• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

The sound of chassis?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
analog_sa said:


There are no such requirements for red book cd. Not enough error correction code to make up for all reading errors. And much more importantly there is jitter.
This topic has been discussed hundreds of times. Why don't you do some reading?

You know, there are CDROMs which can reliably do perfect DAE. Plextor stands as a front runner, although a lot of drives these days do just as good a job. Even for drives which don't do such an amazing job, there's tons of room for jitter correction at playback speeds.

I don't understand why instead of putting tons of effort into asymptotically approaching perfect extraction via case dampening, etc. people don't make up an interface with one of these drives and be done with the whole thing.
 
By making an interface i assume you mean something other than spdif. It's the silly interface with embedded clock that's probably the cause of 99% of the problems. I've tried some of the software based 'perfect' extraction proggies. If your CDROM is not capable of perfect extraction the softaware approach doesn't seem to work so well.

cheers

peter
 
Excuse Me Sir, But Your Ignorance Is Showing...

"I don't understand how the air platform makes any difference with a CD player."

From a previous post -
I'll Say It Again.......
"Another scenerio where damping could be seen to benefit is: without damping we get more vibration on the disk making the laser mechanism work harder to track the disk. This causes more noise and load in the power-supply degrading the analog section of the player."

Yup, thats what I said earlier on .
Additionally, digital power supply fluctuation causes modulation of the analogue waveform datastream slicer (jitter), modulation of the clock oscillator (jitter) and modulation of timings in the DSP chip.
Seperated power supplies for servo, digital and analogue audio stages pays big benefits.
Mechanical damping helps, but the electronics require attention also.

Eric.

There are portable cd players that have a buffer memory such that the audio being decoded is delayed by two or three seconds after being read from disc.
This gives enough time for disc sections to be re-read if corrupted, and the disc spin speed is servoed to keep the buffer full (or near to full).
This buffer memory effectively isolates the focus and tracking servo induced power supply pertubations from the D/A and audio output stages, and allows substantially less jitter.
When servicing these players on the bench, when switched to normal operation, the audio output is similar to any other domestic cd player, but when switched to anti-shock buffer mode, the audio output is remarkably cleaner, sweeter and smoother due to substantial jitter reduction.
The signal coming off the disc is not a clean data stream, and the differences are in how the following stages behave, and much is to do with power supply cleanliness and earthing techniques.
Want some more ?.
Just do some searching or reading.

Eric.
 
Analog
sounds like the biggest problem with red book cd format is that it is crap. Thanks for clearing that up for me. Since you seem to be up to date with the literature on the subject, do you know about how often/sec are we listening to interpolated music as opposed to what's recorded on the CD? I am sure the SONY and Philips engineers have tested and characterized the influence of vibrations on the reading mechanism before lauching the red book format, not that discovering how bad it was would have stopped them. Can you suggest relevant literature that is more technical and less anecdotal to better help me understand?
I have a genuine interest in understaning this subject and I apologize if I came off the wrong way. Thanks for slapping me back into line! ;)

MrFeedback, my ignorance is deep indeed. I always thought I was listening to the corrected, checked and re-checked buffer signal. I wonder why it isn't so by default.
 
Grataku,

I understand your dilemma. However, the thing about vibrations in the reading of digital material is not in the mixing of signals as you mention with subsonic plus audible analog, but in the error correction the CD player now needs to undergo. I do a bad thing. I place my CD player right on top of one of my speakers and occasionally it will stop playing and resume due to vibrations of bass. But this is rare and only in my lab.

Basically, the error correction does an approximation of what the stored data is supposed to be rather than what it originally is. So the sound is different, or worse.

BTW, it is good to read that someone else also is aware of the presence of subsonic sounds and how it mixes with sound sources.

Cool!:cool:

What I would wonder is how, even though the table dampens vibrations travelling across the floor, the CD player doesn't vibrate from the sheer transmission of sound through the air, since he plays sometimes at full volume?:confused:

Gabe
 
Gabevee,
my basic dilemma comes from not knowing anything about this, really. I am not factoring the analog part, the servo part, into the equation at all which I thought was irrelevant to the software, digital part. The error correction by interpolation which, I think, is simply taking the preceeding and following point and averaging them to obtain the missing point in the middle, implies that there is a certain amount of "math" going on while the CD is reading this "math'" takes time to do. As a result, I thought that the sound was always buffered, that the interpolation was really the last resort to be used when repeat reading doesn't work, and that this doesn't happen on a continuous basis to completely change the overall character of the sound .
 
Up To Date ? - Old News Really.....

"Since you seem to be up to date with the literature on the subject, do you know about how often/sec are we listening to interpolated music as opposed to what's recorded on the CD?"
Gabe, most mass procuced CDP DSP processor chips have error flag output pins, but these are usually left unused.
Some high end machines have front panel indication of different categories of data errors, but this is rare.

"MrFeedback, my ignorance is deep indeed. I always thought I was listening to the corrected, checked and re-checked buffer signal. I wonder why it isn't so by default."
Just a little bitch slapping you back to reality - nothing personal. ;)
All cdps have a buffer memory, but on anti-shock machines the delay is much, much longer and this seems to make a big sonic difference.
These machines run an external (to the DSP chip) memory ic, and this is a cost function, of course.

"What I would wonder is how, even though the table dampens vibrations travelling across the floor, the CD player doesn't vibrate from the sheer transmission of sound through the air, since he plays sometimes at full volume? "
Because when the cdp is isolated, the only acoustic energy recieved is into the cdp cabinet itself, rather than being coupled from an excited shelf, which represents additional acoustical coupled energy, and further elimination of resonances between the shelf and cdp cabinet.
 
Re: Interpolation Is Audible....

mrfeedback said:
" The error correction by interpolation which, I think, is simply taking the preceeding and following point and averaging them to obtain the missing point in the middle"
This causes a temporary halving in bandwidth. ...... Hmmmm I hear you say.

Eric.

I enjoy a little 'bitch slapping' from time to time, keeps me awake. Now you have your chance to spank me, too.

Hmmmm I hear you say

No. Initially I'd say NO WAY, then I'd ask you what the hell is "temporary halving in bandwidth" then I'd say NO WAY again.
Please explain IN DETAIL what do you mean... Reciting the Nyquist Theorem won't do it. Learning is exciting.
 
Let's Go To the Stoning.....

Ummm, I forget references and details (Panasonic training manual maybe ?) but whereever it was stated interpolation causes a halving in available bandwidth for the sample period concerned.
To think about it, linear interpolation is the intermediate average of the two outside sample points, and by definition this will cause halving of possible variation betwen the two valid samples (the before and after points) - halving of BW IOW.

Eric.
 
mrfeedback,

"Since you seem to be up to date with the literature on the subject, do you know about how often/sec are we listening to interpolated music as opposed to what's recorded on the CD?"

Gabe, most mass procuced CDP DSP processor chips have error flag output pins, but these are usually left unused.
Some high end machines have front panel indication of different categories of data errors, but this is rare.

I didn't say this.


Gabe
 
You Need To Some Study Boy....

grataku said:
MF
maybe you are paraphrasing incorrectly, the whole point of interpolating is so that you don't lose the point by making an educated guess of it. Anyway nobody answer my original question. How often does this interpolation actually happen?
You need to read up on CDP DSP theory and techniques.
Anyway, all cdps have a buffer memory, and partly for this purpose.
When a particular sample is determined to be uncorrectable, the dsp will linealy interpolate a value between the previous and post sample numerical values.
This causes an automatic halving in bandwidth resoloution.

Eric.
 
This causes an automatic halving in bandwidth resoloution.
Bandwidth and resolution are 2 different things.
If you interpolate over only 2 bits, then the resolution is halved for that moment in time.
However, the system bandwidth remains (almost) as great as before. IMO The reason for this is that a sudden change in bandwidth takes time. Time that is not available for such a rate of change. If interpolatin takes place on every other bit, then the system bandwidth is halved.
 
analog_sa said:
By making an interface i assume you mean something other than spdif. It's the silly interface with embedded clock that's probably the cause of 99% of the problems. I've tried some of the software based 'perfect' extraction proggies. If your CDROM is not capable of perfect extraction the softaware approach doesn't seem to work so well.

Yep, I mean pulling out via SCSI or ATAPI.
 

Attachments

  • artisa1.gif
    artisa1.gif
    68.4 KB · Views: 296
Hiya. So awhile back you guys were talking about case materials and shielding and whatnot.. I have a question relating to those topics.

I am making a case for a gainclone, and the RF noise is kinda bad where I live.. So I was thinking of using thin copper sheets as the interior of my case.. Good shielding and good acoustic properties, right? And maybe the exterior would be aluminum... Though there seems to be some anti-aluminum sentiment here--not the most acoustic metal... But if the case had an aluminum interior, would that make up for it? What about a wooden box with a copper inner lining? Would using wood provide enough of an improvement over the aluminum to be worth the extra work?

any info appreciated

--Jordan
 
I was thinking that wood requires less extra work to finish than aluminum (were you have to buff it and anodize which is not that extra work but costly).

If you using heatsinks and your case doesn't act as heat dispersion surface, probably wood be recommended ( I like maple). The other alternative might be acrylic. You can line the interior with self adhesive mumetal sheets, which were available on a forum from Dave (planet10).

Another way to shield just the chip, would be using copper plate when attaching the chip to the heatsink. You cut the small plate, slightly bigger than the chip, and place it on top of a chip using two screws on ea. side. It is much better way to press the chip to the sink (than a single screw), as well as providing extra shielding.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.