Hi Peter - Yesterday I compared (with my ears) XXhighend against Foobar with SoX Resampler - mod version (with resampled to 196) + Convolver with correction filters created by Audiolense.

I am not sure if others have experienced the same thing, but as usual XXhighend had better imaging and a more 'black' background. On the whole it is more musical. However, there were 2 things which XXhighend was weaker: 1) it was not as lively, in that it seemed to be putting its effort into creating a beautiful sound but it seemed to lack the 'oomph' for certain pieces (especially symphonic pieces such as Schubert's 9th or Beethoven's 7th and did not sound as dynamic).  2) it also lacked a certain detail in the middle register (for jazz pieces with a lot of details in it such as the 1st track of Chick Corea's Ultimate Adventure, or Corryell, Asad and Abercrombie's Three Guitars). I think that for 2) was a combination of Convolver and the higher resampled frequency (compared to XX's 176hz) but I can't explain 1).

When I switched to XX last year it was clearly better than Foobar but recently with these comparisons I am not so sure. I will experiment some more and update with more of my experiences, but perhaps if you have the time you can also try and let me know what you think.

hi Peter - I have spent the last few days experimenting with all types of combinations between buffer settings, file sizes, Qvalues and Core schemes. My conclusion is that they all help. In fact with scheme 4 with the highest buffer settings (2094), Q settings (2094) and lowest file sizes (100mb) I get some stability.

But... always.. after a certain amount of time the distortion will return. Also, my system sounds terrible under Scheme 4 (scheme 3 is the best).

But last night I installed 09z1 and used the following settings: adaptive mode, 2094 buffer, Q settings, 100mb file size and Scheme 3, and also moved the XX folder out of 'Program Files' onto my desktop. This combination seems to give me the best quality for the moment with no distortion (even better than Special), but I need to listen somemore. Will report back again with my result.

Unfortunately I am having some problems getting Special to work without encountering distortion / stuttering. 

I find that after I set the device buffer and Q settings to match the DAC buffer, the player will start out okay until it needs to flip to the next track. At that point the distortion will start to occur. 

Also these problems seem to occur more when I am playing FLAC files (and especially if I have other processes or programs running in the background, such as unzipping or file transferring). I have switched off disk indexing so that is not a problem.

I don't experience any of this in Adaptive mode (which in my opinion doesn't sound as good as Special).

Will getting more RAM (I have 2 GBs) or CPU power help??

But isn't it better to have arc prediction play 48/192 rather than 44.1/176?

Appreciated, thanks. The rate is actually 3072, to be precise.


Thanks for the clarifying. My DAC is capable of receiving 44.1/176 (which is what I am using now, but it is capable of going up to 48/192). so..


from this, I take it that it is possible for XX to upsample to 48/192. Do you know how I can do that? The octo slider seems to only allow one to select the multiple (i.e. 44.1X1, or 44.1X2 etc.) but not from 44.1 to 48. Sorry if I am missing something.

Peter is there any chance you can add 3024 to the buffer settings? This is my DAC's highest buffer after 2048. Unfortunately I can only chose 2048 or 4096 in XX but 2048 or 3042 in my DAC's settings. Also, if my DAC is 24/192 why does it default to 174 when I select Arc Prediction? When I select the 96 or 192 only button in the settings area and Anti Image then I get 192 but that option is not available in Arc Prediction? Sorry as you can tell I don't have a great understanding of device buffers and latency..

Sorry can you clarify what this means, as I have some FLAC encoded HDCDs and haven't been able to play them in Arc Prediction mode. Are we only able to play them in Anti image or  Linear mode?

Peter - I'm get a "HDCD decoding problem during decoding" error when decoding a FLAC encoded CD which XX has recognised as an HDCD in AP mode. No problems in AI or Linear mode. what gives?

Love the SQ (especially 'special' mode) by the way.

I like the new layout of the GUI a bit better in that the Q sliders are now more intuitively placed, as well as the double-octo slider thing (even though my DAC doesn't support 8x AP). Not sure about the knobs though or the way panels slide in and out. I find them a bit gimmicky. Am sure you're able to fit all the sliding panels on the same front panel no? Also the colour scheme is a bit.. er like something from the 80s.

These are just minor minor observations though, because I don't know enough to comment on what's under-the-hood other than that it sounds great and well worth the money!

^{I'd call Special 'dry' and Adaptive... well, just 'perfect' actually. It is very, very listenable.}

oh manis.. what would Romy the Kat say if he were here. ha ha!

Am loving Special mode so far. Now all XX needs is VST support so I can mess about with the phasing and X-overs.

Peter - how about instead of me going out and buying an RME AES I just pay you the money instead and you build in the support??

I just checked out the Trinity DACs. What do you think of the dual mono-DAC formation? Anyone know what DAC chips they are using? In the demo how did they connect the controller to the PC, using rca through the headphone socket??

Hi Matt - Thanks for bringing up the subject. I think we would all agree that psychoacoustics plays a big part in audio fidelity. I think, though, there is an element of 'physicality' in the sound waves which reach our ears as well (I believe is Peter's point). Therefore the question is how much of the image or localisation is done 1) 'in-head' and how much is not done 'in-head' because 2) it is already there and all the brain needs to do is perceive it for what it is without having to do any processing. Ambiophonics is an attempt to help with 1) 'in-head', but the ultimate goal for us all here is to achieve 2).

In Peter's example, flagiolettes are perceived as a much clearer image compared to before. Question is: is it because of 1), or 2)? If I understand Peter correctly, it is 2) because waves are interacting with each other in such a way that the aural image of flagiolettes is 'improved' before it is perceived by the ear. Unfortunately I personally do not see how this effect can substantially contribute to or alter the 'other' direct waves which do not interact with each other but travel straight to the ear (and hence are given 'priority' by the brain compared to the secondary reflected or 'interacted' waves). To support my view, I will say that flagiolettes are a good example because they consist of high frequencies which necessarily travel faster in a more directive fashion and are hence perceived by the ear more quickly and immediately than lower frequencies and that 'in-head localisation' can therefore occur in normal rooms when listening to speakers with high directivity / narrow dispersion / and with high frequencies (such as flagiolettes being sounded by horn speakers!).

I hope that I can be proved wrong.

I understand your point (I think). Sorry I forgot your question.

But before I submit an answer (from my very layman's perspective and understanding) I should clarify that my belief (which I do not hold dearly but faintly) is actually that any interaction of waves in the listening space is undesirable. This is not because I do not believe that stereo speakers cannot provide the same level of accuracy and imaging as a live performance (even though that is the case for many types of performances and speakers). It is just that recording engineers already embed spatial cues perhaps in the form of recorded phase shifts in the recording mixes themselves, and that the listening experience is 'purer' (and hence 'better') if one could only listen to direct waves such as in an anechoic chamber or through headphones. I understand that this is not our reality and that reflections are a fact of life. I also understand as well the other argument that resonant room modes can be exploited to enhance the listening experience. If you have thoughts in this area I will be grateful if you will share it.

Regarding your question, I would say that if a direct wave comes earlier from the left speaker and interacts with a direct wave from the right speaker then the 'collision' would occur at the right plane (assuming that the listener is on axis between the speakers). If, as you say the effect is to increase the sound pressure then unfortunately I do not know in what form the mixed result will ultimately travel to the listener and how it will be perceived. Can you give me another clue?

Peter - From the listener's perspective, sound waves are only perceived in one place: the head. I think it is questionable whether or not the interaction of soundwaves before such waves are received by the listener's ears produce a good result for the listener. If 'interaction' were essential then there would be more inferior headphones to speakers - when in fact the opposite is true!

I have tried it and found the same as you. i.e accuracy is lost. It appears that the processing is subtracting something from the signal. Also from what I understand crosstalk is a product of the interaction of the speakers with you and your room and not something which can be dealt with at source.  The problem with building a barrier (apart from the obvious inconvenience it will cause) is that it will only work if you sit with one ear on one side and the other ear on the other side of the barrier, and hence only you will be able to enjoy it and no one else. I have read that apparently the best way to deal with this is to use a properly configured surround system (see Floyd Toole's book on Psychoacoustics). Romy the Cat's forum also contains interesting information on this subject (if you can take his attitude).

Peter - Thanks for the explanation. It is all much clearer to me now (I think). If I understand correctly, the "filtering" function now occurs at the software playback level (as opposed to the DAC level).

I have some further questions, if I may. Can you please expand on why you say:


I am not sure I understand how multiple samples can mess up output.

Also if I don't hear much of an improvement between Arc Prediction / Quad mode and non Arc Prediction / non Quad mode, could it be because my DAC is filtering the output of XXhighend and making them both sound the same?

And finally I thought about Manisandher's question as well and it seems to be that the arc prediction + NOS solution is designed to replicate whatever is recorded on the CD faithfully so to the extent that the recording is not as good (whether because of delta-sigma or anything else), there will be a limitation, but an Arc Predicted / NOS DAC outputted sound should still be better than a filtered DAC output in any situation.

Admiring your quest for better and better sound.