Yggdrasil Glitch Fixed

[The Computer Audiophile]

This one shouldn't slip under the radar. Mike at @schiit fixed this a while ago, but only recently announced it. I've had the fix in my Yggdrasil but couldn't say anything :~(

 

I'm sure @manisandher will jump for joy about this (not really) :~)

 

 

Here is Mike Moffat's post in full:

 

"I admit it, yeah, the rumors of modification are true. Over a year ago, we implemented a software modification for the notch which used to be visible at -90db waveforms on all multibit DACs. This modification had a user at another forum in a state of high dudgeon that we could claim our units were high end with such a glaring anomaly. Of course he championed a much more expensive competing unit over and over again which was free of such pox. @atomicbob did not seem to be unduly concerned, measurements champ that he is. I didn't either, completely unconcerned about an inaudible flaw. I corrected it with software tweaks and we have a newly clean -90db measurement.

You may wonder, why didn't we announce it? Well, it goes like this: on our upgradable DACs, we from time to time make running changes, such as the software change above. A few are are compelled, due to parts end of life. The majority, however, are minor tweaks we make for the purpose of improvement. It makes little sense to call an Yggy back for a minor upgrade where $10-$15 worth of upgrades are involved, either for the client or ourselves. So we remain muzzled and do the upgrades when the units come back for more major upgrades.

On major upgrades, such as one which would involve new analog, digi, or input boards, we would implement the changes and let them sit several months to a year before announcing anything. This is so they have time to settle. This is the biggest reason to avoid purchasing a used Yggy. Not that I am attempting anything punative – it is just a system that works reliably, and is fairest on an hourly basis in client cost.

Exceptions are when we offer an accross the line upgrade such as the Gen V USB which pretty much fixes USB, at least until we come out with our own chipset implementation, which will take a year or two. Oops – never mind. These upgrades are inexpensive and effect huge change. So we offer them.

Are we going to offer analog upgrades? Is Apple going to upgrade their phones? Of course. The difference is that we will conceal our upgrades for some period of time before we offer them."

 

 

7 hours ago, Speed Racer said:

I just read that over a year ago, Mike Moffat at Schiit Audio implemented a software modification to remove the "notch" or "glitch" from their multibit DACs.

 

https://www.head-fi.org/threads/what-a-long-strange-trip-its-been-robert-hunter.784471/page-335#post-13778331

 

This means you can buy any new Schiit multibit DAC and have no worries about the infamous "glitch" or any impact it may or may not have on the audio you listen to.

 

It also means that my Yggdrasil has the same modification.

 

[firedog]

17 minutes ago, The Computer Audiophile said:

This one shouldn't slip under the radar. Mike at @schiit fixed this a while ago, but only recently announced it. I've had the fix in my Yggdrasil but couldn't say anything :~(

 

I'm sure @manisandher will jump for joy about this (not really) :~)

Wow, so that whole ridiculous thread was for nothing....

But the "glitch" it out there on the Net, so reports of it will never die. I'm sure it will come up again, and someone will claim he hears it, even though it isn't there.

[mbain]

Finally a definitive conclusion.  This is indeed a rarity.  We need to arrive at more of them.

[bobbmd]

What exactly was this 'glitch'-something you could actually hear or something perceived during listening involving the sound quality? Thanks-just curious because I have never had any problems heard or perceived with any of my my Schiit products from Magni/Modi/Fulla/Gungnir except I thought the Magni became tiresome/harsh after hours of listening 

[firedog]

7 hours ago, bobbmd said:

What exactly was this 'glitch'-something you could actually hear or something perceived during listening involving the sound quality? Thanks-just curious because I have never had any problems heard or perceived with any of my my Schiit products from Magni/Modi/Fulla/Gungnir except I thought the Magni became tiresome/harsh after hours of listening 

Generally thought to be inaudible, and logically (measurements) is inaudible.

 

Some individuals who didn't like the sound of the Yggy, or certain aspects of it, attributed what they didn't like to the glitch. But there wasn't actually shown a connection between what they were hearing and the glitch. 

 

At this point we don't even know if some of the people who said they could hear the glitch were even listening to a DAC that has the glitch, or one where it had already been (unannounced) fixed in software. 

[mansr]

I'm curious how they fixed it. Did they change some setting in the DAC chip, or are they manipulating the audio data somehow?

[opus101]

My guess is they employed the same trick Philips did in the SAA7220 - move the zero crossing point by adding an offset. Only a guess mind, I have no other idea how a hardware problem can be fixed in software.

[Speed Racer]

Hardware problems are fixed or bypassed in firmware/software all the time. 

[The Computer Audiophile]

I received a new DSP chip for my Yggy when the Gen 5 USB arrived. Not sure if the two were related or what. 

[mourip]

2 hours ago, The Computer Audiophile said:

I received a new DSP chip for my Yggy when the Gen 5 USB arrived. Not sure if the two were related or what. 

 

Did they send you the chip and USB board or an upgraded unit?

[The Computer Audiophile]

Chip and board. 

[ssh]

Hmm, I just got the board.

[Speed Racer]

They probably used the logic that this was a reviewer reviewing the Yggdrasil with the Gen 5 USB board and he might take measurements. So let's give him a new chip too just in case he does measure. 

[The Computer Audiophile]

Not sure if the DSP chip was required to make the Gen 5 board work in the Yggy. Anyone else know?

[JoeWhip]

No, I think the new DSP chip you received contains the new software update to "fix" the glitch as your DAC is older. Schiit does not send the chip out normally as they do it in house when the unit is sent back for updating or repairs. It has nothing to do with the new USB board. MM has stated that the new software does not change the sound of the Yggy, just removes the glitch from showing up in measurements. 

[FredM]

I can understand the advice to buy a new device, to be sure of the latest/best quality. For me -personally- this affair somehow harms the trustworthiness for future/second hand users, ehat are the odds one buys a good functioning model?

 

I'll try to explain myself. Firstly by separating minor upgrades and bug fixing. For minor upgrades, i.e. due to an end of life of certain parts, are fully understandable. For bug fixing I'm surprised that in recent situations it's hard to find out if an known issue is fixed:

- DAC glitch (which models are fixed)

- Issues early Eitr models (i.e. cable, frequency sensitivity. It seems these issues are fixed now)

- 50 Ohm BNC connector on Yggy (can be -after owner request or at a repair/upgrade moment- be swapped) 

 

It would be nice to have insight in which model and version has the right quality for the use of the second hand buyer, or know issues that can or should be addressed/fixed (i.e. If a buyer couldn't care less which BNC connector is used -because he will only use Coax- fine, but when he plans using BNC and the DAC isn't fixed, he'll get an unexpected surprise. Or of course, a second hand owner has the same right for fixes as a first owner.

 

A nice, perhaps inspiring method is used by Naim. They have an overview with the major product changes (https://www.naimaudio.com/product-history) 

 

 

[mindbomb]

I heard people say this was caused by having 20 bit dacs being fed truncated 24 bit data without dithering. If that was the case, they could have fixed it by just adding the functionality to the dsp.

 

 

[Don Hills]

On 10/17/2017 at 10:57 PM, opus101 said:

My guess is they employed the same trick Philips did in the SAA7220 - move the zero crossing point by adding an offset. Only a guess mind, I have no other idea how a hardware problem can be fixed in software.

 

Your guess is correct - they added an offset in a digital filter.

[opus101]

4 hours ago, Don Hills said:

 

Your guess is correct - they added an offset in a digital filter.

 

That's been confirmed by Mr Moffat himself? If so then its not truly a fix for the glitch merely a move. The glitch is still there just doesn't show up on a  -90dB sinewave any longer as now that waveform no longer passes through digital zero. With the new firmware, perhaps it'll show on a -60dB sinewave, relative to the wanted waveform it'll look a lot smaller.

 

[GUTB]

I think the glitch was more about it being embarrasing after being exposed by Stereophile than it was to do with SQ.

[Don Hills]

3 hours ago, opus101 said:

 

That's been confirmed by Mr Moffat himself? ...

 

 

''The secret to getting rid of a minute amount (-120db) of anomaly was a bit of offset added into our filter, which I had previously done at Theta Digital some thirty years prior to get rid of a significantly larger error. " - Mike Moffat

 

[Speed Racer]

18 minutes ago, Don Hills said:

 

 

''The secret to getting rid of a minute amount (-120db) of anomaly was a bit of offset added into our filter, which I had previously done at Theta Digital some thirty years prior to get rid of a significantly larger error. " - Mike Moffat

 

 

Notice Mike said "getting rid of" and not "moving" or "hiding".

[Don Hills]

Yes. So?

[Speed Racer]

1 hour ago, Don Hills said:

Yes. So?

 

 

Look at this post:

 

 

[Don Hills]

Yes. So?