CharlieMB

I understood this to mean that Quadro K (e.g., K6000) can't be compared with confidence against Quadro P. As a corollary it would follow that Quadro and GTX families can't be compared with confidence either. Thanks

To shop for a card, is it better to focus on FP64 performance over number of CUDA cores? The two are usually related, but sometimes the FP64 performance surprises, especially when looking at older cards.

Wow. Yesterday I was able to go from Redbook 44.1K / 16 to DSD512 using Closed-form-fast and ASDM7 512+ fs. Again, this was with a Quadro K620 and "auto rate family." Closed-form might have also worked, but -fast sounded better.

Thanks. Yes I only changed the modulator. Come to think of it, other than to simply try it, I don't know why I would like to play with ASDM7-512+ when I'm only running DSD256. Anyway, it does work with lighter modulators, IIRC. Thanks. Noted. BTW, I have the Spring 2, KTE, which does any rate. Still, my intention is to always stay within family to reduce load. Ahhh, thanks. Your answer is important because it allows me/us to buy a 2nd card and know it will be used. Does your 1080 get loud during playback? (I have the same ASUS ROG Strix GTX 1080 for my workstation and it's too long to fit in my playback machine.) Thanks again

Yes, I meant GPU memory. Thanks. So then I assume 4GB should be safe, and 2GB iffy. Right now I have 96/24 PCM to DSD256 running Poly-sinc-xtr-2s and DSD7-256+fs with only a Nvidia Quadro K620 and Intel i7/3930K @ 3.2 GHz. Redbook also works at those settings. This is the best sound I've had to date. This card is only 24 GFlops of FP64 performance and what's nice is it only burns 45 watts. But it won't do these settings with ASDM7-512+. I'd like to try something around 5x that performance or ~ 125 GFlops, which is about 1/3 of a GFX 1080, and less power consumption than a 1080. The purpose is to try more of the one stage filters or get further toward DSD512 (some work at 512 even now). From this tread I know that I need to look for graphics cards / drivers supporting CUDA 3.0 or above. If I install a second card, will HQPlayer allow me to choose which card to offload to? And if not, will it pick automatically the more powerful card? Thanks again, and in advance

Does the amount of memory matter? 3, 6, 8 GB ? Perhaps more memory is not better and adds heat for nothing.

I have to walk this back. The answer is no as it turns out, at least not for now. When I removed the ISO Regen the sound improved with this firmware upgrade, 3.12. This has never been the case since I've owned the ISO Regen, which has always improved things. Yesterday and today were the first time the opposite was true. I partially attribute this to another change I made recently but never tried without the ISO Regen and that is moving my wireless keyboard w/ touch pad from the USB port. The only thing on the USB port now, in my whole system, is the Spring 2 KTE. Both my keyboard and mouse are now wired and connected with PS/2 cables to the PS/2 ports on the back of my music server (XXHighEnd PC). Removing the keyboard/mouse from USB made a big change for the better all by itself. I had such faith in the ISO Regen that I never questioned it. Normally I run the ISO Regen off of a battery, since the included power supply is easily improved upon. The ISO Regen is sensitive to its power source. So it wouldn't surprise me if I tried a better power source that things could change again. So I have that to explore or the Intona Industrial for a next step. This is after two days of listening. If things change I'll report back

I get the impression that the answer is Yes, it improves further. But I have not switched back to confirm. I will eventually switch back but for now please assume that it improves things.

This dual USB firmware and driver update has resulted in a hugely significant improvement in sound quality in my system. It's amazing to me that software updates can make this much of an improvement. [To give you an idea of the magnitude of the improvement] To update the firmware I had to remove my ISO REGEN, per instructions. I listened last night without the ISO REGEN and can say that my system sounded better after this update without the ISO REGEN than it did before with it. And the ISO REGEN had always provided a big improvement, as you may have read, and such was true in my system. If I had to state a reason for the improvement, I'd guess that USB's 8K packet noise is further suppressed with this update, somehow. Yes, hat's off to Tim and Jeff. Big thanks guys.

Thanks Matias

Not yet. The goal is complete isolation so I like the idea of a simple 6-cell pack of Eneloop AAs (NiMH) yielding 7.2 volts. I've got plenty of them lying around already. I can attempt this because I know the load of my DAC is light as seen from the 5V USB input. My DAC is the Wyred 4 Sound DAC2 DSDse.

I'd like to report a very positive experience with the ISO REGEN. Wow!... Inserting the ISO REGEN into my chain with the provided power adapter and the included "USPCB" USB A>B adapter has made a world of difference in sound quality. This report comes after only two nights of listening. Two nights is not normally long enough for a listening impression but the improvement is not subtle at all and worth reporting. I clearly hear more details in the music, extended highs, more realism; everything is more palpable. The instruments are closer and more spread out in the room and they are truer sounding. I had planned on getting the follow-on product to the original REGEN for a long time. I delayed because prefer simplicity and have been avoiding the extra components and especially an additional power cord and power supply. I can easily say now that it is well worth the extra components. I'm also very happy with fast shipping. I ordered it Wednesday of this week with two day shipping and it arrived overnight, much faster than expected. Hats off! Many thanks to Alex Crespi for a fine product.

In theory, yes. In practice? No, this is supposed to work. Glad to read that someone else has experienced this behavior. I see it too. With JRiver, it happens almost every time I switch between DSD 64 and 128. I've seen it also with HQPlayer, although much less often, like only two or three times in total. Along with the faint sinus and noise, I hear low-level music playing at about 1/2 or 1/4 the rate. In other words, faint music in slow motion; as if the data stream were DSD 256 or 512 and accidentally played at 128. Over here the fix is easy. I just click stop, and click play again. This solves it every single time. From that point in, ensuing tracks play correctly at 128. Try it. It only takes four seconds. Nope. I play Redbook CD as 2xDSD through the ASIO driver all of the time.

I am maybe 60 hours into the break-in period and at this stage I have to concur with jt25741 that for DSD 64 anything higher than default of "<50KHz" is objectionable and suboptimal. There are four choices: <50K, 50K, 60K, and 70K. I find even the 2nd setting of 50K to be suboptimal. "<50K" sounds great. But with the other three settings I hear a very high frequency electronic-sounding signature that must be related to the noise shaping. DSD 128 should improve this by a factor of 18db, which is like a ratio of 1/8 (meaning almost 10x lower). But if I recall it did not help that much. I have to recheck to confirm. Interestingly, the DSDse manual only lists the filter-behavior curves for the 50, 60, and 70 KHz filters, and does not show the filter behavior for the <50K setting, the one that works best for me. Speculating, perhaps the <50KHz setting is a Wyred specialty or secret sauce. So jt25741 and I are hearing the same thing with regard to this and with regard to the goodness of PCM Roll = SLOW. Another possibility, of course, is that after the full break-in period, the 50, 60, and 70 settings will become usable. I was actually preferring 60 for a while, until I noticed the signature. I interpret what jt wrote above to mean that, if I for example use HQPlayer to upsample to DSD 128 using highly sophisticated filters, then not much can be done about this signature because it is being added later in the chain, after the fact. This remains to be seen (heard). Whereas, the same does not appear to be true for PCM at 384. Upsampling PCM does not have this issue.

I find 384K8 closer and wider, more intimate and more intense; you could say more vivid and certainly richer harmonically (well, depending on the upsampling filter). Both PCM and DSD are deep, but DSD is probably deeper. DSD sounds a little more analog, I'd say. Both have their place and for me both are favorites, if that's possible. DSD 128 sounds more analog than 64. Over here, both PCM and DSD are wider than the speakers, by far. But PCM has the appearance of being wider probably because it presents itself as being closer, or starting out closer. Today I had the house all to myself and I listened for over 7 hours, mostly straight hours. I experienced not even a hint of fatigue. With only ~50 hours of break in, this DAC surely does not sound like it is breaking in. But I'm sure it will continue to change somehow. I'd be happy if it settled right where it is.