cPlay vs Foobar

[vijwilso]

I am really happy with the SQ from cPlay and would like share my experience .

My setup is like this HTPC -> cPlay(cMP) -> VIA ASIO USB --> Audio GD NFB 3.3 -> Exposure 2010S2 ----> ProAc Studio140MK2

I recently got the Audio GD , and prior to that was using a DIYDAC ( burrbrown ) and for some time with Schiit Modi and using foobar WASAPI as the player . The sound was ok for certain genre of music like HD files from Diana Krell , Nora Jones etc primarily the Jazz , but never liked the Movies songs and other country and Rock files . I always felt kind of bass attached to the voice , voice itself was not being natural and harshness in the high which i can never listen to more than 30 mins and get head ache . I was hearing ( feel Not audible) some kind of noise between the notes in the back-ground , but wanted to have the blackest back-ground possible .
I was blaming on the room size which is small ( that is still a factor ) , but after received the Audio GD which comes with the ASIO driver I was exploring with Various SW players .While foobar was good with the VIA_ASIO output , there is still the harness in the music and it is thin on certain aspect (may be this is called digital music ) . Then I tried the cPlay ( freeware ) along with cMP ( which optimizes the windows by killing unwanted process ) , I felt the sound itself more organic and I could listen to high volumes in the same room ( but still need bigger room to enjoy the capability of the speaker ) , the harshness is gone and back-ground is more dark . I never expected this from a SW player . Even the movies songs sounds better and voice is much better highs are natural . Pls note I am very sensitive to bright sound at the same time I cannot listen to warm sound as it is boring which was the case with the Schiit Modi. So I need all the details at the same time it should not end by in fatigue . I hope, I achieved this with this player . I still to try with the Jplay in Xtream mode for which I need more RAM. Currently my setup is with 2 GB RAM

As part of this I also optimized the windows system by un-installing all un-wanted programs ( It has only windows and XBMC and FOOBAR and cPlay) and also removed all the windwos components like IE etc and disabled all the un-used devices in the device manager , disabled the paging and other indexing services . There is a detailed document in the cPlay site itself . By doing all of these it gives less interruption to the cPlay and cPlay load the complete song into memory , so no disk interaction .

The DPC Latency checker shows a constant ~90 microSeconds. I am not sure if this is good enough or still improvement is possible . One good thing is that the DPC latency never shoot up in my HTPC , however when I tried all the steps in the laptop it periodically shoots more than 1000 micro Seconds . So far it is good with the HTPC as the jitter is not high ( a constant latency )

One issue with the cPlay is that it does not play bitperfect , It re-sample based on the setting . For example , if I play a 44KHz files and setting in cPLay is set at 96 KHz then it upsamples . So for every song , I need to set the right setting in cPlay based on the sampling rate to make it bit-perfect . Also it needs a Cue file for multi-file playback .

I am not sure if I hear much of a improvement with staring cPLay from cMP , but I prefer to use that way as cMP allows me to kill most of the process . Less interrupt is always better .

Though I liked the up-sampling by cPlay (I tried 192 KHZ , SoX setting ) , I feel i am missing the spice and too smooth ( like towards the warm sig ) , but I hope it is good for high volumes playback . It is early to comment as I did not spend much time in this .

So this is definitely a step-up from foo-bar when properly set , I never believed in Jitter theory before , but this changed my perception now

Please share your experience with the SW player and methods to qualitative measure the ltency/jitter and improvement based on it .

 

[AmitKumar]

congratulations for buying new dac. Please share more info about this dac in the forum.

i am very happy to know that finally you are liking the music from your set up after multiple changes.

Best wishes!

 

[vijwilso]

Thanks Amit

I think DAC shows the improvement only when the downstream elements are good .

When i earlier compared the chinese DIY DAC with the Schiit Modi , i could not find any difference with Jamo Entry level FS , the DIY DAC sound loud due to the output levels are high compared to Modi.

Now Jamo replaced with ProAc ( which is more revealing compared to Jamo ) i could see Schiit Modi is lot better than DIY DAC . I could relate the improvement in the sonic signature ( w Schiit ) to the popular reviews in the Web . I liked it for long listening sessions .

Now I dont have Schiit with me to do side by side comparision with Audio Gd , but compared to Schiit, the Audio GD is more detailed ( not bright ) , seperation is better and bass is tighter ( voice clarity is better ) , but i did not like it with the foobar and i easily get into listenin' fatigue . Now the cPlay changed it , retaining the details

 

[matbhuvi]

cPlay is quite an upgrade over Foobar. Especially in the smoothness and texture of Bass. JRiver is quite close to cPlay in performance and hard to tell the difference blindfold. Even though i luv cPlay, i bought JRiver for convenience.

 

[vijwilso]

Have you compared the jPLay with jRiver and cPlay .? I presume the jPlay should be better compared to these , but could not experiment due to the RAM limitation

cPlay is quite an upgrade over Foobar. Especially in the smoothness and texture of Bass. JRiver is quite close to cPlay in performance and hard to tell the difference blindfold. Even though i luv cPlay, i bought JRiver for convenience.

 

[matbhuvi]

Have you compared the jPLay with jRiver and cPlay .? I presume the jPlay should be better compared to these , but could not experiment due to the RAM limitation

Yes, i felt JPlay to be tad smoother in highs than cPlay. Some what better than JRiver as well. But, JRiver wins hands down on convenience. Probably my equipments were not able to reveal stellar differences between the two. It is just humble Xonar STX soundcard and FA-011 headphones

 

[vijwilso]

Attaching some link that helps to reduce the DPC latency . I used this and reduced the DPC latency from 140 microSec to 50 microSec . This should help in for any audio player also .

cMP | CMP / 07Optimisations


Tools to measure the DPC Latency

DPC Latency Checker

Resplendence Software - LatencyMon: suitability checker for real-time audio and other tasks

 

[Thad E Ginathom]

Do not believe all you read about latency! The marketing men know exactly how to tweak the worries of the hifi customer, and tweak it they do. Take all with a pinch of salt --- and remember, that when trusting your ears, you'll still hear what you expect to hear. There's nothing like two hours of tweaking your operating system for getting better sound. It will work. Whatever you do!

DPC latency, on the other hand, is a killer. This is not a subtle effects thing; it can make your PC entirely unsuitable for music listening. If you have clicks, dropouts, etc, dpc latency check is part of the diagnostic toolbox. If you do not, it is not worth worrying about. If you don't have a DPC latency problem, then just be glad and enjoy the music. I can assure you: it is a nightmare that would probably lead to just buying a new machine!

I don't and won't use Windows, so cannot comment on whether there is a best player or not. Even in Linux I currently have an occasional dropout problem which is associated with memory filling up. I'm fairly sure that Firefox is the culprit, but I want to browse when listening, so what to do? PCs are imperfect things.

 

[vijwilso]

Do not believe all you read about latency! The marketing men know exactly how to tweak the worries of the hifi customer, and tweak it they do. Take all with a pinch of salt --- and remember, that when trusting your ears, you'll still hear what you expect to hear. There's nothing like two hours of tweaking your operating system for getting better sound. It will work. Whatever you do!

DPC latency, on the other hand, is a killer. This is not a subtle effects thing; it can make your PC entirely unsuitable for music listening. If you have clicks, dropouts, etc, dpc latency check is part of the diagnostic toolbox. If you do not, it is not worth worrying about. If you don't have a DPC latency problem, then just be glad and enjoy the music. I can assure you: it is a nightmare that would probably lead to just buying a new machine!

I don't and won't use Windows, so cannot comment on whether there is a best player or not. Even in Linux I currently have an occasional dropout problem which is associated with memory filling up. I'm fairly sure that Firefox is the culprit, but I want to browse when listening, so what to do? PCs are imperfect things.

I hear occasional click in the sound ( No dropout) even after tweaking the DPC levels to this level . Setting the ASIO buffer to higher value seems to solve the issue ,however i want to keep the ASIO to as low as possible .

Now the DPC is 50microSec , after un-installing all SWs including IE , Services , i can use this PC only for Audio , nothing else work in the current state . But still this does not look perfect .

Can you share some information on the Linux dist. that you are using . How about AP-Linux which (AudioPhile Linux | Audio Perfection on Linux) which looks more suitable

 

[Thad E Ginathom]

First, my machine is a general purpose machine, and not just used for audio. If I were configuring a small box for the hifi rack, then maybe there might be different decisions.

However, I do not think there is any reason why a general-purpose machine should not play good audio, and, given compatible hardware, do so out of the box.

Essentially, my system is Ubuntu 12.04. I have a firewire interface, and Ubuntu support of firewire audio is not good. This led me to KXStudio, which is available as an independent Ubuntu-based distro, or parts of it can be installed as required.

Whilst my PCI card (or the onboard, but most of us don't want to use that) was, with ubuntu, plug'n'play, the Firewire device was a nightmare until I came across KXStudio. All that stuff about IRQs really matters and has to be taken account of. KXStudio has made that setup fairly automatic.

Firewire devices need a set of drivers called FFADO, and this only works with an audio server called JACK (the Jack Audio Connection Kit) which can, itself, be a nightmare to configure. KXStudio has some excellent tools, and even though I have recently been using a USB DAC, I have remained with KX, Jack, etc. I use a low-latency kernel, as distributed by KXStudio. I hang out on Linux-Musicians forum, where the recent low latency kernels are considered quite good enough, without going for a real-time kernel. Those guys who do music making and recording know when stuff matters and when it does not.

Have I put you off Linux audio? :lol: I hope not: there is no reason that it should be difficult, unless one selects, or chances upon, difficult to set up kit.

The worst thing about Linux audio is hardware support. Most manufacturers don't even think of us, and some do not want to support open-source developers of drivers. The worst thing is, every time I want to even daydream about buying something, I have to spend time with google to find out if it will work for me at all.

Case in point was recent temptation towards a high-end DAC/HP amplifier. I focussed on the Burson Conductor. Would the USB be compatible with linux? Burson themselves told me that it would not. They did say that could use optical, but then I need a soundcard to output the optical. You windows guys would be able to open the box, plug it in and enjoy!

The Windows (and Mac?) world is simpler, just because the manufacturers support it, and with the exception of occasional pieces of Mac-only (or win-only?) hardware, everything just works or comes with drivers.

One sometimes wonders: on a given set of hardware, which operating system would sound the best? Would there be a difference? I honestly don't know --- and nor do I trust my ears/brain to do that kind of A/B-ing with any accuracy, over reboots. Perhaps, sometime, someone will do some properly-set-up blind tests with identical hardware, and I will be very interested in the results. But I really, really, really don't want to go Microsoft again!

BTW: I use Aqualung as my player (available across platforms) because it plays music without doing too much else, is fully compatible with the Jack system, and handles gapless (eg live music split into tracks) playback perfectly.

 

[vijwilso]

First, my machine is a general purpose machine, and not just used for audio. If I were configuring a small box for the hifi rack, then maybe there might be different decisions.

However, I do not think there is any reason why a general-purpose machine should not play good audio, and, given compatible hardware, do so out of the box.

Essentially, my system is Ubuntu 12.04. I have a firewire interface, and Ubuntu support of firewire audio is not good. This led me to KXStudio, which is available as an independent Ubuntu-based distro, or parts of it can be installed as required.

Whilst my PCI card (or the onboard, but most of us don't want to use that) was, with ubuntu, plug'n'play, the Firewire device was a nightmare until I came across KXStudio. All that stuff about IRQs really matters and has to be taken account of. KXStudio has made that setup fairly automatic.

Firewire devices need a set of drivers called FFADO, and this only works with an audio server called JACK (the Jack Audio Connection Kit) which can, itself, be a nightmare to configure. KXStudio has some excellent tools, and even though I have recently been using a USB DAC, I have remained with KX, Jack, etc. I use a low-latency kernel, as distributed by KXStudio. I hang out on Linux-Musicians forum, where the recent low latency kernels are considered quite good enough, without going for a real-time kernel. Those guys who do music making and recording know when stuff matters and when it does not.

Have I put you off Linux audio? :lol: I hope not: there is no reason that it should be difficult, unless one selects, or chances upon, difficult to set up kit.

The worst thing about Linux audio is hardware support. Most manufacturers don't even think of us, and some do not want to support open-source developers of drivers. The worst thing is, every time I want to even daydream about buying something, I have to spend time with google to find out if it will work for me at all.

Case in point was recent temptation towards a high-end DAC/HP amplifier. I focussed on the Burson Conductor. Would the USB be compatible with linux? Burson themselves told me that it would not. They did say that could use optical, but then I need a soundcard to output the optical. You windows guys would be able to open the box, plug it in and enjoy!

The Windows (and Mac?) world is simpler, just because the manufacturers support it, and with the exception of occasional pieces of Mac-only (or win-only?) hardware, everything just works or comes with drivers.

One sometimes wonders: on a given set of hardware, which operating system would sound the best? Would there be a difference? I honestly don't know --- and nor do I trust my ears/brain to do that kind of A/B-ing with any accuracy, over reboots. Perhaps, sometime, someone will do some properly-set-up blind tests with identical hardware, and I will be very interested in the results. But I really, really, really don't want to go Microsoft again!

BTW: I use Aqualung as my player (available across platforms) because it plays music without doing too much else, is fully compatible with the Jack system, and handles gapless (eg live music split into tracks) playback perfectly.

Thanks Thad for the detailed info.

I tried one more tweak in the Win 7 system . By setting the process priority of the player ( foobar/cPLay) to 'real-time' , it looks like the problem with the occasional 'click' is resolved

I am still interested in experimenting with the linux dist. But not immediately. I think , It is a mammoth task to make everything to work fine in a Linux . I use the same Desktop for movies with the HW decoding in the MB , So making It to work for video and audio ( exteral USB DDAC ) would be challenging.
.

 

[Thad E Ginathom]

So long as you have space on a hdd, it is easy to set up a dual-boot system. One can also get a flavour using "live-desktop" dvds to boot from without installing.

Stay away from firewire and unsupported hardware and Linux does not need much work to play. It isn't a mammoth task. It might be just bias on my part (every possibility of that! ) but Linux, with its Unix grandfather going back to the 1960s, has a design that was doing multi-tasking when Windows couldn't even do two file copies at once. For audio and video, please see XBMC. There is a thread or two about it here, but because I don't do video, I didn't take much notice. Probably this is the OS for an AV system.

By the way, the PC/hardware combo that gave me DPC-latency nightmares wouldn't work properly with Linux either. Sometimes it is just the hardware

 

[vijwilso]

I spent some more time experimenting with these player ( primary objective was to test my new DAC with the Dedicated CD player )

Setup consists of
1) Exposure 2010 CD player -> Exposure Pre Power -> Focal Chorus 826V
2) (Laptop)FooBar/cPlay ->(USB 32) Audio GD 3.33 --> -> Exposure Pre Power -> Focal Chorus 826V

This time Abhishek from Audio dynamic was with me ( thanks to him for this ) , we ripped couple of tracks from the CDs using EAC and played the same track from Laptop and CD players at the same time . We used the input selector of the amp to change the source which seemlessly played the same track at the same time point .

My observation - Exposure CD player is better compared to the Audio GD in certain areas.

Second , with the limitation of the DAC in place, the sound from foobar is almost same as the CD player , which shows it is more neutral compared to cPlay ( assuming CD player is netral )

With the cPlay , the mid was more prominant , to me it was more enjoyable to listen compared to the foobar. True to the recording or not , we both felt like , focal is performing with the goodness of the ProAcs like Mids + strength of the Focal in Bass/highs, Who would say no to this . I will continue to use the cPlay.

 

[Anurag]

Have you tried foobar's ASIO or Kernel streaming output plugins? i find all of them have their pros and cons and not totally satisfied with jriver either.

Although i didnt try cPlay as it only supports Asio, but i tried couple of other players like HQplayer, XXHighend, StealthAudioplayer, Jplay, all of them are different and none is perfect, given there are so many variables to tweak with at the OS level itself.

The 4 primary tweeks on windows system that i find makes the most difference in sound are
-disabling the pagefile
-prioritising background process in advance system settings
-disabling the processor power management from Bios as well as OS settings
and enabling maximum processor state always
-disabling all the unneeded system devices from device manager, unneeded USB universal host controllers

apart from that
-Turning off HDD write cache buffer flushing
-disabling USB power management
-disabling unnecessary services and anitvirus


Then there is also a linux distribution, based on Mint OS
AudioPhile Linux | Quality audio on Linux

Compared to win 7 it sounds richer, cleaner and full bodied with bigger soundstage, may be a little laidback not sure

Absolutely minimal OS only audio related services, less variables, i feel its better..

 

[vijwilso]

Have you tried foobar's ASIO or Kernel streaming output plugins? i find all of them have their pros and cons and not totally satisfied with jriver either.

Although i didnt try cPlay as it only supports Asio, but i tried couple of other players like HQplayer, XXHighend, StealthAudioplayer, Jplay, all of them are different and none is perfect, given there are so many variables to tweak with at the OS level itself.

The 4 primary tweeks on windows system that i find makes the most difference in sound are
-disabling the pagefile
-prioritising background process in advance system settings
-disabling the processor power management from Bios as well as OS settings
and enabling maximum processor state always
-disabling all the unneeded system devices from device manager, unneeded USB universal host controllers

apart from that
-Turning off HDD write cache buffer flushing
-disabling USB power management
-disabling unnecessary services and anitvirus


Then there is also a linux distribution, based on Mint OS
AudioPhile Linux | Quality audio on Linux

Compared to win 7 it sounds richer, cleaner and full bodied with bigger soundstage, may be a little laidback not sure

Absolutely minimal OS only audio related services, less variables, i feel its better..

Yes I always use the foobar and cPlay with the Native ASIO driver provided by the DAC vendor.

Also i have done all the optimization as in your list and the DPC latency is quite good and does not vary .

I wanted to try the ap-linux , but their web-site states the latency is in the milliSeconds , shouldn't that be in Micro seconds for Audio ?

DPC latency in my Win7 is around 45 MicroSeconds , i assume the overall end-to-end latency( between PC ans DAC ) is also the same as the native ASIO directly sends it to DAC and according to DPC , it gets the CPU cylcle not greater than 45 microSeconds .

That was the reason I did not try the ap-linux.

Also i tried foobar with kernel streaming , to me ASIO with foobar is better compared to kernel streaming

 

[Thad E Ginathom]

Latency is confusing. Latency of the ordinary sort does not really belong in the same sentence as latency of the DPC sort.

Latency of the ordinary sort is simply the amount of time it takes data to get from A to B, just like the amount of time it takes a person to walk from their house to the end of their road. Distance is involved, it necessarily takes time, but what is important is that they reach there, not how long it takes.

Latency of the DPC sort is like being held up, having to wait for something else, half way along that road. That's not good.

DPC latency checker is a trouble-shooting tool: unless you just like watching hopefully-green bars, you don't really need it unless or until there is a problem. **It would be useful, for pre-evaluating a system for Audio

Latency of the ordinary audio sort doesn't matter --- and I disagree with the AP-Linux guys if they say it does. It does not matter for listening. It may matter very much for recording, where you may be having to listen to sound that is not only "walking to the end of the road," but "walking back again" as well: passing throuhg your interface's ADC, processing, and being played back through your interface's DAC.

Jack (the Jack Audio Connection Kit) is a low/no latency audio system, because it does not add latency itself, but it still cannot be helped that it takes time for a signal to get from A to B. Things like buffer size are configurable: the smaller the buffer, the less the latency, but traded off against lower stability. **Latency certainly matters if the system cannot fill/empty/fill the buffers in time.

Consider that those who use systems that buffer whole songs are adding whole seconds to latency!

 

[vijwilso]

The DPC latency is the indirect indication of how good the Audio Data could be tranferred . If hte DPC latency varies to a high levels , it means the Audio Streaming application may not get the cpu cycles at the right time , and results in varying latency which is jitter,

Pls correct if this wrong .

 

[Thad E Ginathom]

In my last post I said that latency is confusing. Lest anybody mistakes me, let me say that I am confused, and that every time I think or post about it, I am trying to understand and explain it to myself!

In that spirit, I off the following:

The DPC latency is the indirect indication of how good the Audio Data could be tranferred

Deferred processing is the CPU saying, Wait! I'll deal with you later, I have other things to do first. It's never good when that happens to audio or video. They are real-time, but a PC, physically (whatever the OS) is not, so there will always be some compromise somewhere.

If hte DPC latency varies to a high levels , it means the Audio Streaming application may not get the cpu cycles at the right time

And, in the worst case, it will simply stop until it can. If the wait is long enough, there will be drop outs, if it is less there may be crackles. This is not a qualitative thing, but much more polarised. If a CD player behaved like this we'd send it back!

and results in varying latency which is jitter

Not "jitter," which is a specific timing problem encountered elsewhere in the digital audio chain. But whatever we call it, the results can be nasty. Been there, Done that, Written the thread

By the way, and to demonstrate the limit of my understanding of these things, I am still (or again, it didn't happen for a while) getting occasional playback dropouts. I know it is associated with memory filling up, and I also know that Linux should be able to handle that correctly, but to get much further is beyond me. I need a Linux guru!

 

[vijwilso]

Not "jitter," which is a specific timing problem encountered elsewhere in the digital audio chain. But whatever we call it, the results can be nasty. Been there, Done that, Written the thread

Ok, I am more confused now .

All along this is my understanding on the Jitter. And the scenario is PC is connected to the External DAC over USB ( Not with the inbuilt AudioBoard)

1) Jitter is the variance in the latency because of which the receiving device does not get the data in the expected time. If there is a constant latency all the bits would be received at the same time lag .

The worst case jitter , we will hear the drop-out , the minor jitter results in poor audio qualiy .

Worst case Jitter : A chunck of information missed could result in the drop out ( like a complete range of 1 to 15 bits is missed to receive on time )

Non-worst case Jitter : A total of 15 bits missed in the receiveing side over the range of 5000 bits ( randomly distributed ) and non-contiguous . This case we will not hear a dropout , but loss in the over all quality where the DAC could not reconstruct a exact analog due to random bits missed to receive on time.

Both the cases are indirectly related to the DPC Latency . If there is a variance in the DPC latency , the sending application will not get the cpu cycles ALL the time . In those cases , latency of the data transmitted also varies ( cause sending application simple can not send on time ) If the variance in the DPC latency is worst , we hear the drop out , otherwise it is just a loss of few bits every now and then which does not result in drop-out , but may result in poor audio quality.

I dont know if there is a way to messure end to end latency ( like connecting two PC connected over USB and capture the audio stream on the receiving end with the time-stamp and plot the larency and Jitter . I am more interested in this as this results in the utlimate quality ,

When i say loss of bit , it does not imply the real loss , rather it is the data not received On-time

Sorry , to extend this , but i don't understad this completly

 

[Thad E Ginathom]

Neither do I , Killerbrain

I'm not saying there is never any relationship, but I think you need to compartmentalise those things.

DPC only happens if something is able to insist on processing time from the CPU. It isn't the variance that you should look at, it is whether it is ever bad. Your DPCLat window may be solid green for half an hour, but one red spike might indicate a point where DPC Latency was bad enough to cause crackling or drop out --- or it may pass by unheard. It is hardware interupts that are causing this, so it is a hardware (or what is enabled in software) combination issue.

Audio latency is always there, because signal processing cannot be instantaneous. You can't [?] measure it with a DAC, but if you have a soundcard there are ways to loop the output back to the input and calculate the time taken by the round trip. Never needed to do that. it doesn't matter to us. Really.

The word jitter probably has origins in telecommunications and signal processing going way back, but I feel it is a sad choice for audio (for us non-engineers, anyway) as it sounds onomatopoeic, like it should sound like j-j-j-jitter. Actually, if it is audible, I'm told that it can present in a number of ways, but is often like a muddiness in the sound.

Here's a long article on jitter. Of course, the net has many more.