• 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!

Apogee Element 24 thunderbolt interface: quick measurements

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
I was curious about this device for a while, and jumped when a used one came up for about half the list price. This is a thunderbolt / Mac-only pro-level device with all the controls from software. An iOS app is available at no cost for basic remote control. I was actually more curious about the ADC performance, but tested both, DAC and ADC.

Tests were run on iMac using REW as signal generator and measurement software. The unit was used as DAC and ADC simultaneously for these tests. Line out of the device was fed into microphone 1 input. In effect, the plots below are a combination of DAC and ADC performance, combined.

1kHz distortion plot looks good. No mains noise visible even at -150dB. Power was supplied by a provided 12v DC adapter. Balanced output was set to just above 4v, as the software provides control in 1dB increments. Sampling rate was 96kHz throughout (limited by REW on macOS, otherwise Apogee works up to 192kHz).
THD2.png

THD of -114.5dB, THD+N of nearly -107dB. There are also no major surprises all the way up to 48kHz.

I also measured the headphone output in the same loop-back mode, and the performance was nearly identical to the above, with just about 1-2dB worse THD+N result.

Jitter using 12kHz Jtest signal at just below 0dBFS looks pretty good:
12k.png

And a multiple-tone result:
multitone.png

Although there were some higher level artifacts between the tones below 100Hz, they were at a low level of -130dB or below. The 'forest' in between the tones also increased with frequency, but still remained well below -130dB all the way to 20kHz.

[EDIT] Adding a 48kHz frequency sweep and phase response:

sweep.png


As a subjective test, I listened to the headphone output using HE560s. Sound was clean, no noise or hum at all, at any volume setting. Element24 drives the headphones perfectly, as loud as I was ever willing to go. There was much more room left to increase the volume, if I had any desire to destroy my eardrums. Sound is as good as I've heard from any DAC, and it seems that the THX 789 AAA isn't needed, as it improved nothing that the little Apogee couldn't do all by itself.

I think I'm going to like this little unit! ;)
 
Last edited:

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
I was curious about this device for a while, and jumped when a used one came up for about half the list price. This is a thunderbolt / Mac-only pro-level device with all the controls from software. An iOS app is available at no cost for basic remote control. I was actually more curious about the ADC performance, but tested both, DAC and ADC.

Tests were run on iMac using REW as signal generator and measurement software. The unit was used as DAC and ADC simultaneously for these tests. Line out of the device was fed into microphone 1 input. In effect, the plots below are a combination of DAC and ADC performance, combined.

1kHz distortion plot looks good. No mains noise visible even at -150dB. Power was supplied by a provided 12v DC adapter. Balanced output was set to just above 4v, as the software provides control in 1dB increments. Sampling rate was 96kHz throughout (limited by REW on macOS, otherwise Apogee works up to 192kHz).
View attachment 26036
THD of -114.5dB, THD+N of nearly -107dB. There are also no major surprises all the way up to 48kHz.

I also measured the headphone output in the same loop-back mode, and the performance was nearly identical to the above, with just about 1-2dB worse THD+N result.

Jitter using 12kHz Jtest signal at just below 0dBFS looks pretty good:
View attachment 26037
And a multiple-tone result:
View attachment 26038
Although there were some higher level artifacts between the tones below 100Hz, they were at a low level of -130dB or below. The 'forest' in between the tones also increased with frequency, but still remained well below -130dB all the way to 20kHz.

As a subjective test, I listened to the headphone output using HE560s. Sound was clean, no noise or hum at all, at any volume setting. Element24 drives the headphones perfectly, as loud as I was ever willing to go. There was much more room left to increase the volume, if I had any desire to destroy my eardrums. Sound is as good as I've heard from any DAC, and it seems that the THX 789 AAA isn't needed, as it improved nothing that the little Apogee couldn't do all by itself.

I think I'm going to like this little unit! ;)
That's really great performance! What kind of dac chip does it use?
 

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
410
Likes
472
Wow, these can be had used from low $400's. Wall wart and the performance is still amazing.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,747
Likes
37,569
Thank you for posting this........I think? These Elements have had my eye since they were released. I like the idea of minimizing the cost of the box so more can be put on what is inside. Plus Apogee usually makes good stuff, and that appears to be the case right here.

I almost purchased one a couple times, but really don't need it. Plus I dislike that it works only with Mac even though I have one Mac. It looks like if you need such a thing in this below $1000 range this is the one to beat if you are in the Mac universe. Probably this or one of the Motus.

But I must complain, you didn't do anything to dissuade me from wanting one in the future.

BTW, using a basic good music track what kind of Difference RMS do you get with Deltawave on this in loopback.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
Thank you for posting this........I think? These Elements have had my eye since they were released. I like the idea of minimizing the cost of the box so more can be put on what is inside. Plus Apogee usually makes good stuff, and that appears to be the case right here.

I almost purchased one a couple times, but really don't need it. Plus I dislike that it works only with Mac even though I have one Mac. It looks like if you need such a thing in this below $1000 range this is the one to beat if you are in the Mac universe. Probably this or one of the Motus.

Mac-only support made me think twice. But I do like the remote control iOS app. And since I'm not going to use this as a stand-alone interface anyway, having all the control in software made sense to me. This thing doesn't even have a physical power switch!

There are larger versions of this available with more inputs (Element 46 and Element 88), but more expensive.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
BTW, using a basic good music track what kind of Difference RMS do you get with Deltawave on this in loopback.

I'll let you know when I have a chance to test it :)

Just added a frequency sweep from REW. I'm afraid that also doesn't help to dissuade you...
 

JohnPM

Senior Member
Technical Expert
Joined
Apr 9, 2018
Messages
344
Likes
919
Location
UK
Sampling rate was 96kHz throughout (limited by REW on macOS, otherwise Apogee works up to 192kHz).
I've done a bit of work to improve behaviour at 192 kHz using the Java drivers, the next beta build will have 192 kHz as a sample rate option. Device needs to be configured for that in Audio Midi Setup as well of course.

If you increase the right hand window length it will smooth out the slight ripple at the low end of the sweep response.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
I've done a bit of work to improve behaviour at 192 kHz using the Java drivers, the next beta build will have 192 kHz as a sample rate option. Device needs to be configured for that in Audio Midi Setup as well of course.

If you increase the right hand window length it will smooth out the slight ripple at the low end of the sweep response.

That’s great news, John!
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
BTW, using a basic good music track what kind of Difference RMS do you get with Deltawave on this in loopback.


Not bad, but not great, either :) dBA versions are better, so it seems the greater differences are outside the more 'audible' range. This is using the same track that is used in the Gearslutz DAC/ADC loop evaluation thread for all their measurements. I captured the looped output at 44.1kHz and let DeltaWave compensate frequency (EQ) to account for the effect of DAC/ADC filters on the comparison.

This is generation 1 compared to the original digital file:
1557680824795.png
 
Last edited:
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
I'm not sure how to measure crosstalk, other than feeding a near 0dBFS signal into one channel and capturing the output of the other. Here it is. 1kHz, -0.8dbFS signal fed into channel 1, captured from output 2.

1kHz signal is about -145dB down, while the second harmonic at 2kHz is at about -133dB and seems to dominate in the undriven channel. Is there a better the way to measure crosstalk?

crosstalk.png
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,747
Likes
37,569
I'm not sure how to measure crosstalk, other than feeding a near 0dBFS signal into one channel and capturing the output of the other. Here it is. 1kHz, -0.8dbFS signal fed into channel 1, captured from output 2.

1kHz signal is about -145dB down, while the second harmonic at 2kHz is at about -133dB and seems to dominate in the undriven channel. Is there a better the way to measure crosstalk?

View attachment 26098
Only way to measure cross talk. Well I actually will do a sweep and look at the other channel. But the same thing. Put signal in one channel and see what leaks into the other.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
Only way to measure cross talk. Well I actually will do a sweep and look at the other channel. But the same thing. Put signal in one channel and see what leaks into the other.

I'll try a sweep as well. I tried a multi-tone signal, but that produced nothing at all in the second channel except low-level noise, so I'm not sure if the interface simply turned off the channel sensing that there was no input.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,747
Likes
37,569
Not bad, but not great, either :) dBA versions are better, so it seems the greater differences are outside the more 'audible' range. This is using the same track that is used in the Gearslutz DAC/ADC loop evaluation thread for all their measurements. I captured the looped output at 44.1kHz and let DeltaWave compensate frequency (EQ) to account for the effect of DAC/ADC filters on the comparison.

This is generation 1 compared to the original digital file:
View attachment 26095
Thanks for doing this.

I expected better results. I do notice you required 10 db of gain adjustment. Could different output and input gain settings reduce that?

The Zen Tour I have doing a loopback with that same file comes in around -56 db without EQ. It picks up a couple with EQ. If I instead drop EQ and chop off the lower 20 hz it jumps to -66 db. So I would have expected the Element to get at least -50 db and hoped for better.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,747
Likes
37,569
I'll try a sweep as well. I tried a multi-tone signal, but that produced nothing at all in the second channel except low-level noise, so I'm not sure if the interface simply turned off the channel sensing that there was no input.
Probably the individual levels of each tone were low enough the crosstalk fell below the noise floor in the other channel.
If you need to you can always stick a noise signal at - 140 dbFS to prevent it turning off. In an FFT that will drop well below the real noise of the device.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,699
Likes
10,386
Location
North-East
Thanks for doing this.

I expected better results. I do notice you required 10 db of gain adjustment. Could different output and input gain settings reduce that?

There's a -10dB input setting on Element that raises the level (also a +4dB that lowers it). For some reason there were a few clipped samples (intersample over?) in the recording when -10db setting was engaged, so I turned it off to see if this would help. The null value was nearly identical with and without the -10dB setting. I'll play around with the other settings to see if I can get a better result.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,747
Likes
37,569
There's a -10dB input setting on Element that raises the level (also a +4dB that lowers it). For some reason there were a few clipped samples (intersample over?) in the recording when -10db setting was engaged, so I turned it off to see if this would help. The null value was nearly identical with and without the -10dB setting. I'll play around with the other settings to see if I can get a better result.
Yes that file has some intersample overs. Going from memory you will have to leave about a decibel at least. When I got within less than a db there were overs.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,300
Location
China
Please test with imd 20hz+1000hz, multitone NID pink spectrum 20hz-10000hz.
 
Top Bottom