Life after Torreys


#1

Hi Ted/Paul

would like to find out what’s road map after Torreys? Is there still cards hidden in your sleeve :slight_smile:


#2

Yep, I’ve got at least three new things to try, but it’s way too soon to talk about them.


#3
Ted Smith said Yep, I've got at least three new things to try, but it's way too soon to talk about them.
41_gif41_gif Can't wait for it.

#4
Ted Smith said Yep, I've got at least three new things to try, but it's way too soon to talk about them.
Does one of them include a user selectable filter with added midrange richness? (hopeful)

#5

Pretty unlikely - there are no digital filters that play with the response in the unit now. They are all as flat as possible over the audio band. The philosophy in general is to make things as neutral as possible and indeed each time we find something to make it more neutral more people seem to like it. When I first started the project I wasn’t looking forward to voicing things, but as it turned out being as technically correct as I could at each juncture produced the best sound.

Further I don’t think that a linear operation (e.g. filtering) can add the kind of midrange richness that you are talking about. I suspect that you want some added distortion which, once again I work to lower with each release.


#6

clapping_gif

105_gif


#7

FYI: I really like DSD with Torreys, it is much better than previous releases. I think the noise floor with Torreys is much lower compared to previous releases.


#8
adriaan said FYI: I really like DSD with Torreys, it is much better than previous releases. I think the noise floor with Torreys is much lower compared to previous releases.
Thanks. The noise proper only measures a little lower (a few dB in band, a few more out of band), but there's certainly less jitter which makes quite a difference IMO and in Torreys Final low level signals are a little more linear which can't hurt.

#9

Aha, for a non-expert it is difficult to tell if the improvements are caused by reduced jitter or reduced noise…

Especially the colour of the sound and the more realistic soundstage with high quality DSD recordings is really something special, and a lot more realistic / musical compared to 24 bit recordings.

I wish that more DSD recordings will come out in the near future. My personal favourite DSD recording is:

https://challengerecords.nativedsd.com/albums/CC72584-symphonies-nos-3-and-4-first-version


#10
Ted Smith said .....

Further I don’t think that a linear operation (e.g. filtering) can add the kind of midrange richness that you are talking about. I suspect that you want some added distortion which, once again I work to lower with each release.


Ted, I think there could be two different possible demands, one could be a vinyl-like “distortion”, producing a certain harmonic structure vinyl listeners cherish. I think this probably makes no sense to add into a digital design as you say and is not wanted by most non-vinyl listeners anyway.

Ok, some producers add tube stages to their DAC’s, probably for a similar effect and tube haters usually argue the same way, that tubes (among other effects) just produce harmonic distortion, no one really “should want”. And anyway tubes are very well regarded commonly and now even used i.e. by PSA within the amps. However, as I said, I understand and agree, using added harmonic distortion to a DAC most probably makes no sense (if it’s positive-efficiently possible at all).

The other demand of few/some/quite some DS owners I think really just could be a little different level of certain frequency areas to give another weight to the sound, independent of or other than any added "distortion’. This could probably be done with a linear operation and optional, maybe even multi-optional and therefore not mandatory.


#11
jazznut said The other demand of few/some/quite some DS owners I think really just could be a little different level of certain frequency areas to give another weight to the sound, independent of or other than any added "distortion'. This could probably be done with a linear operation and optional, maybe even multi-optional and therefore not mandatory.
Not that I disagree in principle with what I think you (and yacheah) are trying to say, but I really do believe it's more complicated than that. People have reported that various beta's and released versions have varying amounts of bass, midrange, highs, etc. and I don't dispute what they hear and report: even so, all of the betas and releases have the same flat FR from, say, 10Hz or 20Hz to 20kHz within +/- 0.01dB. I do know that (at least with the current FPGA code) adding another filter to do a little "optional" FR tailoring would add jitter and noise to the output (as does any non-trivial addition), but that doesn't mean that that will always be the case. (The problem isn't that filtering PCM takes too many resources as much as filtering late enough in the processing chain to also filter the DSD inputs takes interesting resources and add more noise, jitter...) I can say that none of the engineered changes in the base response from release to release affect the FR but instead do (variously) involve phase response (group delay), THD, jitter, noise, etc..

[edit:]

Not that you were suggesting it, but for those who missed my previous posts about adding a tube to the DS: there’s no architectural place to put a tube, nor any need for any function that a tube would provide. The DS needs an output that’s linear to at least 200MHz, not the strong point of tubes, it doesn’t need current or voltage amplification (it uses a passive output stage, i.e. it has all of the current and voltage it needs by the simple virtue of switching one power rail or the other to the outputs.)


#12

If one should start to ‘flavor’ the DS sound, then you are really entering a minefield. It’s like high fidelity with sugar on the top. I prefer not to do this every day of the week and twice on Sundays!


#13
Ted Smith said . . . FR from, say, 10Hz or 20Hz to 20kHz within +/- 0.01dB.
Mind boggling.

. . .

There are tube buffers on the market if you want to deliberately add a tube sound (and you can roll tubes to further tweak the sound).

You can also use a plug-in or stand alone EQ to dial in your own version of Harbeth’s Gundry dip (a mild depression from 1kHz to 4kHz) if you want a frequency response which exhibits a richer sound.

IMO. neither should be deliberately introduced into the sound of a source component.

Query for those who want a richer sound: Do you happen to be listening in the nearfield?


#14

Is there any way to play a multichannel DSD file over 3xDirectStream (or DSJr) using a computer?


#15
jazznut said

Ted, I think there could be two different possible demands, one could be a vinyl-like “distortion”, producing a certain harmonic structure vinyl listeners cherish. I think this probably makes no sense to add into a digital design as you say and is not wanted by most non-vinyl listeners anyway.

Ok, some producers add tube stages to their DAC’s, probably for a similar effect and tube haters usually argue the same way, that tubes (among other effects) just produce harmonic distortion, no one really “should want”. And anyway tubes are very well regarded commonly and now even used i.e. by PSA within the amps. However, as I said, I understand and agree, using added harmonic distortion to a DAC most probably makes no sense (if it’s positive-efficiently possible at all).

The other demand of few/some/quite some DS owners I think really just could be a little different level of certain frequency areas to give another weight to the sound, independent of or other than any added "distortion’. This could probably be done with a linear operation and optional, maybe even multi-optional and therefore not mandatory.


I agree. I think more weight to the sound would work for me.

#16
Elk said

Query for those who want a richer sound: Do you happen to be listening in the nearfield?


Nah no nearfield listening for me. My system sounds better played loud. The low gain of the DS is already a problem in this regard as I need to turn the volume up.

I’d venture that nearfiled listening with the DS wouldn’t work - already I want more midrange richness, with nearfield listening and lower volumes there would be less of that and less of everything all round.


#17
DoggieHowser said Is there any way to play a multichannel DSD file over 3xDirectStream (or DSJr) using a computer?
People do it, I suspect mostly with a multichannel digital audio computer adapter, there are a lot of MC adapters out there that do multichannel A/D and D/A thru USB (or Firewire), but ones that just do multichannel USB -> 3 x S/PDIF or 3 x AES/EBU, etc. are a little harder to find. They are often a little more work to setup for non-digital audio editing software, e.g. foobar2000 or JRiver, but people do it.

#18

After having had the pleasure of visiting PS Audio and being shown around by Duncan, I can tell you that based on what I heard there and what I hear here, especially with the LANRover in place, there is no lack of midrange anything. Superlative sound quality, if your audio equipment is up to snuff. The LANRover is a big deal to me, but I guess I’ll write something about it on the other forum.


#19
yacheah said Nah no nearfield listening for me. . . . I'd venture that nearfiled listening with the DS wouldn't work - already I want more midrange richness, with nearfield listening and lower volumes there would be less of that and less of everything all round.
Listening in the nearfield does not decrease midrange, but can lead to a perception of too much energy around 2kHz; that is, too much upper midrange.

I suspect for whatever reason your system is a bit pronounced in this area. Try a bit of EQ decreasing the energy from 2kHz to 4kHz by 2dB. I bet you will be happier.


#20
Ted Smith said
jazznut said The other demand of few/some/quite some DS owners I think really just could be a little different level of certain frequency areas to give another weight to the sound, independent of or other than any added "distortion'. This could probably be done with a linear operation and optional, maybe even multi-optional and therefore not mandatory.

Not that I disagree in principle with what I think you (and yacheah) are trying to say, but I really do believe it’s more complicated than that. People have reported that various beta’s and released versions have varying amounts of bass, midrange, highs, etc. and I don’t dispute what they hear and report: even so, all of the betas and releases have the same flat FR from, say, 10Hz or 20Hz to 20kHz within +/- 0.01dB. I do know that (at least with the current FPGA code) adding another filter to do a little “optional” FR tailoring would add jitter and noise to the output (as does any non-trivial addition), but that doesn’t mean that that will always be the case. (The problem isn’t that filtering PCM takes too many resources as much as filtering late enough in the processing chain to also filter the DSD inputs takes interesting resources and add more noise, jitter…) I can say that none of the engineered changes in the base response from release to release affect the FR but instead do (variously) involve phase response (group delay), THD, jitter, noise, etc…

[edit:]

Not that you were suggesting it, but for those who missed my previous posts about adding a tube to the DS: there’s no architectural place to put a tube, nor any need for any function that a tube would provide. The DS needs an output that’s linear to at least 200MHz, not the strong point of tubes, it doesn’t need current or voltage amplification (it uses a passive output stage, i.e. it has all of the current and voltage it needs by the simple virtue of switching one power rail or the other to the outputs.)


First, thanks much Ted, that you as a worldwide respected professional and genius talk such things about your product with us lightweights!

I also couldn’t follow the tonal differences people heard between the firmware releases. If any, they were very minor to me.

Just a concrete example, which I point out in order to try to understand what could be the thing we talk around:

I had most listening experience with Accuphase players so far (from the old 12k DP75V to the current 21k DP600). Accuphase digital gear I think is known for being on the side of strong tonal colors and a richer sound in comparison to some others on the opposite side.

My vinyl player (mainly due to an exceptional separate and clearly above standard-supply motor/power unit, also very good resonance control and tonearm) always was less rich and more accurate sounding than the Accuphases, with all common strenghts and weaknesses against digital. I never was a friend of a mainly charming, bloomy, covering vinyl sound, but to me vinyl still provides a harmonic structure to the sound, that finally makes it more real (just in what I hear, maybe not in theory).

With the DS DAC I found a digital gear, improving in all aspects towards the Accuphases, especially its typical strength in ambiance and separation, but also very much vinyl like timing and also in being able to shape clearly more tonal shading out of instruments and voices.

So even if I personally have less of a low end richness problem than maybe few others with the DS (which I can compensate quite good with my active speaker leveling if necessary), I wished for a bit more midrange richness and I noticed a very obvious difference between the DS and the previous Accuphase gear. So I easily understand what happens to some, who switched from such gear to the DS with no option to compensate.

As you say, I think it’s more complicated and I really don’t understand, what makes such differences. I certainly believe you that the DS FR is flat in all firmwares.

Equally I believe the Accuphase guys, that their FR is flat as well as their reviewers writing things like:

„The DP-600’s non-invasive soundstage is perfectly aligned with its handling of frequency response, which is both completely linear and a model of integration.“

But at the same time they write:

The new voicing is evident in the DP-600 SACD Player. You can’t fail to take note of its powerful and dynamic low-end alongside the legendary finesse the brand is known for.

http://www.positive-feedback.com/Issue62/accuphase_dp600.htm

So, complicated for me to understand is:

Where does a rich sound or as metioned above „powerful low end“ come from, if at the same time the FR is flat? Why does different digital gear sound very different in tonality if they all claim their FR is flat? Until resolved this means to me, a flat FR does not necessarily mean a flat sound.

Thanks for your patience to try to explain such puzzels.