So, perhaps there needs to be more break in time, perhaps the Touch digital signal is so good that the Monarchy Dip is not necessary.
I will give it a few more days.
Anyone else have any thoughts or suggestions?
it might be that the Squeezebox & DIP need more breakin time but....
OK, there were 2 Monarchy products - let's see if I get this right ;-) - the 24/96 interface processor & the 48/96 which was an upsampler.
The 48/96 product that you have is an upsampler that uses the Cirrus Logic CS8420 receiver as a standalone unit. OTOH, the 24/96 product is designed specifically to reduce jitter & amplify the digital signal.
Monarchy claims both products to be asynchronous & in a way they are: the CS8420 has an on-board PLL that is used to recover clock from the incoming data stream. By using a precision clock reference for the PLL, the PLL can clean up the jitter from the incoming data & reclock the data using its cleaner clock as the reference. Note that in sort of scheme jitter is reduced to as low as the PLL clock reference is & no lower. In much more expensive jitter reduction units, a 2nd precision clock reference is used to encode the outgoing data so that this outgoing data is on a totally different clock domain where the jitter can be carefully controlled. Such hi-end jitter reduction devices often use a FIFO to do the job, which the Monarchy DIP products do not.
So, the jitter reduction capability of the Monarchy 48/96 might be limited & the Squeezebox Touch input might be at the limit of what the DIP 48/96 can do?
I was reading JA's measurements of the SB Touch-jitter performance looked OK to me. The way JA did his jitter measurements was that he wanted to see how much suppression to added jitter the SB Touch was capable of. So, he added a (slow) 229Hz sq wave to the LSB i.e the 16th bit of a 16-b word of the 11.025KHz sine. The effect of adding the 229Hz is what we would call a dither & here it is forcefully adding jitter so that 11.025KHz is not precisely 11.025KHz anymore. Then, JA looked to see what the SB Touch was capable of doing in terms of extracting the 11.025KHz at the analog output & yet suppressing all the 229Hz harmonics. This would be possible only if the SB Touch had implemented a good jitter reduction scheme. -122dB - (-6dB)= -116dB suppression of the 229Hz harmonics, which is not bad at all for a $299 unit. So, the SB Touch's jitter performance is quite good to start off with & it might be at the limit of the DIP 48/96 unit.
I also have a feeling that you probably need the DIP 24/96 (jitter suppressor) unit more than you need DIP 48/96 (upsampler) unit. Many people use both in tandem - 24/96 followed by 48/96 & then into your DAC. This upsamples a low jitter signal & it gets a 2nd chance for jitter reduction.
I found similar jitter performance plots for the SB Touch on computeraudiophile.com:
http://www.computeraudiophile.com/content/Logitech-Squeezebox-Touch-Review.
search for the word "jitter" & you'll come to the 05/28/2010 post by JR_Audio. He's posted 3 JPEG graphs for the SB Touch jitter performance. Very similar to JA's.
If you read the (lengthy) set of posts on the SB TOuch you'll see that many people are buying upgraded power supplies & junking the provided switched-mode wal-wart power supply 'cuz it's just too noisy. A good old-fashioned linear regulator is much quieter & will probably also help the overall jitter performance (it will allow you to realize the SB Touch designed jitter suppression performance). It is possible that the wal-wart supply is so noisy that it is adding hash to the music signal & creating an artificial noise floor + creating digititis for you. Needless to say, Logitech did manuf this unit to a price-point hence several compromises & it should be no surprise that you need to spend more money to actually make it a higher-fi unit......
Just my rambling 2 cents FWIW....