Holo Audio May DAC

Has anyone had problems getting a response from Tim/Kitsune? I have emailed him twice with questions/lead time and have not received a reply. I actually live near him in Seattle and directly picked up some small products from him a few years ago - he was very helpful/responsive back then.

Thanks for pointing that out @georgehifi. I am thinking about a Holo May to front my single ended Line Magnetic 805ia and was concerned that I would be missing out on optimal connectivity, that the SE connections would be compromised.

@tinear123 thanks. I am currently using a Lampizator Amber 3 with my LM SET amp. I want to either upgrade to the Lampizator Baltic 3 or the Holo May. Lampi's have a certain house sound that is palpable and vivid but possibly a little too much at times. I have owned a couple of lower end R2R's before (Denafrips Ares and an Audio-GD) and recognize that the typical R2R sound is quite different from the Lampizator house sound, but not entirely sure to what degree the Holo May bridges that divide. Guess there is only one way to find out ...

So I have acquired a slightly used Holo May ( 200 - 300 hrs in mint condition) and have been listening to it for the past few days, comparing it to my Lampizator Amber 3. I am on the fence so far to be honest. Earlier in this thread, more than one person suggested that the May would be a sideways move from Lampizator DAC’s and I am beginning to agree with this. While it’s incredible detailed, accurate, with a black background and neutral to a fault, it doesn’t have the full, 3-dimensionality, or organic sound of the Amber 3. No doubt it’s technically superior in every way - but it’s just not as *pleasant* to listen to, is more forward sounding, and very unforgiving of poor recordings - etched and bright at times. I will let it run for a couple of hundred more hours before re-evaluating.

A few days later and what I have done is create an A/B setup with Roon playing to two endpoints as a grouping, one going to the May and one going to the Lampizator. Each of which is connected to different inputs on my Line Magnetic 805ia SET amp. Same PEQ (but no upsampling!) is applied to both endpoints in Roon with the headroom adjusted slightly differently to level match the volume within 0.1db. So now all I have to do is flick the input on the amp to switch between the two while live - the changeover is absolutely imperceptible in both volume and timing. (One of the cool things you can do with Roon).

There is very little between the two much of the time. The May possibly has a slightly wider, taller soundstage and thus instrument separation, but does not have as much depth and thus less 3-dimensionality to instruments. The May is a bit clearer with more bite (particularly the bass), and has a bit more air to the sound. (i.e. frequency extension no doubt better). Tonally, the May is really neutral across the entire frequency range whereas the Lampi has a bit of that tubey mid-range bloom and richness. The Lampi is a bit rounder and fuller. Timbre is a toss-up. Punchiness is a toss-up. Detail a toss-up. May has a wee bit more PRAT, Lampi a wee bit more romantic. May a bit more forward than Lampi.

Nothing really surprising here given their topologies and builds. My LM amp is rolled with best-of-class tubes such that it’s extremely clean/punchy/dynamic - dare I say almost solid state like (but still with that SET magic). Speakers are very neutral/revealing Spendor D7.2’s. The May and D7.2’s might just be a too much of the same ilk, or maybe I just prefer the slightly smoother/fuller/organic sound of the Lampi tube DAC with my system, giving up very little in other sonic measures. When I first plugged in the Lampi after owning a Schiit gumby I had a huge wow moment and was surprised by how much better it was. Not so with the May, though the gap was obviously not as big. But I do get why many think that the Holo May is a reference level DAC.

Still going to let the May run with music for another week or two before deciding. I do think it has improved a bit since I first plugged it in 100 hrs ago. Will follow up again in a week.

I should provide an update after another 10 days of testing the May alongside my Lampizator Amber 3. In straight head-to-head comparisons there are pros and cons to each. But where the May has shined for me is with HQPlayer upsampling and digital volume control. It’s extremely low noise floor certainly allows for a lot more flexibility and tuning in this regard and obviates my desire for a better analog volume control/pre-amp in between DAC and amp. But the Lampizator is absolutely no slouch at all and I would still like to hear the new Baltic 3.

One other factor is that my amp is single ended and after some reading and talking to others about the Holo architecture and such, I am convinced that I am not getting the most out of the Holo May. The May actually only uses the output of one side of the R2R ladder for RCA outs, and Jeff Zhu even says the balanced outputs sound better. To this end, I have ordered a quality balanced to SE transformer, the Black Box from AmpsandSound. It is basically designed for this exact problem with many current DACs and the recently updated version Justin is sending me has an additional stepped attenuator to dial in the excess gain. Will update with my observations once it’s in the chain this week.

@dvdboulet

So with that in mind... is there an affordable preamp or "analog volume attenuator" that is affordable? It would seem that in its most pure sense all I’m really after is a way to attenuate the sound without degrading the quality (i.e. what’s basically happening with the volume control of the wavelight)... any such device? Or is an audiophile stuck having to buy a $$ preamp to basically serve that purpose? Any input appreciated :)

If that’s all you want then why not just use the digital volume control in Roon or HQPlayer?? That’s what I have successfully done while testing the May into the pre-in of my Line Magnetic amp. It has provided an incremental gain in SQ over the somewhat basic/mediocre pot volume attenuator in my integrated. The May's extremely low noise floor combined with 64 bit volume control provides for tons of headroom.

@jcarcopo I go back and forth with HQPlayer - not all of the filters are better than pure NOS. I am really loving the recently added sinc-mx with LSN15 upsampled to 768hz pcm. I also have a $140 intel pc stick arriving this week that I will use as an endpoint so I can try 1.536khz. Really trying to see how much I can wring out of this dac.

I don't think USB cables should be making a difference given the May's implementation unless they are simply out of spec. I use a Supra USB cable without issue. And again, I have not noticed a difference between 3 low-powered (and LPS powered) endpoints.

@nquery did you tried the EC modulators as well? do you still like the pure NOS over EC mods with HQPlayer?
I ask because per your post I noticed you are using PCM not SDMI’m just curious that’s all

@luisma31 I have not yet as my current core server is not powerful to do it justice and I also want to stick with PCM upsampling first to dial in the sound before making comparisons to DSD. Right now, I am really liking the new poly-sinc-gauss-long filter that Jussi just added. It’s a filter that he has been working on for 6 years, It’s supposed to be the best of both worlds. It sounds fantastic, better than pure NOS.

I am also using a cheap intel pc stick as my endpoint so I can do 1536kHz. This was also a test of the May’s USB input to see if it can back up claims that a fancy USB chain is no longer required. I have had a Sonore OpticalRendu and Allo USBridge Signature in place before this. So far, I can’t tell a difference which is great. One can not worry about USB source as much anymore. That said, I still went with a low-power endpoint and am using an LPS with it.

The other change I made (and mentioned above) was to get a quality Balanced XLR to Single Ended RCA transformer (cinemags) and see if that improved the sonics on my SE amp. I think there is an incremental improvement but haven’t really A/B’ed it yet.

Along with another 400 hrs of burn-in and ability to do digital volume attenuation, I am now much happier than when I first plugged the May in a few weeks ago. I do sometimes miss the organic mid-range and spatial depth of my tubed Lampizator but in most other aspects I am getting better sound from the May.

@luisma31 no you can't do 1.536 with the Pi. Basically has to be Intel USB. Also note you will need to black flash to 30.12 firmware if you want to try 1.536. But it was easy to flash the firmware to 30.12 and then run the HQP NAA image off of a usb stick. 

@luisma31 Pi's are their own architecture/hardware. As far as Holo firmware vs chipset drivers, I think it's a mix of effort on both sides of the equation. Right now, the only combination that really works is 3.12 Holo firmware + intel USB chipset.

@redwoodaudio as per my prior posts above, I have been comparing the May against a Lampizator Amber 3. In pure NOS mode with no external upsampling I generally would agree with you - the Lampizator held its own and was better in that natural/organic way that is obviously part of the tube output, even if not strictly neutral. However where I have found the May to shine is with the ability to feed it better upsampled/filtered audio via HQPlayer and the huge SNR = headroom for digital volume control and DSP. There is a ton of flexibility/tuning available there. With the Lampi the gains of upsampling weren’t as obvious and they generally have very poor noise levels. Bass was definitely not as good. My current pipeline to the May now sounds better than the Amber 3 in most aspects. That said, the organic/natural sound is still slightly missed and I will probably give a higher end Lampi a tryout at some point - Baltic 3 or Atlantic TRP. 

@eckie77
I didn’t notice a big initial difference with the Balanced to Single Ended transformer to be honest but I also haven’t done any A/B testing yet. I put it in and have left it there. I think that the reason that the Lampizator had a more “holographic” sound was due to spatial depth provided by the tube output stage. That is really the only thing I miss from it. But nothing to do with balanced/SE.

I think that Jeff Zhou’s comments are true in general, but the May (and likely Spring 3) are in fact ideal candidates for digital volume control under the right conditions. The May has a very high SNR which allows for a lot of extra headroom when doing digital attenuation. It also has a very low output impedance for driving an amplifier. But you need the proper dsp setup using HQPlayer or similar with a high upsampling rate + noise shaper and ideally an amplifier that’s not too hot (lower gain) so you can minimize attenuation. In my case, the pre-in input on my amp has a 10gb attenuator built in so I can play within the top 20db. And I am using HQP to upsample to 1536 PCM with proper noise shaper into the May and thus already doing headroom attenuation - adding an extra variable 5/10 db’s doesn’t hurt whatsoever. To my ears, this sounds better than the Alps volume control on my amp (via regular line-in). But again, this only works well under the right conditions. eg. this was something that was simply not even remotely possible with my Lampizator tube dac. See this thread/post from Jussi of HQPlayer - https://community.roonlabs.com/t/death-of-the-pre-amp-and-hqplayer-digital-gain/18992/17

Speaking of power (cords), how many folks are running their May through a line conditioner or similar? Like some other posters elsewhere, I find that it sounds better plugged directly into the wall. The slight hint of grain/sibilance is removed. 

You use the pre-in on the LM amp so the LM volume pot is removed from the equation. And then you can set your max volume level in HQPlayer to prevent it from going too high. You can also set ’comfort level’ in Roon which requires manual override to go higher.

Alternatively/additionally, as mentioned before, I also bought a high quality passive transformer with stepped gain control to take the balanced outputs from the May into the SE input on the LM. Aside from slight increase in SQ, the stepped attenuator allows one to dial in max gain into the amp. Which in turn allows one to run HQP volume control at the top of the scale. But honestly, I don’t know if this is worth the extra $$ and box/cables. https://ampsandsound.com/products/black-box-xlr-to-rca-balanced-converter-with-input-transformers-an...

Bonus idea: I then installed free Ropieee.xl on an cheap rPi endpoint and added a $20 FLIRC IR usb plug. You can them use Ropieee remote function to send mapped Holo Remote control volume buttons to Roon/HQP. The funny thing is that I am not using this endpoint for NAA output like most - just as a way to add volume control via the Holo Remote control.

The end result is that I can actually control volume with the Holo May remote in a very well defined gain range. It provides for a nice integrated setup and one would be none the wiser that it wasn’t coming from some sort of pre-amp setup.

I sometimes question whether these extra 2 boxes are worthwhile ... but even without them, I already do some room correction in HQP and I do think HQP volume control sounds better than LM volume pot. And the most recent benefit is that HQP has added automatic loudness adjustments which increases the low shelf as volume is turned down - absolutely fantastic for low-level, late night listening. All in all, this is the best digital sound I have had in my system.

I have owned both. It is definitely much better. No comparison.