exaSound e18 - e20 - e28 - Info and Experiences Post All Here

[LewinskiH01]

Sounds like the best bet will be to go back to the exaSound ASIO drivers vs. the third party ones.

 

I tried the third party route once with the exaSound e28 and found the performance and sound was better staying with exaSound's drivers. It may be the story here as well.

 

Not sure why you say I'm using third-party drivers. I'm using exaSound's G3.

[LewinskiH01]

The problem might be AudioPhil's optimizer under clocking your CPU to 2.2 GHz and not allowing it to speed up to 3.5 Ghz when beneficial.

 

Have you tried something like CPU-Z to monitor the clock speed during playback ?

 

you may also want to try Foobar with SACD plugin that sounds significantly better than JRMC doing Redbook to DSD 256 (make sure integer mode is selected).

 

The sound quality champion is definitely HQ Player with its Polysinc filter. HQ Player is not compatible with some, if not most of AudioPhil's optimizer's OS tweaks

 

Good point. I didn't use CPU-Z or anything like it, but I did watch the resource monitor while playing music and indeed speed was staying at 2.2 GHz. Was it because it didn't need more speed or because it wasn't allowed to? i don't know. But it's easy enough to uninstall the Optimizer and see how it goes.

 

Regarding HQ Player being the sound champ: is this in the context of exaSound DACs? I'm not attached to the Optimizer to the point of being unwilling to try something better.

[john57]

Your math score is quit low. One things you can do with J.River is to set the buffers to 500ms like this:

[LewinskiH01]

I tried uninstalling the Optimizer but it didn't help. Nevertheless, EuroDriver might be onto something: again I kept Task Manager open while I was playing music and noticed that without DSD upconversion speed oscillated in between 2.15 and 2.25 GHz, yet during DSD upconversion it was dead fixed at 2.42 GHz, with 2.49 GHz being noted as the maximum speed. So it does seem like something is limiting speed indeed. Any further ideas as to what could be?

 

On a different note, today I took some time to evaluate the e22 vs. Audiophilleo + Metrum Octave. These impressions are by no means final, but bottom line I ike the AP+Metrum better and they are less expensive.

 

My server is optimized for the Audiophilleo, which also has the PurePower option. I did the comparison with AudioPhil's Optimizer installed, and Windows Server 2012 in GUI mode (Core mode sounds better, but more cumbersome for the comparison). I also had a screen connected, which I usually don't have and that yields even better results.

The Audiophilleo is fed from a Paul Pang USB card and thru an inexpensive USB cable (Ultraviolet) that is the weak link in that chain.

The Metrum Octave feeds from a dedicated power line, without power conditioning.

The e22 feeds from the supplied SMPS.

 

The Audiophilleo with PurePower retailed for about $1k when purchased, and so did the Metrum Octave (Mk.I). These were purchased 2 years ago. I was expecting a $3.5k current DAC to trounce them, but it just isn't happening for me. Both options sound very good. The e22 exhibits some sibilance/glare, especially noticeable on female voices. It might have better bottom end. Listening was done with 16/44.1 material and some 24/192.

 

Of course I'm not calling the comparison finished. And I'm hoping to be floored with DSD after everything I've seen written. In a way it might not be a fair comparison for the e22 as it is designed to work with unoptimized servers. Maybe accounting for the price of the server plus USB cable plus DAC would be more apples to apples, and there the cost of the whole chain would be evened out.

[Miska]

I tried uninstalling the Optimizer but it didn't help. Nevertheless, EuroDriver might be onto something: again I kept Task Manager open while I was playing music and noticed that without DSD upconversion speed oscillated in between 2.15 and 2.25 GHz, yet during DSD upconversion it was dead fixed at 2.42 GHz, with 2.49 GHz being noted as the maximum speed. So it does seem like something is limiting speed indeed. Any further ideas as to what could be?

 

OS power saving functionality is trying to keep CPU frequency just high enough to perform the tasks at hand. If the CPU load is not approaching 75%+, the CPU frequency usually doesn't reach 100%, not at least constantly.

[EuroDriver]

yet during DSD upconversion it was dead fixed at 2.42 GHz, with 2.49 GHz being noted as the maximum speed. So it does seem like something is limiting speed indeed. Any further ideas as to what could be?

 

on my PC the clock speed was limited by a BIOS setting

 

On a different note, today I took some time to evaluate the e22 vs. Audiophilleo + Metrum Octave. These impressions are by no means final, but bottom line I ike the AP+Metrum better and they are less expensive.

 

on 16/44 playback which the E20/E22 performs with no upsampling, my T+A DAC 8 trounces my E20. Not so surprising as the DAC 8 is upsampling to 24/384 and using a Bezier filter with "rininging amelioration"

 

the Exasound DAC's are a good if not the best DAC option if you are wiling to go down the road and up the learning curve of SOTA playback software with format conversion

 

Regarding HQ Player being the sound champ: is this in the context of exaSound DACs? I'm not attached to the Optimizer to the point of being unwilling to try something better.

 

I have heard HQ Player converting 16/44 to DSD128 playing to an OPPO 105D with and without HQ Player's NAA running on a SOTM SMS-100 and on an Antelope Zodiac Platinum which internally upsamples from DSD128 to DSD256 as well as on my E20

 

the relative ranking of software players was the same, JRMC - good, Foobar + SACD significantly better, HQ Player better still. The difference the NAA setup made to the OPPO set up was not subtle, but NAA made no audible difference to play back thru GRIM LS1 digital active speakers which is perhaps SOTA in its data stream noise isolation

 

the sound quality with HQP converting to DSD256 playing to the E20 and HQP Player converting to DSD128 and playing to the AZP which then upsampled to DSD128 was a very hard to distinguish toss up, both were superb with a similar flavor, and stunningly realistic

[LewinskiH01]

Your math score is quit low. One things you can do with J.River is to set the buffers to 500ms like this:

[ATTACH=CONFIG]14697[/ATTACH]

 

Hello John.

 

Indeed, increasing buffering to 500 ms got rid of stutter. Not sure this was the cause or the added buffer is acting as a band-aid to mask the real issue, but at least it's working. I tried this late at night and didn't have the opportunity to critically listen yet, but it didn't jump at me as a significant improvement on sound quality.

 

thanks for the tip!

[LewinskiH01]

the Exasound DAC's are a good if not the best DAC option if you are wiling to go down the road and up the learning curve of SOTA playback software with format conversion

 

Those are big words. Posts by experienced users who have proven to be brand-agnostics do not support this statement. And my personal experience so far suggests this is not so either. But I am willing to be proven wrong - actually interested in being proved wrong. After all I bought the e22 with high expectations.

 

What I'm understanding from the above is HQPlayer sounds better than JRiver. I have also seen comments saying the interfase isn't as polished, and not having a JRemote type application to operate HQPlayer headless would definitely be an issue for me, but let's stay with the concept. But are you saying the improvement HQPlayer brings to an exaSound is larger than to other DACs? Or maybe it's HQPlayer's way of upconverting that sounds better than the way employed by JRiver?

 

I was under the impression exaSound's buffering inside the DAC would make the DACs less prone to unveiling differences in sound from playback software, just like they are lesss immune to differences in hardware or operating system.

[john57]

Hello John.

 

Indeed, increasing buffering to 500 ms got rid of stutter. Not sure this was the cause or the added buffer is acting as a band-aid to mask the real issue, but at least it's working. I tried this late at night and didn't have the opportunity to critically listen yet, but it didn't jump at me as a significant improvement on sound quality.

 

thanks for the tip!

 

Stuttering is a common symptom of buffer underrun issues. It can happen with video as well with audio. Increasing the J.River buffer to 500ms is recommended for most cases when playing back DSD files. This issue does happen with other DSD DAC's as well. I do not think of it as a band-aid just a limitation somewhere in the system. I think it is a combination of application and hardware that is causing this underrun somewhere in the pipeline from the DSD file by the time it reaches the DAC by the way of USB transfers. For some reason playing back DSD files is more resource intensive than some users realized. I just updated my PC to a Intel I7 system with higher speed memory because even converting PCM to DSD256 can overtax some slower system setups. I wanted to make sure I have enough headroom for smooth playback and avoid problems with processes that will cause short spikes in peak loads on the system as a whole. PC's like to use packet interrupt delivery of data and just takes a bit of work to get it stream smoothly. I also setup J.River for one second silent sync on start of playback. If you using DoP for DSD playback, setting J.River to play silent on DoP stop can help in some cases. You just have to find the optimal setting for your setup. I pay attention to the number of process and start-up items on my PC as well.

[exa]

If you use an exaSound DAC, there is an easy way to identify buffer underrun conditions. When the internal hardware buffer gets empty, the DAC display shows "---" instead of the sampling rate. If you notice blinking on the display during stuttering, the issue is slow or interrupted data flow over the USB connection.

[Stereolab42]

I was under the impression exaSound's buffering inside the DAC would make the DACs less prone to unveiling differences in sound from playback software, just like they are lesss immune to differences in hardware or operating system.

 

No, the point of exaSound's technology is to eliminate issues caused by jitter or noise, from non-software sources. No matter what DAC you have, if your playback software can't generate bits fast enough, you're going to get dropouts. (A large software-side buffer can paper over CPU spikes to some degree.) HQPlayer is going to be even harder on the CPU, depending on what algorithm you choose. Honestly, you need a faster computer if you want to play around in this arena.

[Miska]

Xeons should be pretty fast, but there could be some other hardware blocks or device drivers causing issues (commonly Bluetooth/WiFi drivers). Thesycon's DPC Latency Checker helps checking out that the hardware and OS are performing as they should:

DPC Latency Checker

 

Plot should stay green...

[Synfreak]

... Thesycon's DPC Latency Checker helps checking out that the hardware and OS are performing as they should ...

 

Attention..., the DPC latency checker doesn't do very well under Windows 8(.1).

[bmoura]

Stuttering is a common symptom of buffer underrun issues. It can happen with video as well with audio. Increasing the J.River buffer to 500ms is recommended for most cases when playing back DSD files.

 

Hmm, I'm using JRiver 20 with the exaSound e28 and it is working fine with DSD files using the default 50 millisecond setting.

[john57]

Hmm, I'm using JRiver 20 with the exaSound e28 and it is working fine with DSD files using the default 50 millisecond setting.

 

I am not using a exaSound DAC and on three of my DSD DAC's with the older AMD 4 core chip I had to increase the buffer size when playing DSD256 otherwise I will get stuttering once in a while doing other things on the computer.

[bmoura]

I am not using a exaSound DAC and on three of my DSD DAC's with the older AMD 4 core chip I had to increase the buffer size when playing DSD256 otherwise I will get stuttering once in a while doing other things on the computer.

 

OK, that's a key difference. I'm using an i7 based laptop plus the hardware buffer in the e28. Two advantages there.

[EuroDriver]

But are you saying the improvement HQPlayer brings to an exaSound is larger than to other DACs? Or maybe it's HQPlayer's way of upconverting that sounds better than the way employed by JRiver?

 

IMHO I am suggesting both. There are not too many DACs that can playback DSD256, ifi micro DSD and the Exasound's , so HQ Players up converting to DSD256 capability can only be taken advantage of by these two DACs?

 

and HQ Payu

[bmoura]

IMHO I am suggesting both. There are not too many DACs that can playback DSD256, ifi micro DSD and the Exasound's , so HQ Players up converting to DSD256 capability can only be taken advantage of by these two DACs?

 

Very true. Along with the Oppo HA-1 Headphone Amp/DSD256 DAC and a few more.

 

exaSound Brings Quad DSD to Mac, Support for Linux and Windows Server Enhancements

[johndoe21ro]

Nice article, Brian. I hope the guys from Audirvana will implement ASIO support too.

[bmoura]

Nice article, Brian. I hope the guys from Audirvana will implement ASIO support too.

 

Thanks. I'm sure that George & the team at exaSound would love to work with Audirvana and the other Mac Audio Player companies on adding that support for their drivers!

[kennyb123]

I'm sure that George & the team at exaSound would love to work with Audirvana and the other Mac Audio Player companies on adding that support for their drivers!

 

I intend to purchase a new DAC in December. The e22 would likely move to the top of my list if ASIO were to be supported by either Audirvana or JRiver. To the best of my knowledge, neither of the two Mac players that support ASIO now can be controlled with an iPad, which makes them of no use to me unfortunately.

 

Surprising that JRiver Mac does not support ASIO, as I thought their goal was to have feature parity between the Mac and Windows versions. JRiver isn't on par sonically with Audirvana in my system, due I think to Audirvana's Direct Mode offering greater transparency. I suspect that ASIO would level the playing field between these two.

[Matt Ashland]

I intend to purchase a new DAC in December. The e22 would likely move to the top of my list if ASIO were to be supported by either Audirvana or JRiver. To the best of my knowledge, neither of the two Mac players that support ASIO now can be controlled with an iPad, which makes them of no use to me unfortunately.

 

Surprising that JRiver Mac does not support ASIO, as I thought their goal was to have feature parity between the Mac and Windows versions. JRiver isn't on par sonically with Audirvana in my system, due I think to Audirvana's Direct Mode offering greater transparency. I suspect that ASIO would level the playing field between these two.

 

 

ASIO on Mac is in the works. We're working with George at exaSound.

[elcorso]

ASIO on Mac is in the works. We're working with George at exaSound.

 

A very wise decision...!

 

Roch

[kennyb123]

ASIO on Mac is in the works. We're working with George at exaSound.

 

Very good news!

[kennyb123]

(duplicate post deleted)