Q how much does upsampling improve the sound (detail etc)?
and wouldn't this need to be done as close as possible to the data source?
-----------------------------------------------------------------------------------
The CS8420 based upsamplers all sound different and don't improve detail; I think the improvement may be due to the DAC having a brick wall filter at twice or four times the base frequency.
You need to invest in a serious upsampler; a used Assembalge D2D should not cost too much; but the dCS Purcell or dCS972 is something else - much much much better.
You also need a serious dac. I like CD upsampled to 176.4 much better than 192 ie 4x44.1 kHz.
and wouldn't this need to be done as close as possible to the data source?
-----------------------------------------------------------------------------------
The CS8420 based upsamplers all sound different and don't improve detail; I think the improvement may be due to the DAC having a brick wall filter at twice or four times the base frequency.
You need to invest in a serious upsampler; a used Assembalge D2D should not cost too much; but the dCS Purcell or dCS972 is something else - much much much better.
You also need a serious dac. I like CD upsampled to 176.4 much better than 192 ie 4x44.1 kHz.
Gentlemen, please refrain from remarks that refer to income, cashflow, taste etc. One simply can't argue about such things.
Just to point out - all upsamplers are digital filters. The only difference is if you are an engineer or a marketdroid. But there are filters and there are filters. A carefully implemented, many tap, FIR will beat any of the off the shelf single chip wonders that are commonly used to create oversampling filters. Oversampling and upsampling are very closely related although not quite identical issues. The main difference being that a properly designed oversampler can take into account the exact realisation of the DAC architecture in the filter design.
Heavens only knows why belt driven CD drives arrived. In a correctly designed system there is clearly no possibility of them changing the data timing. In a badly designed one, yes. One could theorise about lower servo power supply noise etc etc. But unless the only thing one changed in the system at all was the drive mechanism - and left power supply, optical assemply, servo mechanisms, servo drivers, digital control, all the same - well you actually don't know what made the difference.
Heavens only knows why belt driven CD drives arrived. In a correctly designed system there is clearly no possibility of them changing the data timing. In a badly designed one, yes. One could theorise about lower servo power supply noise etc etc. But unless the only thing one changed in the system at all was the drive mechanism - and left power supply, optical assemply, servo mechanisms, servo drivers, digital control, all the same - well you actually don't know what made the difference.
Just to throw in a potentially crazy idea, but this is a DIY forum after all. 
Has anyone considered using a CPU for the digital filter, or a general DSP? I haven't done any calculations, but supposedly a modern CPU would be fast enough to do the job, and it would offer a lot of flexibility to experiment with whatever type of digital filter you could imagine, by not being specifically tailored to implementing certain types of filters? I am sure this wouldn't be economically or practically reasonable for any, but possibly the most high-end commercial models, but for DIY, why not?
Another equally crazy idea would be to use upsampling to very high frequency and use a computer graphics card for the DAC, or maybe just use such a DAC. They're usually only eight bits, but that should be compensated by the upsampling. I do realize there could be problems to make a low-jitter clock at those frequencies, though. Maybe the graphics card could even be programmed to do the filtering? Using OpenGL to program your CDP, chew on that one.
Has anyone considered using a CPU for the digital filter, or a general DSP? I haven't done any calculations, but supposedly a modern CPU would be fast enough to do the job, and it would offer a lot of flexibility to experiment with whatever type of digital filter you could imagine, by not being specifically tailored to implementing certain types of filters? I am sure this wouldn't be economically or practically reasonable for any, but possibly the most high-end commercial models, but for DIY, why not?
Another equally crazy idea would be to use upsampling to very high frequency and use a computer graphics card for the DAC, or maybe just use such a DAC. They're usually only eight bits, but that should be compensated by the upsampling. I do realize there could be problems to make a low-jitter clock at those frequencies, though. Maybe the graphics card could even be programmed to do the filtering? Using OpenGL to program your CDP, chew on that one.
Christer
look at the meridian gear
ash_dac
seems to have wondered off
clock yes done
"What I want to know about is the varying impedance with frequency of the feedback opamp IV...."
opamp iV i think desiigners use them to balance for output load,
probably easier?
we need that answered from an EE
as for ground
my question was iec cord (3 pin) rather than 2 pin cord
allan
look at the meridian gear
ash_dac
seems to have wondered off
clock yes done
"What I want to know about is the varying impedance with frequency of the feedback opamp IV...."
opamp iV i think desiigners use them to balance for output load,
probably easier?
we need that answered from an EE
as for ground
my question was iec cord (3 pin) rather than 2 pin cord
allan
Absolutely, look at Wadia gear.
They have filters that don't ring in time, absolutely fantastic stuff.
But thread is 3 things to improve on CD players sound, and does not have to be limited to the very 1st posted opinion.
I hope someone can provide an order of importance.
I just replaced some 5532 in my CD player, goodbye soggy sound.
Tom
EE
tmblack said:
tom
which opamps did you goto?
hope you used sockets.
If opa's, carlosfm's info on decoupling is worth it.
have you removed the mute transistors yet?
as for order of importance.
start with the simple and cheapest thing first?
allan
Christers idea of processor controlled
I think this maybe this
http://www.meridian-audio.com/p_g61.htm
search for "jocko low noise oscillator " finds it everytime:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6194&highlight=
also here:
http://www.diyaudio.com/forums/showthread.php?postid=32772#post32772
and while you're at it, you can peek here, too ...
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6121&highlight=
mlloyd1
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6194&highlight=
also here:
http://www.diyaudio.com/forums/showthread.php?postid=32772#post32772
and while you're at it, you can peek here, too ...
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6121&highlight=
mlloyd1
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.