Certainly. It’s important to voice equipment and code on a variety of different speakers/systems. I use the IRS, Harbeths, and my personal designs whenever I’m voicing or choosing compiles.
1 Like
If there’s one thing that stands out in my system, it is Snowmass’s ability to properly communicate the presence of minute pressure changes due to vibrations.
For example, vibrations in a singer’s throat, tambourines shaken to and fro, guitar strings being plucked, congas being pounded on, they all come through with amazing life-like honesty, realism and “air.”
Instead of simply hearing the sound of a singer’s voice waver during sustained loud or high notes, I sense changes in the atmosphere as a result of those vocal chords wavering, as though I was with the singer in the recording environment.
Instead of simply hearing the sound of a tambourine being shaken, I hear one physically present in my listening room. And I can clearly tell if the shaking motion is forwards and backwards or side to side. It’s uncanny!
I have never have I heard such realism out of a DAC, and I certainly did not hear this with Redcloud. Only Snowmass was able to bring out these qualities, and I find myself sneaking in more and more time to listen, much to the behest of my wife. Well, at least she knows exactly where to find me! 
6 Likes
Nicely put. I’m enjoying the great “air” too, but I can’t put my finger on another (if it is a different one) quality that may the combination of the lower noise, greater linearity, etc.
The quality is a kind of tube sound. Seems that the DS had that quality from the beginning (one of the things I liked and why it is still in my rack – the previous two DACS had tubes). Redcloud made the quality more obvious and Snowmass even more.
For those who purchased the DS for $4k with Snowmass loaded – congratulations, you got a smokin’ deal for a truly spectacular DAC. Is there anything close at that price (or at $6k) that also can be run straight to amps?
Indeed this is one of the things that let me know that I was on the right track. Hearing small bits of vibrato I never heard, small bends in pitch on various instruments, To me it conveys their emotion much better and (at times) demands my attention.
6 Likes
Exactly! These are the micro-expressions (dynamics, timbre, pitch, transients, etc.) I keep harping on.
2 Likes
Ayon dac’s and CD players, which have tube output stages and can be run directly into amps. My previous CD player was the Ayon 07s, which is at the lower end of their line. I ran it directly into my amp, and it sounded wonderful, although nowhere near as good as the DSD and DMP, in terms of soundstage, detail, and resolution. But the tonality was spot on. Thought about trying some of Ayon’s higher end products, but I’m so happy with the PS Audio gear, I’m in no hurry to go anywhere else. Plus there’s the added advantage of free upgrades, and my close proximity to Boulder, home of PS Audio.
1 Like
I hear them and I FEEL them, @tedsmith. That’s what I find so amazing about Snowmass.
Here is an album that demonstrates it nicely.
Track 6, “Dance With Me.”
On other DACs I’ve heard, some of them quite expensive, when Earl plucks the strings of his acoustic guitar more forcefully, it merely sounds louder. But with Snowmass, i experience a tangible increase in air pressure in the room as a result of the more forceful plucking, as if Earl and his acoustic guitar was right there in front of me!
How you came up with the complex math to achieve that effect is a tribute to your genius! Thank you for Snowmass. Really.
3 Likes
interesting. Ayon was on my list when I purchased the DS, as a trial, instead. First OS was good, and 1.2.1 sounded just right in my system and to my ears. I assumed there would be another upgrade or two. Didn’t expect so many new OSs and the continuing effort to get the last drop out of the design as implemented.
How many DACs get better with age?
2 Likes
there’s nothing close and its even better with a good preamp…
1 Like
OK I got one… success for Snowmass (and Ted!).
Ted, your Snowmass just made Patti Smith’s productions less crappy. I am listening to Wave… yes, the production is crappy… but now it is … well… hmmm… how about “less crappy”.
She is one of my top 10… nothing like her.
Peace
Bruce in Philly
2 Likes
I did the Snowmass upgrade yesterday on my main system. After a few hours of listening I agree with the general assessment in this discussion of the sonic characteristics, more detail, separation, sound stage moves forward, brighter and bass slam increase.
However, on my main system system Redcloud Senior sounds much better so rolled back this morning. I think this has a lot to do with the room which has a bass resonance issue due to many glass walls and a stone floor (even with a rug). Redcloud is so smooth and sweet it just works perfectly in this room. For those interested the chain is Roon > DSD > VTL 5s2 > Bryston 3BSST2 > Vienna Acoustics Imperial List speakers. Cardas Golden Ref cables everywhere. Also use a PSAudio P5 for power clean up.
That said, I as an engineer I am so impressed with Ted’s ability to program FPGA’s and make these changes and improvements. I also guess that on my second system (office) which uses a DSD Junior would benefit greatly from Snowmass because it sounds a bit too restrained after the amp change I made earlier this year.
Regards, Dan…
Hi Ted, thanks for the update. I have also noticed less bass. I have noticed less bass dynamics and impact. Question the version of of Snowmass I have is v3.0.0. Are there later versions? If so, how do I get them?
Analytically Snowmass’s bass is just the same level as Redcloud. The phase is more accurate which can be either louder or softer depending on your position, and the speaker’s positions relative to your room.
If you haven’t already consider adjusting your system as if you have new speakers, check all of the things you used to dial in the base in the past, e.g. try changing the toe-in, listening position, speaker position, sub position, phase setting on your sub, etc.
Some people that initially complained about bass discovered they could get better bass than they could with Redcloud if they spent a little time setting up the system again.
1 Like
Variable phase dial on my subs is/are invaluable. Arguably the most important.
I agree in my listening room, the bass level didn’t change but the phase and coherence did. I reduced phase/delay in both subs but one more than the other and now have better, more articulate bass than with Redcloud.
1 Like
Dan,
If you know someone with a mic’, a laptop and REW software they can run a sweep to work out the room modes then create a convolution filter to put into Roon. I have some “gentle” filters that I use in my room to fix some peaks. I don’t bother trying to fix the nulls as I can really hear them and it chews through amplifier power.
As with anything, there are best not overdone but they can bring a better in room response without room treatments (though of course room treatments are the ultimate solution).
Mark
Hi Jim, I think you hit it on the head - playing the room and spending sufficient time to get all of the aforementioned right can have the most dramatic effect on sound quality. I’ve found Jim Smiths book/DVDs to be quite helpful in this regard - “Get Better Sound”. There are probably other references folks on this forum can recommend as well. I’ve also read many discussions regarding the use of Golden ratios for speaker placement and room design to account for acoustics (I.e reflections). It can be a time consuming process but one that can help you avoid throwing many away on snake oil or unnecessary tweaks/upgrades. My gut tells me that once you get the basics right that only minor adjustments would be required to optimize room settings/alignment/etc around each successive firmware upgrades to the DSD. (because the basic room characteristics never change - only what is coming out of your speakers)… I would be curious about the thought process you go through in addressing subsequent upgrades. Do you take purely an empiracle approach to authoring new firmware (I.e. use your ears - very subjective), or is there an exacting theoretical basis for processing/filtering that must be applied to achieve the result you are striving for? Maybe a little bit of both😊
After I know what I’m going to work on the work itself is informed by doing the best technical job I can at that juncture. Then it’s, “Trust, but verify.” I’ve made great progress by assuming things like a ruler flat response works sounds best, etc. I.e. that no tweaking of the technically correct approach is needed for best sound quality.
Figuring out what to work on next is the hard part most of the time. Paul often askes for something that seems impossible at the time. There’s the feedback I get here on the fora for each release that lets me know what could be improved. There are always ideas that I’ve implemented in the past that I can revisit with more time or more knowledge about tradeoffs or simply more space that’s opened up in the FPGA. And last, but certainly not least: I stare at the wall and think about the thing that’s the most likely to get the biggest change in sound quality or that will address the goal for this release the best.
6 Likes
I often thought we are in an extremely lucky situation that someone like you works on these improvements who probably works far beyond normal work times and self-driven.
I think all this would never happen within a normal 9-5 company situation and that it’s why it doesn’t happen anywhere else in fact. No one could pay this. Thanks again Ted!
tedsmith
Chief digital guru
November 26
gorm.yoder:
Do you take purely an empiracle approach to authoring new firmware (I.e. use your ears - very subjective), or is there an exacting theoretical basis for processing/filtering that must be applied to achieve the result you are striving for? Maybe a little bit of both😊
After I know what I’m going to work on the work itself is informed by doing the best technical job I can at that juncture. Then it’s, “Trust, but verify.” I’ve made great progress by assuming things like a ruler flat response works sounds best, etc. I.e. that no tweaking of the technically correct approach is needed for best sound quality.
Figuring out what to work on next is the hard part most of the time. Paul often askes for something that seems impossible at the time. There’s the feedback I get here on the fora for each release that lets me know what could be improved. There are always ideas that I’ve implemented in the past that I can revisit with more time or more knowledge about tradeoffs or simply more space that’s opened up in the FPGA. And last, but certainly not least: I stare at the all and think about the thing that’s the most likely to get the biggest change in sound quality or that will address the goal for this release the best.
6 Likes
Curious about that… As you get more familiar with the problems does that simplify the use of the configurable logic blocks - and thus use less of the FPGA’s capacity? Did you recover space when you hand wrote the current filters? Or do new/different capabilities use more resources. I could imagine that at some point you’d run out of space in the FPGA if you kept adding. I’d also imagine you’d need some extra space in any FPGA because otherwise you couldn’t compile a good sounding version. Do you still have sufficient headroom to produce more updates? Or will most of the really exotic changes need to go into your signature DAC that has a more recent (bigger?) FPGA?
Software is a gas, it expands to fill the space available.
As time goes on I sometimes find better ways of doing things, but at other times when I finish one project it clears the way for another. In the case of the upsampler rewrite I used a few more multiplier blocks but fewer memory blocks than the old upsampler. With some more work I can halve the number of multipliers so that I use fewer resources overall for the new resampler. I just didn’t want to take the time to write the code with fewer multiplies and debug it given that it would delay the release of Snowmass.
I have always been able to shuffle things around enough to add every feature I’ve done without much problem. If the new code isn’t smaller than the old code I can usually find some other code that uses more resources than it needs.
The real problem isn’t exactly space, but the amount of time it takes to find a valid placement and routing. When there’s too little extra space available the tools can take a lot longer to do a compile (e.g. a couple of hours instead of five minutes.) It helps that the tools tell you where the bottle necks are so I usually have a pointer to a place to attempt to save some resources.
4 Likes