Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[kiwi]

Hello,

I am hoping to use MC for the highest quality playback possible.  Ideally, I'd like to get bit perfect output.  I'd like the sound card to be able to output this digital signal and it should be getting that signal to its DACs for analog output.

To do this, I am encoding all of my music in Monkey's Audio Format and will be using a pro grade sound card, so that there are no issues with 44.1KHz to 48KHz upsampling, and that the DACs in the card are of reasonably good quality.  

My goal is to use a computer + MC to be an effective replacement for the audio source in my stereo.

Here are some questions that I have about this process:

1. How do I get the cleanest signal path through MC?  i.e. If I don't want MC to really touch any part of the signal, does simply setting the volume to 100% and turning off all DSPs do the trick?

2. What should the Output settings be?  i.e. Wave Out, Direct Sound, or ASIO?  And what should their respective  settings be?

3. I noticed in the help file that it talks about the "Output Format".  This appears to be an option in the DSP Studio setting.  How should this be set?  Should it be turned on?  If it is not turned on, what are the default settings?  Bitdepth is set to source bit depth?  

4. Crossfade : If this is enabled, does it effect the signal quality when the songs are not being faded together?  I understand that to perform a crossfade, the data has to be combinded between the end of the old song and the beginning of the next one.  However, I don't want this to result in the signals being changed once the fade is complete.  Is it best to turn this off entirely?


Thanks, I'd love to hear what other people are doing for high quality playback and hear about any other issues that I've missed (as I'm sure that there are many.)

-kiwi

[jam]

I'm using MC for the same purpose.  So, I just want to try to answer questions and explain them.  I may be wrong, tho.

1A.  MC's volume is just a driver's master volume.  Therefore, how it works is depend on the sound/audio card driver.  e.g. Digi96/8 has no master volume, so that MC has no volume too.
   I'm not sure about DSP part.  I think if I turned off DSP, MC plays the original sound.  However, I'm not 100% sure.

2A. It's also depend on driver.  Some card has problem with DirectSound, so comparing among them is good idea to find the best.  ASIO should be the best, but it might be too heavy to acheive its low latency.  So, again, try them out is the best idea.  I'm using WaveOut.

3A. If you want to the original wave data, don't touch it.  If you want to convert data to better format (e.g. wanna do oversampling), try it out.  I think it is possible to use DSP to get better sound quality.  However, it's depend on DSP component.  Moreover, how better is depend on how you feel.  So, again, the best idea is to try them out.  Personally, I'm not using any of DSP for now.

4A. I don't hear any differences between Crossfade on and off at the middle of song.  So, I'm using it.  Cross-fade (aggressive) with gapless is the best crossfade I've ever heard.  This crossfade is my primary reason of using MJ/MC.  

[kiwi]

Hi Jam,

Thanks for responding.

In general, how do you find the output quality of the system using MC?  Have you compared it to regular CD players?

1A.  MC's volume is just a driver's master volume.  Therefore, how it works is depend on the sound/audio card driver.  e.g. Digi96/8 has no master volume, so that MC has no volume too.
   I'm not sure about DSP part.  I think if I turned off DSP, MC plays the original sound.  However, I'm not 100% sure.

Are you using the REM Digi96/8?  

Cool.  That makes sense.  I didn't know how it dealt with the volume.

Is that because MC has this setting:
Optimize volume: By default, Media Center is set to optimize the volume on startup. This maximizes the "Wave" volume slider and turns the master slider (the one Media Center runs) down by the corresponding amount.  This gives the best S/N ratios possible at a given volume for most cards.  (From their Help)

2A. It's also depend on driver.  Some card has problem with DirectSound, so comparing among them is good idea to find the best.  ASIO should be the best, but it might be too heavy to acheive its low latency.  So, again, try them out is the best idea.  I'm using WaveOut.

So, if the card had all three output methods, ASIO would be the cleanest because it is the most direct path to the card, however, it can use up more system resources?  Then comes WaveOut, where MC takes whatever input it has and generates a wave output which it sends to the sound card.  And finally, direct sound.  Where the output is left up to the mercy of the driver and windows internals?  

3A. If you want to the original wave data, don't touch it.  If you want to convert data to better format (e.g. wanna do oversampling), try it out.  I think it is possible to use DSP to get better sound quality.  However, it's depend on DSP component.  Moreover, how better is depend on how you feel.  So, again, the best idea is to try them out.  Personally, I'm not using any of DSP for now.

I guess that I am of the opinion that most of the time, dsp settings, generally degrade the quality of a good recording.  If there were more complicated Room Correction software, then I'd be more interested in them.  However, just simply playing around with a ten band graphic equalizer seldom seems to generate better sound.

It would be great if there was a DSP that was written that could do Room Correction like this device:
http://www.tactaudio.com/RCS22X/index.html

But I haven't seen any so far (I hope that I've miss it, and that there is one available.)

4A. I don't hear any differences between Crossfade on and off at the middle of song.  So, I'm using it.  Cross-fade (aggressive) with gapless is the best crossfade I've ever heard.  This crossfade is my primary reason of using MJ/MC.  

Yeah, I really like the cross-fade from MC, particularly when just casually listening. However, I'd give it up in a second for serious listening if it resulted in sound degradation during non-fade periods.

Thanks
kiwi

[Flobalob]

I changed to MC with the same idea of replacing all my hi-fi sources with everything on hard drive.
We talked about setup a few days ago...

http://www.musicex.com/cgi-bin/yabb/YaBB.cgi?board=MediaCenter;action=display;num=1053706706

I had played around with Escient systems etc. but found them all very half baked.
I now have all my music ripped to .ape format, MC sends it out of my computer via S/P-DIF to a Tact RCS which you talked about then to Meridian speakers.
It is more than I had hoped for.
I have AB'ed it with the best digital sources available and it is impossible to tell the difference.
I keep all DSP disabled, my sound card does not even show a volume control in MC (M-Audio Audiophile 2496), I use the Direct Sound output but I don't use the sound cards DAC's, I exit via coax to an external DAC. I think the DAC quality is going to be the biggest deciding factor in your final sound quality.
I don't think DSP per se is to be shunned, it is all in the implimentation. Meridians whole line of high end equipment is all based on Digital signal processing. I prefer though to let dedicated hardware that was designed for that purpose to take care of this aspect of the signal.
Finally, the crossfade is merely accessing two tunes simultaneously. There should be no detriment during no-fade times.

[jam]

Flobalob,

Thank you for refering the interesting thread.  

Kiwi,

In general, I'm comparing them by ears.  I don't have regular CD player in my system, so no AB-test.  I don't want to buy any of them since I think my system sounds better than regular system that I had old days.

Are you using the REM Digi96/8?
ASIO would be the cleanest because it is the most direct path to the card, however, it can use up more system resources?

Yes.  I'm using Digi96/8 and love it.  Yes again.  ASIO should be the cleanest since it is designed and developed for that purpose.  However, it consumes CPU time because ASIO is trying to make latency as low as possible and play music real-time.  Some system may show glitches with ASIO.  You may need to adjust buffer size and other stuff to optimize system.  DirectSound should be the next since it is newer than WaveOut and should be better.  However, some pro grade audio cards like Digi96/8 don't support DirectSound well.  So, I said try them all to hear what is the best for you.  

[kiwi]

Thanks Flobalob and jam.

Yeah, my questions were more about MC and what to do/not to do when trying to get high quality output.  

I agree that DSPs can definitely be used very effectively in improving audio sound.  Adding environmental effects, virtual subs and widening the sound space do not fall into that category.  Sure they might sound "cool."  But they are not an accurate reproduction of the original music.  And usually they aren't done particularly well.  

Maybe in the future, you could record a chamber group in an anechoic chamber and then effectively place them into whatever listening hall you wanted to listen to them in... and have it all corrected for your room.  i.e. allow you to listen in your living room to a performance that is very close to a live performance in any other location.  

Thanks for the information on your systems.  That is very helpful in deciding which sound cards to get.  At the moment, I'm just starting that process and have found quite a few sound cards that all appear to have the qualities that I'm looking for.

Thanks,
kiwi

[Flobalob]

Me Again...
If you leave all DSP disabled including output format then you should end up with an unchanged (from the original CD) PCM stream into the sound card regardless of the output settings. The output settings should only affect the efficiency by which the CPU gets it to the output but obviously the simpler the route the better.
Only two things can happen to the data stream assuming it is intact (all the 1's and 0's in the correct order). one it time slurring and the other is jitter (which in itself is time related).
My intuition and experience tells me that...
a) I would not convert to analog withing the computer due to noise issues and DAC quality issues.and...
b) Some reclocking / dejittering device needs to be used outside the computer. Monarchy Audio make a Stereophile Class A rated device that does this and is affordable

Also leaving the computer with coaxial digital can cause ground loops and hence noise whereas an optical digital out effectively decouples the computer from you Hi-Fi equipment. Optical is more suseptible to jitter but this can be corrected.

I have owned or tried out a bunch of external DAC's and the sound difference is huge.
So if you have not yet purchased a sound card. I would get the simplest card that has a S-P/DIF output and spend the savings on outboard gear to interface with your stereo system.

[knickelfarz]

I always was quite sure that the best audio connection between PC and HiFi would be optical SPDIF. Since a lot of cheap soundcards and even motherboards offer SPDIF output and most HiFi Amps offer SPDIF input (since surround sound is getting popular...) this is also quit affordable - I would spend the money for better speakers...

On the other hand, those "cheap" soundcards and motherboards mostly have a fixed 48kHz SPDIF output, maybe resulting in the mentioned "issues with 44.1KHz to 48KHz upsampling" ... perhaps i was lucky, but I never had those issues - the sound was better and cleaner than analog output. I made this comparison with
- Terratec DMX 6FIRE, which has very good DACs
- Creative Audigy, where analog output is a little bit noisy...
- Onboard coaxial SPDIF output of a VIA EPIA-M ITX motherboard (converted to optical via a cheap adapter), the analog outputs of this board are only average...

The digital output of all three devices seem to sound equal with my Onkyo Amp (I guess It's DAC is quite good, but I have no comparison to another amp with digital input).
The analog sound of the three devices had the mentioned restrictions, the digital variant always sounded better...

I would like to hear, if anybody made comparable or maybe contrary experiences, especially with "cheap" SPDIF outputs, and especially concerning the 44.1KHz to 48KHz upsampling  

[DocLotus]

If it's quality you are after then by all means check out...

PC Power & Cooling... http://pcpowercooling.com/

They have been in business for a very long time.  They make very quite power supplies, whisper fans & a product you may be interested in... the Sleekline System PC's (up to 2.8 GH) with a choice of the ATI Radeon 9700 Pro of ATI All-In-Wonder 9700 graphics boards.

The Sleekline PC's are avaliable in home or office versions.  One might be right for you.

[knickelfarz]

Hey Doc, I checked the power supplies and I have to say: It's the final solution for HiFi popes - A never known before clean ultimate 50Hz signal blasted out my speakers!

Now listening to: 'Frequency Unknown' from 'Crackle Noise' on MB 9.0 (MyBarbecue No. 9) (For this year, yummy)

[kiwi]

Flobalob & knickelfarz :  Toslink Connection.

Hmm... I hadn't thought about that.  Find the least expensive card with a toslink, that doesn't resample music... preferably 24/96 and then get a de-jitter device and a good external DAC.  I'm amazed that external DACs don't automatically reclock the incoming data and output it on their own clock signal.  Hell, if I were designing a DAC (and I'm actually working on the design of a super high fidelity A/D system right now) I wouldn't for a moment think about using some external, unknown clock signal as the reference clock for my system.   It doesn't seem that jitter is ever enough to cause bits to be incorrect.  I'd just figure out what clock was supposed to be coming in... i.e. 32KHz, 44.1KHz, 48KHz, 88.2KHz, 96KHz... pump the data into a FIFO (which wouldn't have to be particularly big) and then output the data to a DAC with my own clock.  Now maybe there is something more to it that I am missing, but (obviously) I can't think of it.

Hey Doc, thanks for the link about the Slimline PCs.  I think that if I were going to spend that much money on a system, then I'd just get one of the SID FutureClients:

If you got the money for an ultimate silent solution you may take a look at the "SID FutureClient" from SignumData:
http://www.signumdata.de/english/index_eng.htm
It's build similar to the Hush ITX but with a P4 inside...

(Check out that thread for a discussion on case and silent systems)



The FutureClient looks extremely sexy and is completely silent, absolutely NO fans! I'm just thinking that it may have to be the next machine that I get...  after I've proven to myself that this system can replace a good source.  (And if the Hush doesn't do everything that it needs to.  Though, I think it may.)


-kiwi

[xen-uno]

Regarding 2)...

In order of preference: ASIO, DirectSound, WaveOut

Why? Read this (hard to read in browser...use NotePad)

It's a toss up between ASIO & DS. ASIO was created mainly to get around the high latency and limited channels of WO.

R_x

[kiwi]

xen-uno: thanks for the info.  I'm just reading through that doc now.  It's very helpful.

kiwi