• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

E1DA Cosmos ADCiso Review

Rate this ADC:

  • 1. Poor (headless panther)

    Votes: 3 2.3%
  • 2. Not terrible (postman panther)

    Votes: 1 0.8%
  • 3. Fine (happy panther)

    Votes: 13 9.8%
  • 4. Great (golfing panther)

    Votes: 115 87.1%

  • Total voters
    132
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,770
Likes
242,419
Location
Seattle Area
What do you have in mind?
Need to have iterative sweep support to get things like this:

index.php
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,434
Likes
2,820
Location
South of China, SHZ area, - Слава Україні
I do all things by myself, design, calibration, assembling, and online support in the discord. When the user mixes up scaler vs scalar, I see that he's not an engineer or DIYer but just a manager who accidentally got my stuff and has no idea what is it. I have no time to explain basic things, so I decided I'd sell my stuff to smart people only.
BTW, Cosmos ADC grade 0 shows me THD+N -125db.. Cosmos APU -131db but the theoretical limit is -135db, hence, a smart person has not only 200x times cheaper device vs AP but in some aspects a lot more precise.
PS: I recollect another example of a similar approach i.e. the product is wonderful but not easy to use without spending some brain effort and forum reading. REW - typical DIYer2DIYers SW! I was shocked at how fast John reacted to my complaint about REW calculating the THD+N incorrectly, John instantly found the solution, maybe within 24 hours if my memory serves well. I don't think that John has a huge money from his REW, just he can improve the product and doesn't wait for another reason to start his work. When somebody asks me to mod FW for them or so, I recollect that good example and push myself to do help rapidly. We are from the same planet, I guess ;)
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,770
Likes
242,419
Location
Seattle Area
I do all things by myself, design, calibration, assembling, and online support in the discord.
So? How is that our problem?
When the user mixes up scaler vs scalar, I see that he's not an engineer or DIYer but just a manager who accidentally got my stuff and has no idea what is it. I have no time to explain basic things, so I decided I'd sell my stuff to smart people only.
I don't know who the heck you are talking about. If it is me, teaches me to not make an exception again in testing your stuff. You badly need to keep your arrogance in check.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,770
Likes
242,419
Location
Seattle Area
PS: I recollect another example of a similar approach i.e. the product is wonderful but not easy to use without spending some brain effort and forum reading. REW - typical DIYer2DIYers SW! I was shocked at how fast John reacted to my complaint about REW calculating the THD+N incorrectly, John instantly found the solution, maybe within 24 hours if my memory serves well. I don't think that John has a huge money from his REW, just he can improve the product and doesn't wait for another reason to start his work. When somebody asks me to mod FW for them or so, I recollect that good example and push myself to do help rapidly. We are from the same planet, I guess ;)
John is an incredibly nice and generous person. He is not at all like you. If you were 1% as nice and appreciative as him, it would be a massive improvement.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,434
Likes
2,820
Location
South of China, SHZ area, - Слава Україні
@sarieri, was another guy who I believe, accidentally bought my stuff and he called Scaler as scalar. I taught this guy about how to measure tiny distortions, one or two months in our discord. Now he can teach others if they wish to learn.
PS: diyaudio.com folk told me about another interesting ADC https://nihtila.com/2020/02/02/adc-ak5572-very-high-performance-audio-ad-converter/
That device has only one sensitivity 3.6Vrms 0dbfs, hence, with "standard 4V" test it will simply fall into a hard clipping, THD+N 10% or so.
 
Last edited:

capslock

Senior Member
Joined
Jun 19, 2020
Messages
354
Likes
160
Amir, what kind of signal generator did you use to test the AP's ADC in high performance mode? To the best of my knowledge, the AP's analog signal generator does not have HHD < 160 dB, nor do the digital channels.

Do the generators used for the IMD tests change when subjected to a low impedance load? There was some discussion about the ESS hump in Cosmos ADC either here or on diyaudio. The ADC was driven from another ESS DAC. It turned out the hump decreased substantially when a Cosmos Scaler was added. The interpretation, AFIR, was that Cosmos ADC was loading the DAC so much it brought out the hump in the DAC.
Amir, may I humbly repeat my question: what generator did you to test the high performance mode of your AP?
 

VintageFlanker

Master Contributor
Forum Donor
Joined
Sep 20, 2018
Messages
5,004
Likes
20,137
Location
Paris
When the user mixes up scaler vs scalar, I see that he's not an engineer or DIYer but just a manager who accidentally got my stuff and has no idea what is it. I have no time to explain basic things, so I decided I'd sell my stuff to smart people only.
This...
1000031135.gif

...This type of mindset & speech is absolutely unsuitable for such a (great) place as ASR.

The very purpose of ASR is all about highly competent people wanting to share their technical knowledge and educate others. I'm sure that's what @amirm had in mind when he created this forum.

Whatever how skilled you are (and your are), how performant your products are (and they are), you are definitely lacking in that regard, Ivan.

VF, out.
 

morillon

Major Contributor
Joined
Apr 19, 2022
Messages
1,391
Likes
281
4v.... and mono card possible etc
 

pavuol

Major Contributor
Forum Donor
Joined
Oct 2, 2019
Messages
1,587
Likes
3,977
Location
EU next to warzone :.(
...This type of mindset & speech is absolutely unsuitable for such a (great) place as ASR.

I have no time to explain basic things, so I decided I'd sell my stuff to smart people only.
IVX: Yass but what is "smart people"? We're here for numbers so I think on every [e1da] product page there should be exact IQ value of potential customer specified so to not cause purchase regrets. We had monetary elitism in high-end audio, so intellectual elitism is great approach to make objective "price independent" audio more inclusive..
he-he [i laugh stupidly ;) I'm "out of range" simple bloke..]
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,796
Likes
39,222
Location
Gold Coast, Queensland, Australia
The high performance analyzer turns on a secondary ADC which just measures the main tone and then subtracts it out from the signal. That way, the primary ADC is just measuring the noise/harmonic distortion which is at far too low level of level to cause any non-linearities. As a result, you basically have no distortion in the analyzer as you can see in the FFT.

The distortion of the AP may be lower, but its THD+N is limited by the noise, not the distortion, whereas the Cosmos is limited by its distortion, not the noise.

Cosmos is an incredible achievement for very little money from one man. If you can be bothered with the messing around, calibrating levels, considering the input impedance for each range and then using an external high quality D/A converter and random free (or paid) software. Add to that, the need to buffer/precondition and attenuate externally and it's certainly not a turn-key solution.

I have a Quantasylum QA-403 which solves all those issues, including having inbuilt range switching for the generator and the analyzer, a standard 100k/200k input impedance as well as having the A/D and D/A synchronized on the same clock. It is also now fully isolated. Sure, having a 200k front end means I throw away 10dB of THD+N, but that doesn't matter for what I do and when the numbers are already down to >-120dB loopback THD, and +N numbers better than -110dB it's perfect for me as a small, mobile analyzer that just works.

It will not compete with an AP and neither will the Cosmos setup.

The QA40x Software is continuously evolving and Matt from Quantasylum is very much on the ball with minor bugfixes/features and add-ons. Just since September 2023, there has been 8 (eight) complete version updates to fix/improve things. It will run on any machine- I even ran it on a 10+ year old Intel Celeron NUC and a cheap Lenovo Pentium with no problems.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,796
Likes
39,222
Location
Gold Coast, Queensland, Australia
This...
View attachment 366084
...This type of mindset & speech is absolutely unsuitable for such a (great) place as ASR.

The very purpose of ASR is all about highly competent people wanting to share their technical knowledge and educate others. I'm sure that's what @amirm had in mind when he created this forum.

Whatever how skilled you are (and your are), how performant your products are (and they are), you are definitely lacking in that regard, Ivan.

VF, out.

I think you are likely mis-interpreting what he said. It wasn't directed at anyone in particular, just clowns in general who don't know what they are doing and shouldn't be buying the devices in the first place.

And it will be the hobbyists and serious DIYers who will lose out if and when Ivan decides it's all just too hard dealing with the general public and their skyhigh high expectations and entitled attitudes.
 

morillon

Major Contributor
Joined
Apr 19, 2022
Messages
1,391
Likes
281
The distortion of the AP may be lower, but its THD+N is limited by the noise, not the distortion, whereas the Cosmos is limited by its distortion, not the noise.

Cosmos is an incredible achievement for very little money from one man. If you can be bothered with the messing around, calibrating levels, considering the input impedance for each range and then using an external high quality D/A converter and random free (or paid) software. Add to that, the need to buffer/precondition and attenuate externally and it's certainly not a turn-key solution.

I have a Quantasylum QA-403 which solves all those issues, including having inbuilt range switching for the generator and the analyzer, a standard 100k/200k input impedance as well as having the A/D and D/A synchronized on the same clock. It is also now fully isolated. Sure, having a 200k front end means I throw away 10dB of THD+N, but that doesn't matter for what I do and when the numbers are already down to >-120dB loopback THD, and +N numbers better than -110dB it's perfect for me as a small, mobile analyzer that just works.

It will not compete with an AP and neither will the Cosmos setup.

The QA40x Software is continuously evolving and Matt from Quantasylum is very much on the ball with minor bugfixes/features and add-ons. Just since September 2023, there has been 8 (eight) complete version updates to fix/improve things. It will run on any machine- I even ran it on a 10+ year old Intel Celeron NUC and a cheap Lenovo Pen ;-)tium with no problems.
is it possible to use other software like mt from pkane, virtins , or rew , "asio" , with quantasylum? or only quantasylum software?
;-)
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,770
Likes
242,419
Location
Seattle Area
Amir, may I humbly repeat my question: what generator did you to test the high performance mode of your AP?
I answered already. It is the high-performance analog generator in APx555. You will never get such low distortion if you use the DAC generator.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,770
Likes
242,419
Location
Seattle Area
It wasn't directed at anyone in particular,
Oh it was. He is talking about me. Same thing he did when I tested his last product. Both times I gave his product the best recommendation I can give and yet, he spits out this kind of poison. I have a short list of companies that I want to have nothing to do with and he was on it. But I decided to give it a second chance only to see a repeat. Like the other people on that list, common sense is not something that was taught to him. Or respect for where he is.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,796
Likes
39,222
Location
Gold Coast, Queensland, Australia
is it possible to use other software like mt from pkane, virtins , or rew with quantasylum? or only quantasylum software?
;-)

I can't speak for him in that regard, but I do know Matt is working on progressing ASIO and it could be a while away. The issues of course are controlling the attenuators/gain stages for the inputs and outputs. Overall, I'm more than happy with the cadence of updates and improvements.

There are multiple ways to interact with the device as it stands. Someone could even build their own entire app if they were so inclined.
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,796
Likes
39,222
Location
Gold Coast, Queensland, Australia
Oh it was. He is talking about me.

Oh dear. Sorry to hear Amir. I always like to think language and it subtleties can make someone sound worse than they intend. IOW, we give the benefit of the doubt where we can, and then if that fails, tear them a new a#s. ;)
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,781
Likes
4,746
Location
Liège, Belgium
I think we see something very symptomatic here.

Ivan is producing unique products, by their quality and ambition.
They truely provide options and performance that were unreachable to typical DiYer or amateurs before them: ADC, APU, Scaler, not to forget his excellent phone load board.

Now, this comes at the cost of setup complexity and you need to have some geek-world affinity to really understand them and get the best of them.

At the same time, the ADC especially has a lot of attraction for "normal" people, who don't understand the complexity.

If Ivan wants to survive - earn enough money - he probably needs this wider audience.
Although his main income, I suspect, is coming from the excellent headphones dongles he produces.

So the paradox: I think we see Ivan frustrated somehow to have to explain the complex details that you need to understand to measure with his devices.
It's very time consuming and, at the same time, the same questions come over and over again.

And it seems he also always believes he has to prove his devices are the best. Which, given their price, is not needed: their price/performance ratio is unrivaled. Nobody is disputing that.
Proving your devices are the best is, of course, always challenged, and this, again, I suspect, gives him (unneeded) frustration


If Ivan was able to hire some marketing person that could also produce easy to read documentation for the 2 different profiles of persons he sales his devices to, I think his frustration would decrease and he would be able to focus on what he's doing best -the technical aspect- and earn more from his efforts.

OK, I don't want to seem condescending or giving lessons here.
Just to share my feelings about what I consider as a sadly missed opportunity, given Ivan's talent and genuine will to provide unique products for his customers vs the result he ultimately gets.
 
Last edited:

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,624
Likes
10,823
Location
Prague
Regarding low input impedance of the E1DA ADC that may be an issue for some measurements (preamps, especially tube preamps) can be easily overcome by inserting a high input impedance buffer in front of the E1DA ADC.

Example:

1. Topping DX5 RCA output directly to E1DA, input mono 1.7V /Topping has no issues with low impedance load, this is as a reference/
DX5 to E1DA (1.7V mono) direct.png


2. Same as (1), but the DIY input buffer with 100kohm input impedance and 50ohm output impedance is placed in front of the E1DA DAC. Please forgive some mains residuals due to long single ended wires.
E1DA (1.7V mono) with 100k input buffer R.png


This is still acceptable for many measuring tasks and allows to measure sources with output impedance up to 1kohm, at reasonably low cost, to a skilful DIYer. As I have said, E1DA is not for beginners, but it is an excellent value for technically oriented skilful users.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,829
Likes
6,357
Location
Berlin, Germany
When plugged on a Linux system, the E1DA Cosmos ADC shows 2 sliders in ALSAMIXER.
There is "Line" and "Line 1". Windows does not shows 2 sliders.
I have never met a documentation explaining what the second slider does.
I just checked and it took me two minutes to find what out it does. Both sliders work in the same way and are interchangeable. The thresholds the mode switching acts upon (sort of) an average of the two settings.
 
Top Bottom