HOLO Audio MAY DAC

[Extreme_Boky]

NUC 11, Windows 10, HQPLayer (also JRiver) "directly" to May (using USB), XLR outs straight to Aleph J (no coupling capacitors anywhere is the signal chain). Very fast, revealing... and of course, with a very loud pop. However, the pop is much quieter with JRiver. 

 

According to what Miska said re: how DSD works, I would expect that UNLOCK <-> LOCK is expected after a stop, and before the playback starts again. 

[Extreme_Boky]

I think the DAC should provide mute... but how would it know that one is feeding it DSD? Once it figures out, it is too late... the pop has already fried your tweeters  :)

 

The digital mute could be implemented so that the mute is active whenever there is no digital stream (UNLOCK), however, I think that DSD locking and unmuting would cause a bit of silence at the beginning of a song. 

 

Chicken, egg and the trade offs. 

[camott]

7 minutes ago, Extreme_Boky said:

According to what Miska said re: how DSD works, I would expect that UNLOCK <-> LOCK is expected after a stop, and before the playback starts again. 

 

I think NAA behaves differently. My May is glued to "DSD256X" even if stopped for hours after the last play. Next test will be from HQPlayer Embedded directly from Fitlet2 and see if HQPlayer itself unlocks/locks with every USB stream change.

 

However, I am now trying DOP and so far my upstream HQPlayerDesktop (M1 Mac) has performed just as well as with DSD native, 96k > DSD256 ECv2 and no more static pop! Try the SDM Pack = DOP setting. If there is no sonic cost to DOP unpacking then this just might do it.

[Extreme_Boky]

Great suggestion, thank you.

[sledwards]

17 hours ago, camott said:

However, I am now trying DOP and so far my upstream HQPlayerDesktop (M1 Mac) has performed just as well as with DSD native, 96k > DSD256 ECv2 and no more static pop! Try the SDM Pack = DOP setting. If there is no sonic cost to DOP unpacking then this just might do it.

Nice fix! I had your exact behavior with my May KTE and DoP did the trick. My "pop" was more like an annoying "click", and only coming thru the left channel for some reason.  Now when the music is paused, display switches from DSD255X to 705.6 PCM. From all the reading I have done, using DoP does not result in any audio quality issues.

[camott]

26 minutes ago, sledwards said:

Nice fix! I had your exact behavior with my May KTE and DoP did the trick. My "pop" was more like an annoying "click", and only coming thru the left channel for some reason.  Now when the music is paused, display switches from DSD255X to 705.6 PCM. From all the reading I have done, using DoP does not result in any audio quality issues.

 

Yes my pop/click only comes through the left channel as well. So definitely the same thing. And I also see the 705.6 PCM in display when paused - because that's the actual wrapper format. When music starts streaming, the May sees that it's actually DOP content inside the PCM and changes the display to DSDxxx.

 

But yeah, DOP is working just fine. Given that the May supports all of the different DSD rates, it's not really a concern using DOP for any of them. Glad it's working for you too!

 

(The fact that DOP works suggests it's not an upstream source issue with DSD native. My only guess is that the issue lies with Linux Alsa driver or on the May itself - what platform/driver is your endpoint?).

[sledwards]

49 minutes ago, camott said:

 

Yes my pop/click only comes through the left channel as well. So definitely the same thing. And I also see the 705.6 PCM in display when paused - because that's the actual wrapper format. When music starts streaming, the May sees that it's actually DOP content inside the PCM and changes the display to DSDxxx.

 

But yeah, DOP is working just fine. Given that the May supports all of the different DSD rates, it's not really a concern using DOP for any of them. Glad it's working for you too!

 

(The fact that DOP works suggests it's not an upstream source issue with DSD native. My only guess is that the issue lies with Linux Alsa driver or on the May itself - what platform/driver is your endpoint?).

I am running an Intel i5-9400T processor; HQPlayer embedded image (USB boot stick) for NAA functionality. Needed to run NAA from embedded image because NAA standalone image would not recognize my JCAT NET Card FEMTO

[Extreme_Boky]

I just used 16 of Black Gate NX Hi-Q 0.1uf 50v types in May DAC power supply unit.

Stunning, amazing, call it anything you want. Mind-blowing.

 

I also checked the price, the last price I found (eBay) was approximately US$100... each.

[MurphMasterPro]

14 minutes ago, Extreme_Boky said:

I just used 16 of Black Gate NX Hi-Q 0.1uf 50v types in May DAC power supply unit.

Stunning, amazing, call it anything you want. Mind-blowing.

 

I also checked the price, the last price I found (eBay) was approximately US$100... each.

Wait, what?  Sounds incredible!  Can you share pictures of what you did?

[Roasty]

Just received and installed the new replacement USB module from Tim @ Kitsune. 

 

For those attempting this for the first time.. Flip the unit upside down. Every screw in the base (except the feet screws) need to be removed. I also removed the screws on the backplate. Then slowly flip the unit the correct way up and slowly remove the cover. Remember to remove the ribbon connecting the board to the front display. Following that, the usb module replacement itself is pretty straightforward. Just two screws holding the board in place. 

 

[Extreme_Boky]

Well, it would be good to see the whole USB card.... but from that section that you posted, I see a completely different processing chip. It is not XMOS any more.... and, the I2S isolator IC's are also different; the new ones may have the series I2S signal resistors imbedded. I can not read the part numbers.

 

So, the changes seem quite drastic. Also, sorry to say this... but that Altera IC solder job looks... well... not good. It could be the photo.

 

If anyone can post in-focus, well lit photos of a new USB card, I will be able to tell much more...

 

Did you notice any changes in sound quality?

[Roasty]

12 minutes ago, Extreme_Boky said:

Well, it would be good to see the whole USB card.... but from that section that you posted, I see a completely different processing chip. It is not XMOS any more.... and, the I2S isolator IC's are also different; the new ones may have the series I2S signal resistors imbedded. I can not read the part numbers.

 

So, the changes seem quite drastic. Also, sorry to say this... but that Altera IC solder job looks... well... not good. It could be the photo.

 

If anyone can post in-focus, well lit photos of a new USB card, I will be able to tell much more...

 

Did you notice any changes in sound quality?

 

Yea sorry forgot to take a pic of the full module. 

 

Don't know if there is a change in sq. But one thing I noticed is that I somehow need more turn on the volume knob to achieve the same levels. Previously my listening was around 9 to 10 o'clock on the knob, but now it is 12 to 1 o'clock. 

[Extreme_Boky]

I think XMOS USB to I2S is still on that USB card... somewhere. Altera CPLD is probably used to provide a better 1.536MHz support.

[musicjunkie917]

So I have a Holo Audio May KTE DAC coming soon. I currently use HQPlayer Embedded on a sonicTransporter i9 and an ultraRendu acting as an NAA. The May will support up to PCM 1.536MHz. The ultraRendu is limited to PCM 768kHz. 

 

With my old DAC, I was using HQPlayer to upsample both PCM and DSD to DSD256 (poly-sinc-ext2, ASDM7ECv2). Is that a good idea with the May KTE? Or does upsampling all PCM to PCM 705.6kHz/768kHz sound better? Or, to get the best sound out of PCM, do I really want a new endpoint that supports up to PCM 1.536MHz?

 

Thanks!

[luisma]

7 hours ago, musicjunkie917 said:

So I have a Holo Audio May KTE DAC coming soon. I currently use HQPlayer Embedded on a sonicTransporter i9 and an ultraRendu acting as an NAA. The May will support up to PCM 1.536MHz. The ultraRendu is limited to PCM 768kHz. 

 

With my old DAC, I was using HQPlayer to upsample both PCM and DSD to DSD256 (poly-sinc-ext2, ASDM7ECv2). Is that a good idea with the May KTE? Or does upsampling all PCM to PCM 705.6kHz/768kHz sound better? Or, to get the best sound out of PCM, do I really want a new endpoint that supports up to PCM 1.536MHz?

 

Thanks!

You need to decide by yourself what you like best if PCM or SDM, try both.

For max PCM I would do any endpoint in which you can load Jussi NAA latest image, don't know if you can load that in the UR but his image is fantastic

 

[musicjunkie917]

1 hour ago, luisma said:

You need to decide by yourself what you like best if PCM or SDM, try both.

For max PCM I would do any endpoint in which you can load Jussi NAA latest image, don't know if you can load that in the UR but his image is fantastic

 

 

I will certainly listen for myself. But, I don't have an NAA that can do PCM 1.536MHz so I can't listen to that for myself. Also, my old DAC was not a NOS DAC and accepted PCM up to 384kHz only. It did its own internal upsampling with PCM and DSD. That's why I was soliciting opinions.

 

The ultraRendu already uses a minimal OS so I don't suspect that running Jussi's NAA image, even if it were possible, would improve sound quality.

 

So, anyone with opinions of upsampling everything to DSD versus upsampling DSD to DSD256 and PCM to PCM 768kHz and PCM to PCM 1.536MHz, please post!

[Bertel]

26 minutes ago, musicjunkie917 said:

The ultraRendu already uses a minimal OS so I don't suspect that running Jussi's NAA image, even if it were possible, would improve sound quality.

 

As you say, the ultraRendu actually has NAA functionality already included, that‘s what you are using already, no need to change that. I have the Signature Rendu and do exactly that, for me the best option. My May KTE does play PCM to PCM 1.536MHz this way, so I‘d expect your ultraRendu to do the same.

[Bertel]

I should also note that, like you, I upsample everything to DSD256 (also with poly-sinc-ext2 and ASDM7ECv2) because as Jussi showed in his measurements that‘s the sweet spot of Spring/May. For me this sounds excellent, no need to change anything.

[musicjunkie917]

8 minutes ago, Bertel said:

I should also note that, like you, I upsample everything to DSD256 (also with poly-sinc-ext2 and ASDM7ECv2) because as Jussi showed in his measurements that‘s the sweet spot of Spring/May. For me this sounds excellent, no need to change anything.

 

 Thank you for the replies!

 

Where does Jussi show these measurements?

[87mpi]

Listen with your ears... 

[Bertel]

9 minutes ago, musicjunkie917 said:

Where does Jussi show these measurements?

 

I think that's where he started: 

 

[musicjunkie917]

20 minutes ago, Bertel said:

I think that's where he started:

 

Awesome! Thank you! Thank gives me a lot of good information with which to start my testing.

 

I will set the DAC bits to 20.

 

I know that Jussi normally says that you want to set the max volume to -3dBFS. I set mine to -6 as my Thriller SACD would increase the limit count when set at -3. With the May being 6 dB quieter with SDM compared to PCM, do I still need to adjust the max volume to -3 or -6? I am not sure the two are related....

[Foggie]

1 hour ago, Bertel said:

 

As you say, the ultraRendu actually has NAA functionality already included, that‘s what you are using already, no need to change that. I have the Signature Rendu and do exactly that, for me the best option. My May KTE does play PCM to PCM 1.536MHz this way, so I‘d expect your ultraRendu to do the same.

I thought the sonore products are limited to 705/768 PCM, how are  you getting 32fs?

[Bertel]

25 minutes ago, Foggie said:

I thought the sonore products are limited to 705/768 PCM, how are  you getting 32fs?

 

No idea, I just tried it a while ago and it works 🙂 I didn't do anything special, no changes or adjustments were required on my side.(However as said I'm using DSD256 instead) I think the 705/768 figure just is the officially supported one

[barrows]

4 hours ago, Bertel said:

 

No idea, I just tried it a while ago and it works 🙂 I didn't do anything special, no changes or adjustments were required on my side.(However as said I'm using DSD256 instead) I think the 705/768 figure just is the officially supported one

Thank you @Bertel for confirming that your Signature Rendu works with 1.536 MHz PCM on the May!  We (Sonore) have only been able to test up to 768 in house, because we do not have access to any DACs which operate at higher rates.  We have suspected that 1.536 MHZ would work, but we could not make make any promises to customers without being able to test such ourselves.

You guys using HQPlayer for DSD 256 with the ASDMEC7v2 modulators and poly sine ext2, have you tried poly sinc gauss?  I am really liking poly sinc gauss here for DSD 256 with my Bricasti M3.