EtherREGEN: All posts about firmware moved here

[Superdad]

22 minutes ago, Dutch said:

Since there’s no way back I now really hope there’s a possibility Alex&John will release a firmware version with just the EEE issue fixed (or even the original with that EEE bug present)

 

That is not going to happen. We are not making a "tunable" switch!

The new firmware code results not only in fixing of the EEE issue, but it improves both transmit and receive performance of the PHYs in the switch for the RJ45 jacks on the ‘A’ side. It is technically—and sonically—better. Period. 

[Superdad]

5 minutes ago, Iving said:

... i don't like the loss of body - the music is technically better but has lost its soul

I'll give it a while - see if there is a cooking effect

otherwise i'd like to understand how *this* firmware change could have such dramatic effects

 

Sorry Martin, but there is no “loss of body or soul” (though that is a great jazz standard ) with the new firmware. I have compared them side-by-side (two new units with same zero hours on them), and while the new firmware is better, the difference is not huge and there is no loss of bass.

[Superdad]

2 minutes ago, Iving said:

likewise whether firmware is better or not

i can tell you categorically that the difference is very substantial and that the bass is significantly diminished

 

Since you can not A/B/A compare, you can not know that for sure. Your mind may be imagining the difference. 

Of course if you are not happy with your EtherREGEN we will cheerfully refund your purchase. 

[Superdad]

10 minutes ago, Iving said:

please don't level that anti-subjective crap at me

 

There is nothing “anti-subjective” about performing simple A/B/A/B listening comparisons. (Not talking about blind, A/B/X, etc.) I’ve been doing such for more than 4 decades, even down to with things like resistors and small parts changes. Sometimes it takes a bit of back and forth to get a handle on if something is truly better, worse, or the same—as well as to identify any trade offs.  And with many actual advances, the change can reveal the true nature of the other components in the system and recordings—for better or worse.

Such is the realm that the EtherREGEN works in...

[Superdad]

Chill guys! 

Since John and I are the only people on the planet currently able to compare the current firmware (A/B/A/B) to the original (and despite what someone implied I have not sent a "rollback" file to ANYONE), I'd say it was amusing how a FEW of you are getting hung up on what you THINK you remember the original sounding like.  Only I am not amused, as all you are doing sowing fear, uncertainty, and doubt.

The new firmware--aside from fixing the EEE bug--improves transmit and receive performance of the PHY circuits of the 'A'-side RJ45 ports. That is all. We reserve the right to produce a technically superior product.  Whoever said they thought we spent months "voicing" the EtherREGEN to its original release "sound" and then rushed in a change without listening is 100% wrong on both counts.

I think one vocal person is using only the optical port on the 'A' side and still claiming the new firmware "ruins" the sound him. But the code change we made does not touch the SFP cage's SGMII interface at all, so it is just imagination. 

 

I am going to listen again myself over the weekend and then make a decision about if we will make the technically poorer performing firmware file available. Until then, please tone it down a bit.  Thanks,

--Alex C. 

[Superdad]

On Tuesday, November 26th we released to the public a firmware update applicable to the first batch of 250 EtherREGENs which were shipped in November. My blog post about it was at our web site here:

https://uptoneaudio.com/blogs/news/etherregen-huge-success-new-firmware-for-november-units-reviews

 

The update successfully fixes a connectivity issue that never showed for us or our beta testers, but did for some users. And as @JohnSwenson and myself have explained, another change was made to the main switch chip code which technically improves the transmit and receive performance of the PHY circuits connected to the 'A'-side RJ45 ports.

 

The revised code does change the sound very slightly.  While most have enjoyed the new firmware or reported not noticing much difference, a very few of the many people who applied the update have expressed that they do not enjoy the EtherREGEN quite as much with the new firmware. However, they are relying exclusively on their last memory of the SQ before applying the update (which for most people required a break of about 20~30 minutes), as they are not able to reload the original code. Aural memory in this realm is a very tricky thing, and while I am by no means a dogmatic blind-A/B/X proponent, I have for decades collaborated on design decisions using sighted A/B/A/B/A testing.

 

As promised, last night I carefully repeated the comparison. (Yes, I have the original firmware file.) Full disclosure, I used two brand new EtherREGEN boards that had no time on them. This seemed more fair than using my personal unit and it allowed me to alternate boards in about 40 seconds.

As I mentioned, and can again confirm: There is a slight difference sonically. But nowhere near what a couple of the very vocal dissenters are claiming. Music with the EtherREGEN is still just as engaging; no bass is lost; no "soul" is lost. 

 

I've been at this game for a long time, and in any A/B/A/B/A design test I prefer to pick a clear "winner."  Yet last night I did not. And generally when I am unable to pick the "winner," it means the difference is not enough to worry about.  UpTone products are about the bigger things, and having folks be "wowed" by the use of our products is our goal.  Just as you all have been expressing in the EtherREGEN Listening Impressions thread and in the listening reports threads for all our other products. So while I respect others differing opinions, this firmware update is nowhere near as big a deal as their posts might lead you to believe.

 

As John explains (in this post), the PHY performance code change was not done to purposely alter SQ. It actually is an important part of the connectivity fix. Thus our extreme reluctance to make the earlier code available. Still, we will continue to consider the matter.

 

Someone suggested that people who wish to discuss the firmware issue ought to have a separate place to do so. I agree, and thus have started this thread for that purpose and moved all the related posts here.

 

Thanks and regards,

--Alex C.

[Superdad]

1 hour ago, simon_pepper said:

Application of the update was straight forward and well documented.  I understand that in-the-field firmware updates was never planned, hence the USB port being internal to the device - anyway it means we were allowed to open and view the actual PCB - nicely done and densely populated.

Thank you Simon.

 

1 hour ago, simon_pepper said:

Back in the system and all port responding - initial listening, I still have the soundstage space, and additional clarity over the unmanaged Cisco 5-port the EtherRegen placed. Is there more? I think so, however system is now playing the new Pink Floyd album and it all sounds mighty fine across the range, with good transient behaviour and fine detail resolution. So no issues here.

Nice.

 

1 hour ago, simon_pepper said:

Now if John has found additional scope on the PHY layer of the A-ports, is there any such opportunity for the B-port?

The very new, very special Ethernet protocol converter/PHY which sits just 2mm behind the 'B'-side RJ45 is not being programmed externally. Rather we use "strapping" resistors to set certain functional parameters.  And as far as I know (John may know differently--his is VERY intimate with that chip's 125 page datasheet) there are no PHY parameters to adjust.

 

[Superdad]

4 minutes ago, Johnnydev said:

I really do not understand that after so many reviews from you, now come with the grounding, whereas this was specifically described in advance? It has only brought doubt and unrest! 🥵

 

No worries gentlemen.  Everything comes out just fine in the end.

I think this has been educational for all of us. 

It is interesting how the insertion or removal of other networking gear seems to affect the perception of even the EtherREGEN's "magic."  Seems that simpler (fewer switches) is better!

[Superdad]

43 minutes ago, Puma Cat said:

I can move the ER just a little bit, and the fit/connection between the optical transceiver can disconnect. I know that it takes quite a bit of force to install the optical transceiver to get it to seat securely, and I apply quite a bit of force when inserting the optical transceiver into the SFP cage. Sometimes it connects up, but many times it does not. What I've consistently experienced is that it takes several attempts and a fair bit of "fiddling" to get it to connect up. 

 

We aside from the obvious--stop moving things around! --I agree that you ought to try another SFP transceiver.  

And John mentioned to me something along the lines of it being possible to have some of the differential pairs of the module not connect well, still have connectivity--but poor sound. Busy day, but maybe he will explain about that more properly at some point.

[Superdad]

14 minutes ago, Midnightrambler said:

Please could you confirm the second batch has the latest software or does it need updating?

You are already running the current firmware unless you received your unit in November (1st batch).