Print

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

[SteveR]

Hello all, new member alert..
I have just started using MC15, having decided to sell my Cyrus CD6SE and go completely PC based. I have a Sony Viao F12 laptop which has the inbuilt optical out which I use to connect to the dacmagic. I have managed to sort out my library, and am in the process of ripping CDs to the PC hard drive temporarily using FLAC level 6. I have WASAPI Event style selected, as I have read this is the optimal setting, it works OK for me. SQ is excellent via my Kandy K2 amp and Spendord S6e speakers.

I just wanted to understand the various output setting and how best to configure things. I have the driver in windows (7 64 bit) set to exclusive control mode, though the shared out put rate seem to make no difference (is set to default DVD 48khz 2 channnel DVD quality). I have set the bit depth to 24 same as the dac, is there anything else I can do tweak the sound..? buffering (I have left at 50ms default).
Any commenst / help appreciated..

[JimH]

Welcome to the forum.  This thread may help:
http://wiki.jriver.com/index.php/Audio_Output_Modes

[SteveR]

Thanks - read that earlier. MC works with either WASAPI  and WASAPI Event Style. I have set it to the latter, on the basis it works (an impressive sond is coming out of the speakers anway....)

Another Q sorry, but I presume that using sbtle EQ adjustments are a little like alteraing mage setting in RAW files, as opposed to converted JPG ("analogue") where any adustments to the signal path or data affect other parts of the signal, the photo analogy being artifacts and general noise...

[mdx1]

Another Q sorry, but I presume that using sbtle EQ adjustments are a little like alteraing mage setting in RAW files, as opposed to converted JPG ("analogue") where any adustments to the signal path or data affect other parts of the signal, the photo analogy being artifacts and general noise...

Any changes, EQ, level or gain in the digital stage (computer) alter the data sent to the DAC.  That would make the data not bit-perfect.  The change may or may not cause audible SQ degradations but the bit-perfectness will be gone.

[SteveR]

Any changes, EQ, level or gain in the digital stage (computer) alter the data sent to the DAC.  That would make the data not bit-perfect.  The change may or may cause audible SQ degradations but the bit-perfectness will be gone.

OK that's interesting. What does bit-perfect mean exactly?

[mdx1]

OK that's interesting. What does bit-perfect mean exactly?

It means the original bits representing music are not altered throughout the stages from the music file (wav, flac, aiff, etc.) read off the disc or disk all way to the DAC.

[Matt]

To the original question, it sounds like you have Media Center configured well.

As for the bit-perfect question, it's true that any changes to the sound like software volume, equalization, etc. mean the output will no longer be "bit perfect" with the input.  All this really says is "changing the sound means the sound will be changed."

Many changes to the sound are desirable, which is why we offer them.  This includes sound changes like equalization, room correction, and software volume.  All processing in Media Center uses audiophile-grade 64-bit DSPs.

[SteveR]

To the original question, it sounds like you have Media Center configured well.

As for the bit-perfect question, it's true that any changes to the sound like software volume, equalization, etc. mean the output will no longer be "bit perfect" with the input.  All this really says is "changing the sound means the sound will be changed."

Many changes to the sound are desirable, which is why we offer them.  This includes sound changes like equalization, room correction, and software volume.  All processing in Media Center uses audiophile-grade 64-bit DSPs.

The analogy as I understand is digital photo editing. A DSLR takes .RAW image files, which can be altered with the correct programmes such as photoshop etc. These have a negligible deterioration of the final processed image, other than the deliberate changes you make. However, if you carry out the same processing on the converted file (e.g. a JPG file) then many unwanted effects are introduced for the same processing changes, i.e. noisy images with artifacts etc. In other words, I would imagine that adding a 16KHz mild lift won't reduce the amount of detail in the midband, whereas using passive tone controls in the analogue you most likely will.

[SteveR]

Hmmm forgto to ask the rather important question about the DSP - I have read that its is important to set up the bit depth to the same as the DAC - in my case it would be 24 bit. Is this the case? I also have the output driver set to default (i.e. both in built speakers or optical driver whichever I have switched on)

[Matt]

Hmmm forgto to ask the rather important question about the DSP - I have read that its is important to set up the bit depth to the same as the DAC - in my case it would be 24 bit. Is this the case? I also have the output driver set to default (i.e. both in built speakers or optical driver whichever I have switched on)

The simple rule is to use the highest bit depth supported by your hardware.  This is 24-bit for most DACs.

Inside Media Center, the data flow looks like:
Input File > Convert to 64-bit floating point > Any DSP, VST, or other changes > Final downmix to output bitdepth (24-bit in your case)

[Matt]

The analogy as I understand is digital photo editing. A DSLR takes .RAW image files, which can be altered with the correct programmes such as photoshop etc. These have a negligible deterioration of the final processed image, other than the deliberate changes you make. However, if you carry out the same processing on the converted file (e.g. a JPG file) then many unwanted effects are introduced for the same processing changes, i.e. noisy images with artifacts etc. In other words, I would imagine that adding a 16KHz mild lift won't reduce the amount of detail in the midband, whereas using passive tone controls in the analogue you most likely will.

Since all processing happens on decompressed audio in 64-bit space, there's essentially no added noise or artifacts from any processing.

In my opinion, it's good to do as much audio processing into Media Center as possible.  This handling is one of Media Center's core strengths, and there's no way an analog circuit could compete with it.

It's a bit of a tangent, but one of our pushes with v15 was to incorporate the processing needed for Media Center to replace the traditional work done by a preamplifier (things like bass management or room correction).  There are still some hardware issues to this approach, but we have made a lot of progress on the software side.

[SteveR]

The simple rule is to use the highest bit depth supported by your hardware.  This is 24-bit for most DACs.

Inside Media Center, the data flow looks like:
Input File > Convert to 64-bit floating point > Any DSP, VST, or other changes > Final downmix to output bitdepth (24-bit in your case)

OK, now You've lost me.

I've got Output Format checked in DSP, and bitdepth set to 24bit, no resampling, everything else unchecked (no subwoofer)

[mdx1]

The analogy as I understand is digital photo editing. A DSLR takes .RAW image files, which can be altered with the correct programmes such as photoshop etc. These have a negligible deterioration of the final processed image, other than the deliberate changes you make. However, if you carry out the same processing on the converted file (e.g. a JPG file) then many unwanted effects are introduced for the same processing changes, i.e. noisy images with artifacts etc. In other words, I would imagine that adding a 16KHz mild lift won't reduce the amount of detail in the midband, whereas using passive tone controls in the analogue you most likely will.

I understand what you are saying and should have answered in my earlier reply.  I have been shooting 100% RAW with my Canon DSLRs since 2003 using Capture One Pro.  Yes, the raw vs. cooked analogy is very clear to me.  Resaving a JPG file will lose quality as JPEG is a lossy format.

However, it is a bit different in the audio world.  The digital audio files ripped from the CD or downloaded (bought) from HDTracks or alike are all "cooked".  The equivalent "RAW" file would be the original recording project file internal to many recording devices.  Once that is converted to wav, flac or aiff, it is no longer "raw".

But with that said, the jpg equivalent in audio would be mp3 or other lossy formats.  The flac is like TIFF.  The wave and AIFF are more like BMP.  Changing TIFF and resaving it should not hurt the image quality but the image has been altered.  The bits have changed.

Similarly changing a flac file and resaving it should not reduce the quality.  However, the bits have changed.  The quality in the audio world has another criteria, the bit-perfectness.  Ideally, when one listens to a recording, the original unaltered bits should be used to convert to the analog format.  When EQ or other DSP is applied to the signal, the original sound is altered.   One may still prefer the altered sound to the original if the mastering engineer had a different take on the sound.

[Matt]

OK, now You've lost me.

I've got Output Format checked in DSP, and bitdepth set to 24bit, no resampling, everything else unchecked (no subwoofer)

That's good.

It means there will be a final downmix from 64-bit to 24-bit (the highest your DAC supports).

[mdx1]

Hmmm forgto to ask the rather important question about the DSP - I have read that its is important to set up the bit depth to the same as the DAC - in my case it would be 24 bit. Is this the case? I also have the output driver set to default (i.e. both in built speakers or optical driver whichever I have switched on)

I use the following under the DSP Studio Output Format panel.

Sample rate
   Resample:  No resampling

Bitdepth
  Bitdepth: Source bitdepth

Channel
  Channels: Source number of channels

To make sure it is using the correct output device, you should select the optical driver explicitly.

[mdx1]

Since all processing happens on decompressed audio in 64-bit space, there's essentially no added noise or artifacts from any processing.

In my opinion, it's good to do as much audio processing into Media Center as possible.  This handling is one of Media Center's core strengths, and there's no way an analog circuit could compete with it.

It's a bit of a tangent, but one of our pushes with v15 was to incorporate the processing needed for Media Center to replace the traditional work done by a preamplifier (things like bass management or room correction).  There are still some hardware issues to this approach, but we have made a lot of progress on the software side.

Matt,

I am glad you shared the information about the DSP used in MC here.  Is it safe then to use ReplayGain for album level volume leveling in MC? 

Has Chris or anyone on the ComputerAudiophile.com reviewed this feature?

[SteveR]

Matt, thanks for the explanation on the photo analogy. I understand it much better now.

There is an ongoing debate about the validity of equalization in most of the hi fi forums, part of me is still in flat earth thinking and that bit perfection is what we should be striving for. However, the pragmatic view is that most modern music is processed in the studio and all manner of EQ and processing applied prior to cutting (burning) the master, so some form of room correction is probably required, especially as the speaker designer will have done the development in a room that is likely very different from yours or mine...

I've yet to play with the room correction part - looks fascinating.

[Matt]

I use the following under the DSP Studio Output Format panel.

Sample rate
   Resample:  No resampling

Bitdepth
  Bitdepth: Source bitdepth

Channel
  Channels: Source number of channels

To make sure it is using the correct output device, you should select the optical driver explicitly.

Please see my reply above about bitdepth:
http://yabb.jriver.com/interact/index.php?topic=60520.msg408560#msg408560

It's best to use the highest your card supports.  If you use ASIO, this happens automatically (so the setting is ignored).

[Matt]

There is an ongoing debate about the validity of equalization in most of the hi fi forums, part of me is still in flat earth thinking and that bit perfection is what we should be striving for. However, the pragmatic view is that most modern music is processed in the studio and all manner of EQ and processing applied prior to cutting (burning) the master, so some form of room correction is probably required, especially as the speaker designer will have done the development in a room that is likely very different from yours or mine...

I don't really have a horse in this race.  I say follow your ears.

Media Center can deliver bit-perfect results, and does so by default.

It also has, in my opinion, the best DSP engine available.

[Matt]

I am glad you shared the information about the DSP used in MC here.  Is it safe then to use ReplayGain for album level volume leveling in MC?

Yes, absolutely.

Like I said above, just be sure to output at the highest bitdepth your hardware supports.

Volume is simple (it's a constant scalar of the amplitude of the sound wave), and Media Center implements it in 64-bit floating point which is the highest precision method practically usable on a computer.  64-bit precision delivers billions of times more precision that your ear can use.

I suppose you could imagine cases where turning down the volume in software could force you to increase the volume of some other external hardware.  In this case, even though the software volume didn't degrade the signal, the final output quality could be worse if the other hardware didn't do a good job when turned up.  I don't think this would be a real-world issue in many configurations.

[SteveR]

I use the following under the DSP Studio Output Format panel.


To make sure it is using the correct output device, you should select the optical driver explicitly.

I can only select either the optical out or the internal speaker/headphone card by changing the default in windows. Plugging the toslink into the shared optical / headphone socket does not mute the speakers when set to internal speaker mode i.e. internal DAC active. Pluggig headphones in of course does mute the speakers and send the signal to the 'phones.
I assume that the changing the windows default in the control panel is simply switching the internal DAC On/Off and correspondingly the optical out driver Off/On. Makes no difference to the SQ in my hi fi set up whether set to default or optical only. I haven't yet tried using the internal DAC into headphones in WASAPI mode.

Sorry more Qs and not directly related - 1/3 0f my library is in wma lossless, the rest in flac. Is it worth me converting the wma files to FLAC? I also want to stop WMP importing and converting files back into its own library again, but thats my problem to fix....

Finally, I have done some headphone listening this evening (Grado SR80) and am simply stunned by the playback quality. Most of my CDs sound pathetically produced (as most are these days), the soundstage on studio albums sounds very narrow on all but a few where they have been processed less in the mixdown, but the overall level of detail is stunning. One track I always check with is Mamma Mia by Abba as a test of how to follow a bass line from both harmony, and rhythm, as its this instrument that drives 50% of the melody. Listening to it through MC and its so easy to track the melody its playing, and as importantly the timing. It's one of the best bits of bass playing in any song IMO in any case because of how much it adds for so little effort.

[JimH]

Steve,
It's fun to follow what you're doing and how you find the experience.  Thanks for the kind words.

Any lossless format will produce the same sound as any other.

[SteveR]

Steve,
It's fun to follow what you're doing and how you find the experience.  Thanks for the kind words.

Any lossless format will produce the same sound as any other.

You're welcome, and thanks for the reply. Great forum by the way, one of the most responsive I have used in any field (and I post on quite a few).

I also think the ripper is good - several times it has re-read and corrected ter data off disc, so I can see the verification process is at least going trough te check phase - ripping in wmp didnt do this once in about 30 discs. Even brand new hardly played discs can cause problems, I had to reload such a disc 3 times before it would rip. All very re-assuring! All I need now is for the £GBP / $USD to swing in my favour in the next 25 days....!

POST EDIT: am getting some glicthes when opening other programmes, and browing the web using Chrome. I get the occasional drop out for a moment on approx 1 in 20 tracks or so. Any ideas? The first is an easy cure I guess, but it's somewhat restrictive, and the second I have none...

[Matt]

POST EDIT: am getting some glicthes when opening other programmes, and browing the web using Chrome. I get the occasional drop out for a moment on approx 1 in 20 tracks or so. Any ideas? The first is an easy cure I guess, but it's somewhat restrictive, and the second I have none...

Increase the buffer setting in Options > Audio > Output mode settings...

I believe you said you were using "WASAPI - Event Style".  If this is the case, try 250ms for your buffering setting.

[SteveR]

OK, that's fixed it. The effect on cross fade etc isn't that noticeable really.

I have also discovered the links to google images - I now have all missing album artwork, and corrected the ones that was wrong from the original download. Touches like this make it really pleasurable to use, great attention to detail.