Tír Na HiFi

Best sse in my setup is still 4.22, tried all the latest sse's. None comes close.

4.22 has a tremendous live characteristic and foot tapping factor. I enjoy this version a lot.

Just thought maybe someone would like to know. -:))

3daudio wrote:Best sse in my setup is still 4.22, tried all the latest sse's. None comes close.

4.22 has a tremendous live characteristic and foot tapping factor. I enjoy this version a lot.

Just thought maybe someone would like to know. -:))

shall check what code it used. Foot tapping seems to be something that could be measured, at last we have a parameter that can be measured. (beats per minute/foot taps per minute) bpm/ftpm - the baft ratio

5.62 perhaps better balanced, might be too much, but can be dialled back

quite like the sound of 5.63, shall call it a day there. surprisingly sweet sound, good bass , dynamics, detail.

sweet and loss of detail, 5.64 fixes it.

toodle pip

sbgk wrote:5.62 perhaps better balanced, might be too much, but can be dialled back

quite like the sound of 5.63, shall call it a day there. surprisingly sweet sound, good bass , dynamics, detail.

sweet and loss of detail, 5.64 fixes it.

toodle pip

No 5.64 on Google drive?

Aleg wrote:

sbgk wrote:5.62 perhaps better balanced, might be too much, but can be dialled back

quite like the sound of 5.63, shall call it a day there. surprisingly sweet sound, good bass , dynamics, detail.

sweet and loss of detail, 5.64 fixes it.

toodle pip

No 5.64 on Google drive?

there now

I'd like to request avx versions of the latest release(s) as well. My AMD CPU can handle avx2, but for some reason avx sounds much better, to such an extent that in direct comparisons I even tend to favour sse over avx2! I seem to be the only one here with an AMD CPU, but in case there are any others, I thought I'd mention it.

This discovery also means that testing with 1644 avx2 on my system, as has been happening the last few weeks, may have given skewed results in my case. I did some tests yesterday on the few available 2496 avx versions and found that 5.51 is indeed better than 5.07, but going back to a 2 month old favourite 3.39, it blew 5.51 away in terms of musicality, timing, coherence, solidity and spacial atmospherics. 5.51 did manage to retrieve quite a bit more detail, but at a price that's way too high for me.

I hesitate to ask this, but I would really love to try some 2496 avx versions of milestones in between 3.39 and 5.51, like my old favourite 4.04 for example. :-)

I think at these later stages of polishing up the code, system synergy may play an important role in the final sound quality, particularly how the PC handles the instructions.

Suspect the cpu architecture will make a difference
3.39 was a 10ms version
5.32 and 5.34 were around 3 ms
5.30 is good too if it works
Although there not perfect But they have musicality flow atmosphere stuff were still missing
Noone else seemed to take to them though

satshanti wrote:I'd like to request avx versions of the latest release(s) as well. My AMD CPU can handle avx2, but for some reason avx sounds much better, to such an extent that in direct comparisons I even tend to favour sse over avx2! I seem to be the only one here with an AMD CPU, but in case there are any others, I thought I'd mention it.

This discovery also means that testing with 1644 avx2 on my system, as has been happening the last few weeks, may have given skewed results in my case. I did some tests yesterday on the few available 2496 avx versions and found that 5.51 is indeed better than 5.07, but going back to a 2 month old favourite 3.39, it blew 5.51 away in terms of musicality, timing, coherence, solidity and spacial atmospherics. 5.51 did manage to retrieve quite a bit more detail, but at a price that's way too high for me.

I hesitate to ask this, but I would really love to try some 2496 avx versions of milestones in between 3.39 and 5.51, like my old favourite 4.04 for example. :-)

I think at these later stages of polishing up the code, system synergy may play an important role in the final sound quality, particularly how the PC handles the instructions.

3.39 is the one the played distorted sound on most installations and played only well on the odd machine.
If it was so good, it may pay dividend the find out why it was so distorted on most machines.

As was 5.30

satshanti wrote:...but going back to a 2 month old favourite 3.39, it blew 5.51 away in terms of musicality, timing, coherence, solidity and spacial atmospherics....

think that 3.39 has bass reaches low enough makes treble very likable.
don't think another version has same tonality..
when listening to enough versions...
one can tell that when bass reaches low enough... magic happens... stage width is widened for example

listening to it again... sounds better than i remember...
more liquid and shinny that it was...

I found an easy way to make MQn playback DSD-files into a DoP capable DAC like a Chord Hugo.

It is possible to encapsulate dff or dsf files into a DoP PCM format offline, before playback. There are few converters around that can do this.

These DoP files can be wav-format and therefore can be played back directly using MQn, if MQn would have a 176.4kHz/24-bit version that is.

I have made a few scripts that can convert an SACD-iso into dff and encapsulate the dff into DoP pcm files.

Gordon could you build a 176.4kHz/24-bit version for 32-bit containers so I try the complete playback chain?

Cheers

Aleg