NVA 'TFS': 'The Future Statement' lands at Marco Towers...

[Figlet108]

Hi Tim,

thanks for your support - you are spot on with all your points. And it's not just R&D costs - the cost for supporting this kind of product is considerable, especially when done properly and with care.

It's not like selling an amp or a cable or a speaker - you sell those and you can more or less forget about them. You sell someone a computer and you are committed for years.

I'm sorry you don't like the case though - I put a lot of time and attention to detail in the case design and have got lots of compliments on it.

However regardless whether someone likes the case or not at least no one can say it looks like a computer...

[Tim]

Hi Tim,

thanks for your support - you are spot on with all your points. And it's not just R&D costs - the cost for supporting this kind of product is considerable, especially when done properly and with care.

It's not like selling an amp or a cable or a speaker - you sell those and you can more or less forget about them. You sell someone a computer and you are committed for years.

I'm sorry you don't like the case though - I put a lot of time and attention to detail in the case design and have got lots of compliments on it.

However regardless whether someone likes the case or not at least no one can say it looks like a computer...

Jeez, you got that right about a computer Jason - I learnt the hard way and now refuse to build them for my mates anymore!

Cases are personal and I'm a bit old fashioned I guess, square small and black for me which is a bit dull really, but yep you are right, it doesn't look like a computer and ultimately it's all about what it sounds like that counts

Good luck with it too, it's a long slog getting people to accept that a computer can produce good audio, but we're getting there

[Figlet108]

Don't be too hasty. For serious listening I still prefer to spin a CD. Music is ripped from a CD at a far higher speed than the disc would normally be played at. I am not convinced that the ripped data is always the same as the slowly played data. The Reed-Solomon error correction employed in a CDP during playback is lacking in documentation as far as its operation is concerned during the CD ripping process. So I am not satisfied that the R-S truth table is adhered to during a rip.

Stanley I thought it worth quickly commenting on this.
The best way to rip a CD is use something like dbpoweramp that basically guarantees you have made a bit perfect rip.
First of all you need to have that error correction switched off for ripping. You don't want algorithms to be guessing at what the right data might be, you want it to be read properly off the CD with 100% certainty.
In addition programs like dbpoweramp compare the rip you have just made with a database of known bit-perfect rips and can validate that you have made a bit perfect rip.
In contrast a CD player will almost certainly never play a CD with bit-perfect precision when in stream mode as it needs to play in real-time.

[Figlet108]

I learnt the hard way and now refuse to build them for my mates anymore!

We've all been there
My friends and family know better than to ask anymore...

[Welder]

Thank you very much for taking the time to try to cover at least some of my questions Jason.

The proof of course is in the listening and I will make some effort to get to hear your product.

I find the whole computer/file based audio thing really interesting. A bit sad in a way that computers didn't figure more highly when I was at school and University.
I used to hate them but now I'm getting rather fond of messing about with them.

As you say, excellent audio can be had from all three of the major operating systems.
This doesn't alter the fact that the best servers I've heard have been Linux based. But, as you write, its a bit of a religion thing and there's plenty of that in audio anyway so I wont feel erm, unusual in this respect.

I'm a fortunate in being able to build my own server and I'm also fortunate in having a friend who develops audio software for a living so you never know, we might just come up with something interesting when I've stopped scrapping boards and burning crystals and spending my money on building linear power supplies and USB interfaces.

Obviously I'm not likely to be a customer, but that doesn't stop me wishing you every success with the project.
Having spent; possibly wasted, quite a few pounds on building then scrapping and building again I can only state again that if at £1800, your server can deliver the benefits of file based audio to those not so happy messing about with computers and compete with CD players and vinyl replay systems way above this price line, you'll have achieved in a single product what I've been banging on about since I arrived at this forum.

Oh, the glowing Perspex tubes, brilliant sales strategy.

[StanleyB]

Stanley I thought it worth quickly commenting on this.
The best way to rip a CD is use something like dbpoweramp that basically guarantees you have made a bit perfect rip.
First of all you need to have that error correction switched off for ripping. You don't want algorithms to be guessing at what the right data might be, you want it to be read properly off the CD with 100% certainty.
In addition programs like dbpoweramp compare the rip you have just made with a database of known bit-perfect rips and can validate that you have made a bit perfect rip.
In contrast a CD player will almost certainly never play a CD with bit-perfect precision when in stream mode as it needs to play in real-time.

I wish I had your level of confidence in the process, but after having had to endure years of training by some CDP manufacturers as part of my work then, I have a different level of insight into some of the "bit perfect" check sum validation process. I can't remember the exact value (it could be 27 or 30), but there is a checksum calculator that is used to validate the data on a CD. It is this validation process that allows for a hole to be drilled into a CD and the data to still be passed as correct.
I refer back to some training manuals that I lent to someone, but which contains some valuable technical information on this. It is infuriating to me that I don't have access to it any more. Basically what it boils down to is that bit perfect rip is a misnomer. A more accurate term would be checksum accurate rip.

[Welder]

I wish I had your level of confidence in the process, but after having had to endure years of training by some CDP manufacturers as part of my work then, I have a different level of insight into some of the "bit perfect" check sum validation process. I can't remember the exact value (it could be 27 or 30), but there is a checksum calculator that is used to validate the data on a CD. It is this validation process that allows for a hole to be drilled into a CD and the data to still be passed as correct.
I refer back to some training manuals that I lent to someone, but which contains some valuable technical information on this. It is infuriating to me that I don't have access to it any more. Basically what it boils down to is that bit perfect rip is a misnomer. A more accurate term would be checksum accurate rip.

Funnily enough, this is my understanding of the checksum too. Checksum perfect may not be perfect at all.
But then again interpolation isn't perfect either and sound still seems to come out.

[StanleyB]

Funnily enough, this is my understanding of the checksum too. Checksum perfect may not be perfect at all.
But then again interpolation isn't perfect either and sound still seems to come out.

That too is part of the whole Solomon-Reed error correction table.
I have somewhere in my CD collection in the loft a couple of test discs that were issued to a couple of engineers like me in the 80's as part of their training course in Belgium (some of the Philips CD players were once made in Belgium, where they also had a training center). The discs have some odd errors embedded in them. The one that is of importance to me when discussing bit accuracy is the 'fingerprint error test". Basically there is very little data on a quite considerable stretch of the disc, but it plays perfect!
BUT: the same musical track is used on another test disc in that collection. If you play both discs at the same time and using identical players, at certain parts in the music there is a distinct difference to be heard. It could be in the dynamics in some areas, or in the instruments. If you swap the discs round, the differences transfer themselves to the other player.
This has baffled me for years. Whilst I have devised my own theory why there is a difference in two identical pieces of music, but with one being covered by fingerprints on the disc, I am of course still short of concrete evidence if my theory is right or not.

[Audioman]

That too is part of the whole Solomon-Reed error correction table.
I have somewhere in my CD collection in the loft a couple of test discs that were issued to a couple of engineers like me in the 80's as part of their training course in Belgium (some of the Philips CD players were once made in Belgium, where they also had a training center). The discs have some odd errors embedded in them. The one that is of importance to me when discussing bit accuracy is the 'fingerprint error test". Basically there is very little data on a quite considerable stretch of the disc, but it plays perfect!
BUT: the same musical track is used on another test disc in that collection. If you play both discs at the same time and using identical players, at certain parts in the music there is a distinct difference to be heard. It could be in the dynamics in some areas, or in the instruments. If you swap the discs round, the differences transfer themselves to the other player.
This has baffled me for years. Whilst I have devised my own theory why there is a difference in two identical pieces of music, but with one being covered by fingerprints on the disc, I am of course still short of concrete evidence if my theory is right or not.

Hasn't there long been a theory that the harder the error correction has to work the more effect there is on sound reproduction ? Stan Curtis used to produce a device that displayed errors that went with his CD1 to show this if I am not mistaken. I think this is the reason behind the various CD cleaning and repair products on the market and perceived (by some at least) improvement in sound with their use. For my part I have always treated CD's with equal respect to vinyl never buying the 'will play with jam on it' marketing.

[Clive]

The whole concept of reading data optically in the way that CD does has always bothered me. It just goes to show that in an ideal world for file-based audio we could do with having the original files rather than relying on rips.