This should be illegal… PC Audio Product Scam

[Nystemy]

46 minutes ago, papajo said:

RAM noise filtering actually would make sense for OC e.g a particular 5950x I had could not do 3800Mhz 1:1 with a particular mobo.. yet most can and the same cpu with the same ram but other mobo could do 3800 1:1

it probably has something to do with noise, like the communication between ram and the CPU's ram phy would look somewhat like this 

and probably on that mobo (where I couldnt get 3800 1:1)  looked a little off like in the left side  but if it would be possible to "iron out" the anomalies by adding some sort of a filtering stick in a slot or somewhere else in the mobo maybe that indeed could help. 

Both the left and right signal looks completely fine. Both have rather exceptional margin to the no go zone in the middle.

In regards to adding filtering into another RAM slot on the same channel.
This wouldn't help. And at worst would cause unwanted reflections and making the signal integrity even worse.
At best, it wouldn't really do anything to the signal of interest. Since the path between our memory module and CPU doesn't go through this slot, it just goes past it. And this means that there really isn't anything meaningful we could do here.

The best thing we can do is to add termination resistors to quench any reflections from happening. But this isn't filtering the signal of interest in the slightest.

 

At these frequencies signals aren't just "a state" that a data line happens to be in. But rather a wave propagating down the line. When our signal of interest reaches the fork in the road, it goes both ways, one part to the CPU, the other up into the second memory slot. If the distance into the memory slot is significant, it will cause distortion as it reflects back. If it is terminated, our wave will just disappear into the void _{(as is the case if one has two memory modules installed on the same channel, since they have termination resistors included)}. And if the stub is suitably short, then the signal won't have much interest propagating that way to start with. _{(as is the case for most unpopulated memory slots. Unless it uses a T type typology... Daisy chain is better if it is unpopulated. However, 1 slot per channel is always better, unless one butchers the board layout...)}

If we make a filter module for the data lines. Then for it to do anything, it needs to react and counteract any unwanted signals. But by the time this reaction starts propagating down to where the fork in our road is, then our signal of interest is already long gone, so our reaction instead just butchers the data that now travels through. This isn't wanted.

So no. We can't add a filter module into a slot. The laws of time forbids it from doing anything of value.

However, a termination module could be of use. But this isn't filtering. _{(honestly speaking, the "for the looks" RGB modules should honestly have termination resistors on them, then they would at least do something more than just look fancy... And for DDR5 this is going to become a little bit more relevant than with DDR4 where it is only marginally noticeable when pushing the limits.)}

[AndreiArgeanu]

It's hilarious that those memory modules use NEC Tokins, even though they basically do nothing. Also the PS3 suffers relatively rare from Tokin failure, it's actually the RSX (gpu) that's the problem typically due to bad underfill leading to a similar issue to the RROD on the 360. The underfill isn't rated for the temperatures the PS3 typically runs at under default fan configuration and is completely useless leading to cracked solder balls. Replacing the tokin's will also heat up the CPU, enough to make a connection again and become usable for a while. The only real fix is replacing it with a newer RSX. And the RSX wasn't the only chip facing this problem, other NVIDIA made chips at the time had the same issue since they used the same underfill and they got sued over this.

[ImorallySourcedElectrons]

2 hours ago, Quasido said:

Putting myself on the line to defend a different product that works for me. The iFi iDefender. I had terrible GPU noise in my headphones plugged into my DAC when my active studio monitors were plugged into the mains which disappeared when I unplugged them. However I had bought an individually switched extension lead so I don't have to keep swapping plugs so I don't want to leave them unplugged. There are probably much better products/companies than what I bought but it allows the USB data to get to my DAC from my PC with power from my phone charger, thus breaking the earth loop that was going through the USB power.

 

There are other USB isolators on amazon but those only seemed to support a part of the USB 2 spec and only 100 mA power.

That says more about the quality of the grounding of your setup than anything else. Might want to call an electrician over to check if all the circuits are earthed properly (I assume the speakers and PC are on a different circuit in your house). Additionally, replace any extension cords, etc. that you're using in that setup. But indeed, a capacitive or optical isolator does kind of work for such problems, it's just not a very pretty solution.

 

1 hour ago, papajo said:

RAM noise filtering actually would make sense for OC e.g a particular 5950x I had could not do 3800Mhz 1:1 with a particular mobo.. yet most can and the same cpu with the same ram but other mobo could do 3800 1:1

it probably has something to do with noise, like the communication between ram and the CPU's ram phy would look somewhat like this 

and probably on that mobo (where I couldnt get 3800 1:1)  looked a little off like in the left side  but if it would be possible to "iron out" the anomalies by adding some sort of a filtering stick in a slot or somewhere else in the mobo maybe that indeed could help. 

Those eye diagrams don't demonstrate what you think they demonstrate I suspect.

 

But filtering wouldn't do much. The signal integrity and power quality requirements for DDR4 and DDR5 are so stringent that if you had any noise on them, the bus wouldn't work in the first place. And not achieving those clock rates will most likely be down to individual transmission lines or the bus going out of spec. Sometimes you got to do odd things to match phase/delay at the desired frequency, which can screw over your total bandwidth on a bus system like this.

[sub68]

I know my view is gonna be weird as heck, But I read alot of Head fi fourms (back in 2020 I think don't remember) But the important part is power points, like psu's (corsair, seasonic) high quality units generally don't have problems this what I read but then again IDK. Anything on DIGITAL SIDE is snake oil IMO.

 

But what I know.

The more important point is surge and power conditioners, This is only gonna imply pro audio/higher end audio when having mixing consoles it depends but these can be sensitive to surges and generally weird stuff (weird stuff is more applicable to analog consoles).

I have to used in the past, ground isolators due to XLR's that were in plugged into some speakers having buzzing issues.

 

Interesting thing I have a laptop that has no battery but I bring it on car trips, I have a 12v DC power inverter (good ancient GE one) that is plugged into the car with that into a surge protector, the problem is this setup is not grounded (neutral ground) and I use wired headphones I hear a very distinct buzz. the cool part is if I touch a cross member of the seat the buzz goes away (or put the inverter on the cross member). 

SO basscily I don't think POWER audiophile products arent snakeoil.

[Avacado]

I would've loved to see the "sata noise removal" thingy attached to a sata CD player with the little output on the front because that seems like an area if anything it would benefit

[manikyath]

even *if* these things worked to smoothen the power supply rails.. there's one glaring issue here:

they do absolutely nothing against EMI inside the case. by design the audio traces are right next to the GPU's PCIe slot.. i dont like this design and i dont understand why they keep doing this.. but it is how it is i suppose.

 

my previous motherboard had the "test file" squeaks quite audible on the microphone signal while gaming... i literally put a piece of tin foil (insulated with tape, for obvious reasons) over the audio components, wedged (with an exposed edge) to the case metal, and all of the squeak went away entirely. 

 

3 minutes ago, Avacado said:

I would've loved to see the "sata noise removal" thingy attached to a sata CD player with the little output on the front because that seems like an area if anything it would benefit

given i've had one of those hooked up to a power supply with a bunch of cap juice caked on the PCB, and still had perfect audio.. i doubt it's gonna make much difference.

 

i'd hazard a guess that the stuff going on in the CD drive itself would make more noise than the rest of the system combined.

[Avacado]

6 minutes ago, manikyath said:

even *if* these things worked to smoothen the power supply rails.. there's one glaring issue here:

they do absolutely nothing against EMI inside the case. by design the audio traces are right next to the GPU's PCIe slot.. i dont like this design and i dont understand why they keep doing this.. but it is how it is i suppose.

 

my previous motherboard had the "test file" squeaks quite audible on the microphone signal while gaming... i literally put a piece of tin foil (insulated with tape, for obvious reasons) over the audio components, wedged (with an exposed edge) to the case metal, and all of the squeak went away entirely. 

 

given i've had one of those hooked up to a power supply with a bunch of cap juice caked on the PCB, and still had perfect audio.. i doubt it's gonna make much difference.

 

i'd hazard a guess that the stuff going on in the CD drive itself would make more noise than the rest of the system combined.

how about a really really high quality cd player and have it really far a way from the other components

[HenrySalayne]

26 minutes ago, sub68 said:

SO basscily I don't think POWER audiophile products arent snakeoil.

I mean you basically stated when you would need these devices:

- bad power supplies

- bad grounding

- bad shielding

Basically all the things any well-made audio product should have already taken care of (we don't need more band-aids, we need less cuts).

[papajo]

45 minutes ago, ImorallySourcedElectrons said:

That says more about the quality of the grounding of your setup than anything else. Might want to call an electrician over to check if all the circuits are earthed properly (I assume the speakers and PC are on a different circuit in your house). Additionally, replace any extension cords, etc. that you're using in that setup. But indeed, a capacitive or optical isolator does kind of work for such problems, it's just not a very pretty solution.

 

Those eye diagrams don't demonstrate what you think they demonstrate I suspect.

 

But filtering wouldn't do much. The signal integrity and power quality requirements for DDR4 and DDR5 are so stringent that if you had any noise on them, the bus wouldn't work in the first place. And not achieving those clock rates will most likely be down to individual transmission lines or the bus going out of spec. Sometimes you got to do odd things to match phase/delay at the desired frequency, which can screw over your total bandwidth on a bus system like this.

It was a random eye diagram to serve as a visual aid to where I was getting at I did not meant to use it as a reference. 


Having that said I am not going to pretend that I know this stuff I am not an engineer working at micron or samsung etc. 

But being a hobbyst overclocker I dwelled a little deeper on how these thingies work and I wouldnt be so negative in believing that filtering couldnt change anything I mean the basic idea of bit flipping in DDR4 can result into a cleaner signal just because when most of the bits are high (cause of the characteristic of the POD being high then nor current flows) the distortion is lower 


(this is actually a screenshot of what I am talking about and not just a random graph for visual aid, notice that the "X"s are much smaller/tighter which means less time variation and thus jitter which I presume could make the difference e.g in reaching 3800 or failing to do so -given that same cpu and ram in a different mobo can reach it at a given temp and voltage- ) 



 

[sub68]

6 minutes ago, HenrySalayne said:

bad grounding

That part is makes sense due to no ground via a car power invertor.

7 minutes ago, HenrySalayne said:

Basically all the things any well-made audio product should have already taken care of (we don't need more band-aids, we need less cuts).

That is true alot of audio products have pretty good power (unless its my over 10 year old mixer that had a cap about to blow.)

15 minutes ago, HenrySalayne said:

bad shielding

Not really, those XLR's are expensive as heck

[Avacado]

1 minute ago, sub68 said:

Not really, those XLR's are expensive as heck

Yeah imagine if expensive equipment just forgot the most basic things such as good shielding it would be ridiculous

 

keep every piece of equipment in a tiny faraday cage

[sub68]

Just now, Avacado said:

Yeah imagine if expensive equipment just forgot the most basic things such as good shielding it would be ridiculous

I am talking about cables that are like 10 dollars per foot.

[papajo]

1 minute ago, sub68 said:

I am talking about cables that are like 10 dollars per foot.

You mean the sucker cables ?

[sub68]

1 minute ago, papajo said:

You mean the sucker cables ?

Huh? I am thinking of morgami

[papajo]

5 minutes ago, sub68 said:

Huh? I am thinking of morgami

morgami salami if its 10+ dollars a foot I am sure that there is a $10  or sub $10,  2+ meter cable that does exactly the same job and will last as long. 

[sub68]

2 minutes ago, papajo said:

morgami salami if its 10+ dollars a foot I am sure that there is a $10  or sub $10,  2+ meter cable that does exactly the same job and will last as long. 

ish

 

[papajo]

4 minutes ago, sub68 said:

ish

 

I pushed play then I saw this guy.... I remember this guy... I saw him in a couple of other videos 100% shilling and always showing stupid expensive low value brands e.g the last video i remember him making is comparing blue yeti mics that cost 400+++ and all are garbage btw (For the price at least) and he found the most expensive of the yetis (I think like 3000? not sure)  much better while they all had the same size diaphragm lol and all sounded the same... That's like showing me an unboxtherapy video or the other guy martis talking about apple 

[sub68]

1 minute ago, papajo said:

That's like showing me an unboxtherapy video or the other guy martis talking about apple 

What ever I watch this guy because he test every mic ever made, look The point is shielding is important, I have had problem with that with bad XLR's in pro setting cables get run all over power ETC and we have had some problems with those but are nicer ones never usally had a problem.

[papajo]

37 minutes ago, sub68 said:

What ever I watch this guy because he test every mic ever made, look The point is shielding is important, I have had problem with that with bad XLR's in pro setting cables get run all over power ETC and we have had some problems with those but are nicer ones never usally had a problem.

Ι gave the video a gander.. besides the cables performing the same whenever they did not it was like on minus double digit db... that's SUPER low noise lol I cant describe to you how low noise that is but take into account that the db levels dont scale linearly but logarithmic, for some perspective.

e.g 


20 db difference from 120db  to 140 is about 99 watts of power 

20 db difference from 20 db to 40 is less (a lot less) than 1 watt of power. 

and the measurements were at negative double digits for all cables... 



Basically the isolation is so good even on the cheap cable that much of the (practically non existing) noise may be due to not having precise or noise free instruments to measure it lol 

Or in other words they are all the same but he chooses to do some BS measurement to showcases that the expensive is a little better because numbers are looking a little better but even if you had bat ears you wouldnt be able to tell the difference and hence this measurement is irrelevant. 

It is like saying "I am faster than you because I can run at 20.0000000000000000000000000000004 mph and you can run 20.0000000000000000000000000000003 mph "  

[sub68]

35 minutes ago, papajo said:

that's SUPER low noise lol I cant describe to you how low noise that is but take into account that the db levels dont scale linearly but logarithmic

Fair, I think the problem we had was very werid and discovered it might have not been the cable, but don't remember as this crap happens every day

[ihatesleepyhead]

What were the wired headphones Linus was wearing?  I didn't see them mentioned in the description.  

[mononymous]

Have an isolated power socket and a "audiophile" grade power cable. 

[xnamkcor]

Not saying this stuff works, but if I remember correctly, "Solo" in Audacity doesn't completely mute the other channels.

[sub68]

2 hours ago, ihatesleepyhead said:

What were the wired headphones Linus was wearing?  I didn't see them mentioned in the description.  

its a rode model.

[bluehawk]

Let's bring back sound cards, going to slot in my sound blaster