Bits Is/Ain't Bits: A Modest Proposal

[YashN]

I have read a number of Swenson papers.

 

Really? Which papers?

[YashN]

I say, show me the evidence in clear terms how that affects the sound coming out and, maybe, I will consider it.

 

If you think you have the baggage to comprehend such evidence, and I highly doubt it, as evidence has been provided by people in this very forum time and time again, why don't you do your own measurements and post them?

 

Said differently, I don't think you would understand the phenomena nor the measurement results if they are presented to you.

 

You want to be spoon-fed evidence and knowledge.

 

The knowledge is already here, you can access it if you're not too lazy.

 

You want to be convinced of something, but most of the people who could convince you are busy making better software and hardware for better SQ, things you can download, buy or use right now for immediate and obvious benefits.

[Fitzcaraldo215]

Not a red herring at all - it's a proven phenomenon and there's a research paper just out on it. It's something to worry about *in addition* to everything else which happens in a computer audio setup.

 

The 'everything else' contains a lot of issues related to noise and jitter and the D/A process even if bit-perfection is observed from source to USB receiver of the DAC.

 

This is the crucial point that many who are too focused on just bit-perfection miss.

 

So, if you follow closely, the RH issue is something to worry about before everything else, because if the bits are already messed up with in the DRAM buffers, you can't have bit-perfection.

 

Now, suppose you do manage to have bit-perfection in the DRAM, then there are a lot of system interactions which affect SQ, many of which are counter-intuitive even for people schooled in IT, because they have nothing to do with bit corruption at all.

 

They have to do with what happens with energy consumption and ground plane noise and signal integrity and they ultimately affect the DAC chip and the DAC clock and hence the SQ.

 

 

Ok, so, you alone, of all the posters here still support the notion of bit flipping. You presented a theoretical paper in that other thread, but I do not see any measurements anywhere of how significant or pervasive this problem is in any of our real world computer audio systems. I do not see any participants in that Row Hammer thread saying, after appearing here in this thread, that they believe in bit flipping, either. Except you, that is.

 

I will actually worry about it if and when someone measures some way, some how that it is an actual problem. And, funny, but it is not ever even mentioned as a problem in computer video applications which move much more data at much higher bandwidth than audio.

 

I understand noise and jitter and their influence, and I said so several times, if you read my comments, above in thus thread.

 

I disagree with you. I do not need to worry about ground plane noise. I just need to buy a DAC that has excellent audio performance when connected to the PC. It is the DAC designer's job to worry about noise, not mine. I just pick the best performing DAC. And, why is galvanic isolation not effective against such noise? Are there actual published measurements that support any of this and make it clear in English?

[esldude]

Ok, so, you alone, of all the posters here still support the notion of bit flipping. You presented a theoretical paper in that other thread, but I do not see any measurements anywhere of how significant or pervasive this problem is in any of our real world computer audio systems. I do not see any participants in that Row Hammer thread saying, after appearing here in this thread, that they believe in bit flipping, either. Except you, that is.

 

I will actually worry about it if and when someone measures some way, some how that it is an actual problem. And, funny, but it is not ever even mentioned as a problem in computer video applications which move much more data at much higher bandwidth than audio.

 

I understand noise and jitter and their influence, and I said so several times, if you read my comments, above in thus thread.

 

I disagree with you. I do not need to worry about ground plane noise. I just need to buy a DAC that has excellent audio performance when connected to the PC. It is the DAC designer's job to worry about noise, not mine. I just pick the best performing DAC. And, why is galvanic isolation not effective against such noise? Are there actual published measurements that support any of this and make it clear in English?

 

+1 and then some. He says you can't have bit perfection, yet it has been demonstrated many times we do in fact quite nearly all the time have exactly bit perfection in transmission. Still, he is going to fix the problem. Great if he can do it and long as it costs pretty much zero. Because it is pretty much a near zero probability with audio signals.

[sandyk]

+1 and then some. He says you can't have bit perfection, yet it has been demonstrated many times we do in fact quite nearly all the time have exactly bit perfection in transmission. Still, he is going to fix the problem. Great if he can do it and long as it costs pretty much zero. Because it is pretty much a near zero probability with audio signals.

 

Transmission errors are in my experience, very rare with normal data.

IMO, Bit perfection is overrated. Who cares , or even notices a few , perhaps masked errors, with either audio or video ?

Many LPs have far more obvious defects that may result in slightly noticeable clicks, but do they detract that much from the overall enjoyment?

I can rip a problematical, but not severely damaged CD, at higher speed in EAC, and the checksums don't agree with the rip that may vary in speed from 0.1 times to say 10 x , but eventually results in an error free rip. However, to my ears at least, the rip made at a higher but much more constant speed usually sounds better.

[YashN]

Ok, so, you alone, of all the posters here still support the notion of bit flipping.

 

Most of the issues we've been talking about in several threads here have nothing to do at all with bit flipping or bit corruption.

 

Simple? It doesn't seem you understand.

[YashN]

+1 and then some. He says you can't have bit perfection, yet it has been demonstrated many times we do in fact quite nearly all the time have exactly bit perfection in transmission.

 

It's been clear to me you still don't get it either, esldude. Bit-perfection is a requirement, but good SQ doesn't stop with bit-perfection...

[trithio]

It's been clear to me you still don't get it either, esldude. Bit-perfection is a requirement, but good SQ doesn't stop with bit-perfection...

 

All good except it does. From a media player developer's point of view and from a PC's own point of view the buck simply stops there. The PC's only job is to deliver the original, unadulterated bits. On time. That's it.

The rest is the DAC's job. And it includes taking care of the eventual noise coming over USB or whatever other interface, includes proper clocking of the signal and so on.

At best you can try to help the DAC by taking care of the electrical noise inside the PC. Not much to do with the bits though and _any_ PC sold in AD 2015 can deliver those bits in spades.

That is not such good news for CA because, as the mainstream hardware evolves, CA has to move further an further into the fringe zone to stay relevant. Identical checksums who sound different, bits are not bits anecdotes, you know the drill. Spent about a month around here and still havent seen anything remotely reliable coming from that fringe zone. Except the to-be-expected monsters.

[Fitzcaraldo215]

All good except it does. From a media player developer's point of view and from a PC's own point of view the buck simply stops there. The PC's only job is to deliver the original, unadulterated bits. On time. That's it.

The rest is the DAC's job. And it includes taking care of the eventual noise coming over USB or whatever other interface, includes proper clocking of the signal and so on.

At best you can try to help the DAC by taking care of the electrical noise inside the PC. Not much to do with the bits though and _any_ PC sold in AD 2015 can deliver those bits in spades.

That is not such good news for CA because, as the mainstream hardware evolves, CA has to move further an further into the fringe zone to stay relevant. Identical checksums who sound different, bits are not bits anecdotes, you know the drill. Spent about a month around here and still havent seen anything remotely reliable coming from that fringe zone. Except the to-be-expected monsters.

 

I agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson), Chicken Littles (the sky is falling), Music Men ( we got trouble right here in River City) and guys writing code on the PC end to solve problems they only believe exist without any generally accepted, measurable proof before or after.

 

I do think the sonic issues are largely on the DAC end, where there is considerable variation in approach. Even there, it is possible to do a lot of wild and crazy things in the name of "it sounds better to me" under totally uncontrolled listening conditions.

 

Audio just seems to bring out the worst in people. As if pre-CA high end audio were not already loaded with myths and staunch, but unproven beliefs, computers seem to make it worse. The mumbo jumbo of seeming technical "experts" does not help. The big picture and concrete objectives (preferably measurable ones) too often get lost and obscured in the bits and bytes detail.

[YashN]

I do not need to worry about ground plane noise.

 

Indeed, don't worry about anything, let people with far more knowledge than you'll ever have build and sell their solution while freely sharing their findings here.

 

This forum isn't for you.

 

Obviously...

[wisnon]

I agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson), Chicken Littles (the sky is falling), Music Men ( we got trouble right here in River City) and guys writing code on the PC end to solve problems they only believe exist without any generally accepted, measurable proof before or after.

 

I do think the sonic issues are largely on the DAC end, where there is considerable variation in approach. Even there, it is possible to do a lot of wild and crazy things in the name of "it sounds better to me" under totally uncontrolled listening conditions.

 

Audio just seems to bring out the worst in people. As if pre-CA high end audio were not already loaded with myths and staunch, but unproven beliefs, computers seem to make it worse. The mumbo jumbo of seeming technical "experts" does not help. The big picture and concrete objectives (preferably measurable ones) too often get lost and obscured in the bits and bytes detail.

 

LoL

 

Bet you thought Tesla was a kooky dilletante too.

[YashN]

LoL

 

Bet you thought Tesla was a kooky dilletante too.

 

Hehe!

 

Disparaging someone like Swenson must be one of the dumbest things I've seen happening here.

[Paul R]

Well, addressing your original question, bits are bits, and they can only hold one value at a time. Unfortunately, when you add in the transmission of those bits, you encounter all sorts of things that I am totally surprised you seem to be unaware of, to the point of challenging. Threshold values, timing, electrical and RF noise imposed on the system, etc.

 

I suggest if you want analogous examples, look more into VoIP. You can certainly get digital noise on a VoIP line, and the number one culprit is packet queuing delays. And only a certain amount of buffering can take place and still have the call appears to be in real time.

 

It is even worse with Audio, and there is more data flowing and the timing is still just as strict. Talk to the DAC makers here if you think that jitter and ground plane noise do not have an effect on the sound of a DAC. Quite simply, you are wrong, they do.

 

Now whether their effect bothers you, is audible to you, or is something you need to consider, that is personal choice.

 

And as for random bit flips - yeah - they *do* happen. Much more often than you would believe. Modern ECC memory makes them a non-issue, but they still happen. Here's an example of one.

 

https://blogs.oracle.com/ksplice/entry/attack_of_the_cosmic_rays1

[PeterSt]

It seems true. CA is River flooded with idiots. My pleasure.

[esldude]

Indeed, don't worry about anything, let people with far more knowledge than you'll ever have build and sell their solution while freely sharing their findings here.

 

This forum isn't for you.

 

Obviously...

 

 

YashN, I don't think anyone made you moderator of who should post here. It is downright rude on your part, and not your place to decide. You haven't been here long.

 

Chris as much as possible welcomes anyone. CA most definitely is not limited to those up to some bar of technical knowledge. He intends the place to be for anyone to further their enjoyment of computer audio.

 

Your out of line here YashN, and if you continue in that manner I would venture it is you for whom this forum isn't the place.

[sandyk]

YashN, I don't think anyone made you moderator of who should post here. It is downright rude on your part, and not your place to decide. You haven't been here long.

 

Chris as much as possible welcomes anyone. CA most definitely is not limited to those up to some bar of technical knowledge. He intends the place to be for anyone to further their enjoyment of computer audio.

 

Your out of line here YashN, and if you continue in that manner I would venture it is you for whom this forum isn't the place.

 

Dennis

Aren't you being a tad oversensitive here ? I fail to see anything in his last few posts that is any more confronting than some from the Objective side of the fence on a regular basis.

Kind Regards

Alex

[esldude]

Dennis

Aren't you being a tad oversensitive here ? I fail to see anything in his last few posts that is any more confronting than some from the Objective side of the fence on a regular basis.

Kind Regards

Alex

 

It isn't so much about being a confrontational post. It isn't one of having a different opinion. It a post saying someone doesn't know enough to take part therefore they should not be involved in the forum.

[wisnon]

That is not what he said.

 

He said that if the poster wants to feel that way (only kooks here), then perhaps this is not the place for him to be. Seems logical to me.

[kumakuma]

That is not what he said.

 

He said that if the poster wants to feel that way (only kooks here), then perhaps this is not the place for him to be. Seems logical to me.

 

I am not sure how you got that from Post #35 above. YashN quoted post #28 which had nothing to do with kooks.

[trithio]

I agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson), Chicken Littles (the sky is falling), Music Men ( we got trouble right here in River City) and guys writing code on the PC end to solve problems they only believe exist without any generally accepted, measurable proof before or after.

 

I do think the sonic issues are largely on the DAC end, where there is considerable variation in approach. Even there, it is possible to do a lot of wild and crazy things in the name of "it sounds better to me" under totally uncontrolled listening conditions.

 

Audio just seems to bring out the worst in people. As if pre-CA high end audio were not already loaded with myths and staunch, but unproven beliefs, computers seem to make it worse. The mumbo jumbo of seeming technical "experts" does not help. The big picture and concrete objectives (preferably measurable ones) too often get lost and obscured in the bits and bytes detail.

 

Expect to be crucified for that. At least twice

Plus, I think you forgot to mention the storytellers and/or salesmen.

Your post is somewhat harsh even for my taste. But you make at least one very good point: from software developers to hardware people to music business and other industry people and to the last joe-poster, almost no one seems to accept even the minimum amount of proof standards.

Everyone is free, beautiful and has a bat's ears . Pure Wonderland

[sandyk]

It isn't so much about being a confrontational post. It isn't one of having a different opinion. It a post saying someone doesn't know enough to take part therefore they should not be involved in the forum.

 

I don't have a problem with that for those who appear to be here mainly to troll, and aren't interested in the experiences of others. Of course, there will always be personality clashes, but if more than a few feel that the person concerned is a Troll, the suggestion that they may be happier in another forum isn't unreasonable.. I find that far less objectionable than downright nasty sarcasm and suggestions of mental instability, like a few on your side of the fence consistently post.

[elcorso]

I agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson), Chicken Littles (the sky is falling), Music Men ( we got trouble right here in River City) and guys writing code on the PC end to solve problems they only believe exist without any generally accepted, measurable proof before or after.

 

Lacks the most important, deaf people ... they are my benchmark.

 

I do think the sonic issues are largely on the DAC end, where there is considerable variation in approach. Even there, it is possible to do a lot of wild and crazy things in the name of "it sounds better to me" under totally uncontrolled listening conditions.

 

Oh my, a prayer for the DACs, must receive all possible garbage? But I could agree some DAC manufacturers make magic to make lovely music from a 'dirty' bit stream.

 

Audio just seems to bring out the worst in people. As if pre-CA high end audio were not already loaded with myths and staunch, but unproven beliefs, computers seem to make it worse. The mumbo jumbo of seeming technical "experts" does not help. The big picture and concrete objectives (preferably measurable ones) too often get lost and obscured in the bits and bytes detail.

 

You are again talking bad things about deaf listeners? Audio maybe bring out the worst from them, I hope that the music does not..!

 

You need fine and trained ears to be able to listen to 'myths'.

 

Cheers!

 

Roch

[sandyk]

agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson),

 

 

 

WTF ? John Swenson is a highly respected E.E. with wide experience, and is big on the measurement side of things.

 

Industry Voice | AudioStream

[Jud]

I agree. CA is like the Wild West and out of control. There are fire and brimstone preachers (Swenson)

 

Audio just seems to bring out the worst in people.

 

John Swenson "fire and brimstone"?? You *are* referring to the fellow who occasionally posts here, right?

 

You do realize how incongruous it is to be hanging around an audio forum saying audio brings out the worst in people?

[elcorso]

Expect to be crucified for that. At least twice

Plus, I think you forgot to mention the storytellers and/or salesmen.

 

You forgot the trollers, didn't you?

 

Cheers!

 

Roch