be had relatively easily, then the choice of DAC will always have a huge impact on resulting sound from there and matters a lot.

The only difference between good and cheap transport is amount of jitter it produces. Jitter rejecting upsampling DACs like Benchmark have so low jitter bandwidth (few Hz) that at frequency of interest (kHz) jitter rejection is in order of 100dB. In addition jitter induced artifacts (jitter=noise in time domain) are, even with poor transport at the level of -80dB. Bel Canto DAC3 is also upsampling and jitter rejecting AFAIK. If you decide to get different type of DAC (like NOS DAC) then get very good transport and good digital cable with good shielding and impedance maching (signal reflects on impedance boundaries).