Is this how oversampling works?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
When the left/right clock signal is 44100 Hz, that would mean no oversampling is performed, i.e. even if the DAC is fed 24 bit data words (this is the case here), the 8 less significant bits will be zeros. Unless you increase the word rate, you will not be able to "smooth" the steps by inserting interpolated samples in between.

Still, in theory, you should get slightly better performance, since a 24bit chip is (or should be) more accurate and linear than a 16bit chip. In a 24 DAC with +/- 0.25 LSB accuracy, the "16th" bit would be 256 x more accurate than in a 16 bit DAC with the same specs (this is simplified).

But, in real-world, DVD players often sound worse than CD players , because sometimes less care is taken for the audio part and more HF noise is present.

There are many people here who prefer non-oversampling DACs, so no worries. Just listen and compare ;)
 
Graham, off subject a little I know, but I have had good sucess in improving two el-cheapo DVD players so far.
I replaced the doubtfull quality caps in the decoding/audio stages with the low esr Hitano caps available from WES - EXR.pdf .
Also changing out the caps on the SMPS pcb improved things further, as did grounding the cans of all oscillator crystals.
The result in both cases was a bloody good sounding CD player thay can play movies too !.

Eric.
 
Off subject? Nahhh. Topic is trying to make my DVD player honk. Ok. The switcher psu is on the extreme left side of the case and the pcb is on the right and they are separated by the width of the drive. There is a little bit of 50kHz (radiated) noise sneaking into the audio from the psu, so one of the things I am going to do is make a little shield along one side of the psu to isolate it frome everything else. The psu has proper X and Y caps and a common mode choke at least, but there is no earth to the case; I think it is suposed to rely on the RCA connector earth! I'll have to make a proper 3 wire mains lead for it. Whoa! Maybe I should get one of those enchanted power cords. ;) I'm just lucky at work we have a shielded room and nifty new HP RF spectrum analyser just for sussing out SMPS's. I may have to hit it with this big hammer. BTW, what's a good dual dual opamp I could stick in it? At the moment it has a JRC4580, IIRC. It is a little wider than a normal 8 pin SOIC.
 
OK. Here's the box in question.
 

Attachments

  • dvd player.jpg
    dvd player.jpg
    31 KB · Views: 137
Interesting thing is that the disc doesn't rotate at 500-200 rpm like a normal cd player, rather it goes at a constant 2500 rpm or so.

Anyway, this is the bitstream fed into the D/A with the remote volume at full blast. Only the 16 most significant bits (starting from the left) are fed in as you might expect, the lowest 8 are all zeros. The Upper waveform is the left/right clock.
 

Attachments

  • 16 bit.jpg
    16 bit.jpg
    8.2 KB · Views: 138
Then I lowered the volume by one step out of 22 and the D/A's get fed 24 bit all of a sudden! A scaling factor is applied to the 16 bit signal and the lowest 8 bits presumably represent the remainder, rather than sticking with 16 bits and losing resolution!
 

Attachments

  • 24 bit hi volume.jpg
    24 bit hi volume.jpg
    8.6 KB · Views: 130
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.