|
|
Post by routlaw on Dec 10, 2021 12:50:36 GMT -5
I'm curious to try it on my XMC-1 using the AES/EBU but I really like the analog out of the ERC-4 so I've been too lazy haha, but now I'm curious to try the RCA to my XMC-1. Cheers, Chad I'll put money on the ERC-4 sounding better via analog out vs the AES with the XMC-1, but its a different story with the G3 processors IMO. Belden 1800F used by Blue Jeans Cable is an inexpensive AES/EBU cable, for a slightly larger investment DH Labs makes a really nice AES cable too which I prefer. I understand Mogami also makes a nice AES cable but haven't tried that one. If you do get around to comparing the RCA to XLR outputs from your ERC-4 you'll probably find the RCA output to be about 4 to 4.5 db louder. I use the Irrational but Efficacious CD by Ayre Acoustics to match volume, works very well with SPL apps for my iPhone. Good luck and let us know what you find. |
|
|
|
Post by routlaw on Dec 10, 2021 12:53:40 GMT -5
Indeed the balance analog connections offer less volume than the analog RCA’s. So we have at least two random ERC-4 players now with the same results, RCA being louder than the XLR and thus the XLR outputs are NOT providing true balanced output. They couldn't in this case. Thanks for sharing. |
|
|
|
Post by AudioHTIT on Dec 10, 2021 13:24:56 GMT -5
Indeed the balance analog connections offer less volume than the analog RCA’s. So we have at least two random ERC-4 players now with the same results, RCA being louder than the XLR and thus the XLR outputs are NOT providing true balanced output. They couldn't in this case. Thanks for sharing. But what about putting on a test disk and measuring the output of the ERC-4 (XLR & RCA), eliminating any other components. |
|
|
|
Post by routlaw on Dec 10, 2021 13:32:05 GMT -5
But what about putting on a test disk and measuring the output of the ERC-4 (XLR & RCA), eliminating any other components. Have had a very busy week and not had time to do this, let alone the confident knowledge. ttocs provided a tutorial and if you are an EE all well and good but it also was more intent on dbu measurements rather than just the voltage output which is really all that is needed. In any case as it now stands I'm willing to bet dollars to donuts these ERC-4 players are only outputting about 1.4 vRMS on the XLR's. Thanks |
|
cawgijoe
Emo VIPs 
"When you come to a fork in the road, take it." - Yogi Berra
Posts: 5,033
|
Post by cawgijoe on Dec 10, 2021 13:35:39 GMT -5
Indeed the balance analog connections offer less volume than the analog RCA’s. So we have at least two random ERC-4 players now with the same results, RCA being louder than the XLR and thus the XLR outputs are NOT providing true balanced output. They couldn't in this case. Thanks for sharing. Have you asked Emotiva about this? Maybe Keith can chime in. |
|
|
|
Post by routlaw on Dec 10, 2021 14:04:33 GMT -5
Have you asked Emotiva about this? Maybe Keith can chime in. Not yet, but will. Honestly I am surprised they have not had something to say about this already… or maybe they are too busy to read the forum lately. Who knows. |
|
ttocs
Global Moderator 
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,154
|
Post by ttocs on Dec 10, 2021 14:20:51 GMT -5
But what about putting on a test disk and measuring the output of the ERC-4 (XLR & RCA), eliminating any other components. Have had a very busy week and not had time to do this, let alone the confident knowledge. ttocs provided a tutorial and if you are an EE all well and good but it also was more intent on dbu measurements rather than just the voltage output which is really all that is needed. In any case as it now stands I'm willing to bet dollars to donuts these ERC-4 players are only outputting about 1.4 vRMS on the XLR's. Thanks Something to consider in all of this is what happens on the preamp end of the interconnect. I understand you're getting lower SPL from a speaker when ERC-4 is connected via the XLR output vs the RCA, but I don't think one may conclude that the output voltage of the ERC-4 is different than stated by Emotiva. I've read elsewhere that other products with XLR Balanced and RCA outputs result in a louder volume from the RCA, but, this doesn't mean the voltage from these source outputs is upside down because it's the preamp that takes the signal and does "something" with it before spitting it out again. Last night I began to setup a test for my OPPO. I'll test the SPL of: SPL of Tone Generator -> OPPO XLR & RCA -> XMC-2 Reference Stereo -> Amp/Speaker, VS VOLTAGE of Tone Generator -> OPPO -> XLR & RCA Outputs. I hope to do this tonight or tomorrow. Another fun experiment! |
|
|
|
Post by routlaw on Dec 10, 2021 17:43:29 GMT -5
Something to consider in all of this is what happens on the preamp end of the interconnect. I understand you're getting lower SPL from a speaker when ERC-4 is connected via the XLR output vs the RCA, but I don't think one may conclude that the output voltage of the ERC-4 is different than stated by Emotiva. I've read elsewhere that other products with XLR Balanced and RCA outputs result in a louder volume from the RCA, but, this doesn't mean the voltage from these source outputs is upside down because it's the preamp that takes the signal and does "something" with it before spitting it out again. Last night I began to setup a test for my OPPO. I'll test the SPL of: SPL of Tone Generator -> OPPO XLR & RCA -> XMC-2 Reference Stereo -> Amp/Speaker, VS VOLTAGE of Tone Generator -> OPPO -> XLR & RCA Outputs. I hope to do this tonight or tomorrow. Another fun experiment! Thanks for the continued input, sounds as though you have some knowledge with this sort of thing. Thus far I have not had any success obtaining a meaningful voltage measurement from my player unfortunately. So still in the dark about its absolute output gain for now. However I am not so sure I am in agreement with your theory above, unless of course a given pre-amp only had XLR inputs but was not a true balanced design meaning it would not have the ability to sum the + and - signals. All of my gear as stated by Emotiva is full differential balanced designs, so at least in this case the RMC should not be fooling around with and reducing a balanced inputs otherwise what on earth would be the point of buying into such devices and technology. All that out of the way I look forward to reading about your experiments above with your Oppo player. |
|
|
|
Post by routlaw on Dec 10, 2021 20:50:32 GMT -5
Later this afternoon I was able to obtain what I felt was a decent and accurate voltage output measurements on both the RCA and XLR outputs. As expected the RCA came in a just a tad over 2 vRMS, while the XLR barely reached over 1 vRMS, doubled that would equate to also 2 vRMS. Additionally measuring Pins 2&3 together did NOT sum to zero or near zero as it should have been.
|
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,154
|
Post by ttocs on Dec 11, 2021 3:00:45 GMT -5
I played a 1kHz tone on a Mac mini using REW and sent the signal to an OPPO 105 via HDMI. The same can be done via Optical to the ERC-4. The first test has the OPPO connected to the XMC-2 using XLR and RCA cables into Balanced In and Analog 3, and the tone plays into a speaker. A second computer recorded the tone with a mic and REW's SPL Meter and RTA. You can see which XMC-2 Input is being used in each image as well as the audio mode which is Reference Stereo. The XLR Balanced Input of the XMC-2 played louder by 1.2dB. This only shows what happens when the OPPO is sent through the XMC-2 using Reference Stereo and measuring a speaker with a mic.   Then the interconnect cables were disconnected from the XMC-2 and oscilloscope probes were connected to each one to measure the voltage from the OPPO's XLR and RCA outputs at the same time, but only measuring one pin of the XLR (so half the total voltage). I moved the traces so each can be seen clearly, but just look below to see the actual voltage of each one. The voltage is right at about 2V for each one, but for the record, the Teal trace is the RCA, Purple is the XLR. This clearly shows output voltage from each output type.  This doesn't tell us what the ERC-4 is outputting, but it's a comparison of both types of outputs on one device so it's something to possibly compare against the ERC-4 operationally. |
|
|
|
Post by routlaw on Dec 11, 2021 11:34:28 GMT -5
Very interesting experiment you have going on here ttocs, thanks for doing this and sharing.
I don't have the level of equipment you have to pull off the same thing, but yesterday with my test I finally located an old test disc with a 1 kHz test tone, at -20, -10, and 0 db and used this in my ERC-4 with repeat play. Given my more crude method I didn't have an extra pair of hands to write the various measurements down but it was clear to me that my ERC-4 in balanced mode measured about half of what it should at 1 volt ± with 0 db, and just as important pins 2&3 when measured together did not sum out to zero. I have no idea what's going on with these players but they obviously do NOT have a balanced output mode. I'll gladly eat a diet of crow and humble pie if and when proven wrong and apologize for creating a ruckus about it.
Perhaps worth noting one other measurement I did which reflects one of your test was with my Holo Spring DAC which has both XLR and RCA outputs with specs of 5.8 vRMS and 2.9 vRMS respectively. While running SPL measurements in my room the XLR outputs (Holo) into the same XLR inputs of the RMC used previously rendered an approximate 1.5 db louder output than the RCA outputs but at least it wasn't 4 to 4.5 db lower as with the ERC-4. One would think the XLR output would be 6 db louder in both scenarios (ERC-4 and Holo) but obviously not, yet another mystery.
In the meantime I have sent a note to Emotiva tech support about this but don't expect to hear back anytime soon given the holiday season. Did more listening with the RCA outputs yesterday again and am inclined to think the player sounds better this way vs the XLR outputs, but should do a few more comparisons.
|
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,154
|
Post by ttocs on Dec 11, 2021 13:09:24 GMT -5
DO NOT DO THIS!I know that the OPPO 105 can competently be connected directly to an amp and be used as a "preamp" with a volume control, and my OPPO was at a very low volume. Not very many products like this can be connected to an amp, so, DO NOT DO THIS !!!!! The amp used is a Krell with both XLR and RCA inputs. This is where you can find the 6dB difference of XLR vs RCA outputs, or in my case, 6.3dB. The first image is XLR, the second is RCA. To measure a tone like this requires a mic on a stand, otherwise if it moves a fraction of an inch the SPL will vary.   This is why I am not concerned with getting only 1.2dB variance from XLR/RCA when running a source signal into and out of a processor. We don't know how the processor inputs are chained to the rest of the circuitry. What matters to me is the actual voltage coming out of a source, in your case, a ERC-4, while it's outputs are not connected to anything other than a meter or scope.
|
|
|
|
Post by routlaw on Dec 11, 2021 13:29:18 GMT -5
What matters to me is the actual voltage coming out of a source, in your case, a ERC-4, while it's outputs are not connected to anything other than a meter or scope. Agreed and is exactly how I measured yesterday. Thanks as always. |
|
|
|
Post by garbulky on Dec 11, 2021 13:38:36 GMT -5
Later this afternoon I was able to obtain what I felt was a decent and accurate voltage output measurements on both the RCA and XLR outputs. As expected the RCA came in a just a tad over 2 vRMS, while the XLR barely reached over 1 vRMS, doubled that would equate to also 2 vRMS. Additionally measuring Pins 2&3 together did NOT sum to zero or near zero as it should have been. Oh boy that's really dissapointing. Looking forward to what Emotiva says but this is not a good sign. |
|
|
|
Post by routlaw on Dec 11, 2021 17:44:13 GMT -5
Later this afternoon I was able to obtain what I felt was a decent and accurate voltage output measurements on both the RCA and XLR outputs. As expected the RCA came in a just a tad over 2 vRMS, while the XLR barely reached over 1 vRMS, doubled that would equate to also 2 vRMS. Additionally measuring Pins 2&3 together did NOT sum to zero or near zero as it should have been. Oh boy that's really dissapointing. Looking forward to what Emotiva says but this is not a good sign. Well it is indeed, but even with this setback truth is the ERC-4 really sounds very good, excellent in fact. One has to wonder though how much the audio quality would improve if the Balanced XLR outputs were performing as they should. Regardless even with the RCA outputs it still sounds awesome. Thanks |
|
|
|
Post by tsruha on Mar 30, 2022 19:38:09 GMT -5
I agree. I am very disappointed with the ERC-4 because of the crappy balanced outputs. It’s unacceptable and irks me enough that I’m close to calling the ERC-4 snake oil. Just for the heck of it last evening upon discovering the considerable difference in the output from the RCA vs Balanced outs I thought a listening session comparing the two might be worthwhile. Again starting out with BJC LC-1 interconnects on the RCA and Mogami 2534 for the balanced outs the difference in sound quality really surprised given how different the two were. More like comparing two different CD players rather than cable/output combinations. As expected based upon my sound level measurements the RCA outputs played louder than the balanced outputs and it seemed more like 4.5 db instead of the measured 4 db earlier in the afternoon. This might be attributed to what I perceived as considerably more profound bass output, not necessarily better just abundant compared to the balanced outs with Mogami. The Mogami/XLR bass was more focused and precise and perhaps a bit more tuneful. After a couple of hours of back and forth listening I gave up on which performed the best, yet still scratching my head on how different they presented the same music. This was not subtle at all. So much for "expectation bias" and at least in this case you could throw that neurosis out the window. The only thing concrete clear was the RCA outputs played louder by some 4 to 4.5 db compared to the balanced outputs when in fact the balanced outputs should have provided a bare minimum of 6 db more output over the RCA. Emotiva, you have a problem. |
|
cawgijoe
Emo VIPs
"When you come to a fork in the road, take it." - Yogi Berra
Posts: 5,033
|
Post by cawgijoe on Mar 31, 2022 7:46:35 GMT -5
I had asked if Emotiva has been contacted to look into this or comment and I don't see a response.
Has anyone officially reached out with their findings to Lonnie/Keith?
|
|
KeithL
Administrator
Posts: 10,261
|
Post by KeithL on Mar 31, 2022 10:45:09 GMT -5
I hadn't noticed this thread before... but now I'm going to respond to it....  The balanced outputs on the ERC-4 come from the balanced outputs of the DAC chip, go through some really nice buffers, and are sent to the balanced outputs on the ERC-4... The unbalanced RCA output is actually derived by a separate differential amplifier connected to both of those outputs... This is the "good" way of doing that...
And, since the gain on that differential amplifier happens to be set to 1.0, the output level on the unbalanced output happens to be double the output on each leg of the balanced output. (That is what you get when you sum the two halves of a balanced differential signal and don't divide the result in half afterwards.)
This is NOT "a problem"... and it is NOT "an issue"... it is simply an arbitrary decision made by the engineer who designed that particular circuit. (Either of those levels could be doubled or halved by changing a pair of fifty cent parts... and doing so would not change the performance or the sound quality of the ERC-4 in any substantial way.)
Some people think that it is "proper" for the level of the unbalanced output to match the sum of the two sides of the balanced output... And some people think that it should match each side when taken separately... And some people prefer for their balanced outputs to be "pro level" and so to be 12 dB higher... Although, after a few beers, some would even go so far as to argue whether "consumer" gear should use "pro" levels or not...
The point is that it doesn't really matter and has no effect at all on the way the unit sounds...
And, although I haven't actually measured it, I would bet that the two legs of the balanced output match within a percent or two. And, with modern gear, there is no specific requirement for them to match more perfectly than that, or to null perfectly, and no significant benefit if they do...
(We're quite pleased with both how the ERC-4 performs and how it actually sounds...  ) I had asked if Emotiva has been contacted to look into this or comment and I don't see a response. Has anyone officially reached out with their findings to Lonnie/Keith? |
|
|
|
Post by 405x5 on Mar 31, 2022 11:02:21 GMT -5
I hadn't noticed this thread before... but now I'm going to respond to it.... The balanced outputs on the ERC-4 come from the balanced outputs of the DAC chip, go through some really nice buffers, and are sent to the balanced outputs on the ERC-4... The unbalanced RCA output is actually derived by a separate differential amplifier connected to both of those outputs... This is the "good" way of doing that...
And, since the gain on that differential amplifier happens to be set to 1.0, the output level on the unbalanced output happens to be double the output on each leg of the balanced output. (That is what you get when you sum the two halves of a balanced differential signal and don't divide the result in half afterwards.)
This is NOT "a problem"... and it is NOT "an issue"... it is simply an arbitrary decision made by the engineer who designed that particular circuit. (Either of those levels could be doubled or halved by changing a pair of fifty cent parts... and doing so would not change the performance or the sound quality of the ERC-4 in any substantial way.)
Some people think that it is "proper" for the level of the unbalanced output to match the sum of the two sides of the balanced output... And some people think that it should match each side when taken separately... And some people prefer for their balanced outputs to be "pro level" and so to be 12 dB higher... Although, after a few beers, some would even go so far as to argue whether "consumer" gear should use "pro" levels or not...
The point is that it doesn't really matter and has no effect at all on the way the unit sounds...
And, although I haven't actually measured it, I would bet that the two legs of the balanced output match within a percent or two. And, with modern gear, there is no specific requirement for them to match more perfectly than that, or to null perfectly, and no significant benefit if they do...
(We're quite pleased with both how the ERC-4 performs and how it actually sounds... ) I had asked if Emotiva has been contacted to look into this or comment and I don't see a response. Has anyone officially reached out with their findings to Lonnie/Keith? What he said…. and if I could add: The loudspeaker The amplification The processor/ AV processor The room, volume and acoustical environment. What’s kicked around here in this thread is at the bottom of the list, or better yet pulled from it all together in terms of meaningful and useful differences regarding audibility |
|
|
|
Post by mgbpuff on Mar 31, 2022 18:45:23 GMT -5
Hmm - Seems a couple of rascals need to apologize to Emotiva!
|
|
)
