• 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!

Review and Measurements of Chromecast Audio Analog Performance

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,395
Location
Seattle Area
This is a follow up to my review and measurements of Google Chromecast Audio's digital performance. As mentioned there, the Chromecast audio is a streaming device with integrated DAC. In the other review, I measured its Toslink digital output for usage with an external DAC. Here, I will be evaluating its analog output performance in case you want to forego the external DAC.

As noted there, the casting functionality of Chrome browser is poor and not bit-exact. So for this testing, I used Roon to play the test files and cast them to Chromecast Audio. This allows bit-perfect functionality so that we can see the performance of the hardware in best light.

The measurements are not extensive as with the digital output because my audio analyzer can't control the Chromecast Audio. But there is enough here to get a good idea of how well it does.

Measurements
As usual, let's feed the unit a 1 kHz tone at 44.1 kHz, 0 dBFS (full amplitude) and see how she does:

Google Chromecast Audio Analog Output Dashboard Measurement.png


We have near nominal output of 1.973 volts which is nice. This means there should be no problem driving your amplifiers to their maximum power.

Distortion is respectable resulting in SINAD (signal over power of distortion and noise) of 91 dB. Here is how that ranks among recently reviewed gear:

Google Chromecast Audio Analog Output SINAD Measurement.png


It is well short of even cheap desktop DACs like Topping D10 but not bad either.

Dynamic Range which is a test of playing (almost) nothing versus maximum signal follows the same suit:
Google Chromecast Audio Analog Output Dynamic Range SNR Measurement.png


Jitter and noise is also uneventful:
Google Chromecast Audio Analog Output Jitter Measurement.png


Yes, we have some jitter spikes around our main tone (symmetrical spikes) but because their levels are very low and so is their frequency, they will most likely be perceptually masked. There is some mains leakage coming from the switching power supply I am using to drive it but again, levels are pretty low at -110 dB and lower.

Conclusions
There appears to be nothing broken in the design of the Chromecast Audio DAC as long as you use the right software to drive it. While performance falls in the bottom third tier of tested audio products, it gets there without any drama or obvious faults.

For $35, you get a DAC and streaming functionality built it. As such, the Google Chromecast Audio easily gets my recommendation!

Seeing how you can improve its performance anytime with the use of a good DAC, it is hard to go wrong. You get to put your computer in a different place than your audio system and then stream what you want using Chromecast Audio. With a system like Roon where you can have multiple concurrent playback threads, this would make for an excellent secondary use/room.

-------------

As always, questions, comments, recommendations, etc. are welcome.

If you like this review, please consider donating using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though PayPal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
 

Umlautica

Member
Joined
Jul 25, 2018
Messages
18
Likes
40
I am using Roon for playback which sends the bits as is, so dynamic range compression is not in play.

Full dynamic range is a feature of the CCA that's disabled by default. It can be enabled in the Google Home app (screenshot attached).

I don't know where its power supply is. :) I am using another switching one.

If we're being picky, this has the potential to affect the results. Either way, thanks for confirming.
 

Attachments

  • 0.png
    0.png
    207.4 KB · Views: 1,279

folzag

Member
Joined
Jul 24, 2018
Messages
99
Likes
131
Apologies if this is a little too orthogonal to the topic at hand, but it seemed to fit in regard to a device with its primary claim to fame being cost.

Is there a simplistic mapping of SINAD to bit depth? For instance, given 16 bits, what is the minimum SINAD, above which the DAC can be considered "transparent"? How much SINAD is required for each additional bit?

And the (maybe) non-obvious follow-up: is there a simplistic mapping of SINAD to dynamic range? For instance, if I'm listening to a 24 bit, high-res song, albeit with DR8, what SINAD is really required for the DAC to be "transparent"?

If these questions are fundamentally wrong, please clarify, as that would mean I've missed understood something fairly important.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,395
Location
Seattle Area
Is there a simplistic mapping of SINAD to bit depth? For instance, given 16 bits, what is the minimum SINAD, above which the DAC can be considered "transparent"? How much SINAD is required for each additional bit?
Sure. Subtract 1.76 from SINAD and then divide by 6 to get effective number of bits:

1536896781324.png


And the (maybe) non-obvious follow-up: is there a simplistic mapping of SINAD to dynamic range? For instance, if I'm listening to a 24 bit, high-res song, albeit with DR8, what SINAD is really required for the DAC to be "transparent"?
No, you can't relate these two. The DR metrics are heuristically based values. In case of DR meter, it creates a histogram of loudness levels and picks some number of them, averages them and compares to second loudest peak (or something like that). In that regard, it is creating a single number for the entire track.

The DAC does not see music that way. It has to reproduce the quietest and loudest peaks individually. Regardless of the distribution of music loudness, we need to be able to reproduce the faintest signal faithfully regardless of how common that is in the track. As such, you can't use DR values to judge how much through dynamic range/SINAD you need.
 

folzag

Member
Joined
Jul 24, 2018
Messages
99
Likes
131
Thx. I had misunderstood DR. I thought it was the difference between the loudest sample and the quietest sample in a given song, with some extra fudge-factors to account for non-linearity of hearing perception.
 

bakker_be

Active Member
Joined
May 20, 2018
Messages
188
Likes
115
Location
Belgium
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,395
Location
Seattle Area
Adding on, there is no simple way to determine the lowest amplitude music signal. For one thing, there is digital mute sometimes at the beginning or the end of the track which would be a false metric to use. On the other hand, music could be below the noise level. And distinguishing noise from music can be hard.

The only method is a statistical one where you have to determine what is noise and what is music.

So in a nutshell, you can't determine the actual dynamic range of music.
 

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,032
Location
Tampa Bay
Wow I'm glad you did actually test this. I was just thinking about asking you to test this!

For $35 this is one hell of a useful unit! I'm glad to see that its performance is acceptable (especially at this price point).
 

folzag

Member
Joined
Jul 24, 2018
Messages
99
Likes
131

jpo

Member
Joined
Sep 29, 2018
Messages
12
Likes
5
Hi... first post here. I only signed up to talk about CCA. I think is a game changer and love to see reviews like this one... Thanks!

My opinion: after fiddling with it, and trying it through a Chord Mojo, and then bypass it to listen directly without finding noticeable differences with my 37 y/o ears. I have very good stereo setup (custom made, but i think there are about 10kusd just in components)... and i find its amazing that the analog chain of my setup is just the chromecast audio and tidal. Now i'm evaluating if tidal + roon or just tidal will be good enough.

Amir, you were about to measure that on the other CCA thread. Can you please update us with results? thanks for the good work and cheers from Chile!
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,395
Location
Seattle Area
Amir, you were about to measure that on the other CCA thread. Can you please update us with results? thanks for the good work and cheers from Chile!
Hi and welcome to the forum. Unfortunately I have so much new stuff to test that I have not gotten around to doing this.
 
  • Like
Reactions: jpo

veeceem

Active Member
Joined
Feb 18, 2018
Messages
284
Likes
158
big question: does it sound as good or the same or better than Schiit Yggdrasil? Based on the above measurements? @amirm
 

PierreV

Major Contributor
Forum Donor
Joined
Nov 6, 2018
Messages
1,437
Likes
4,686
I have been lurking for a while... great site indeed! And wonderful snake oil debunking.

As far as the CCA is concerned, I was an early adopter and own 5 now. Early days were a bit rough in terms of software/firmware with everything being usually brutally resampled to 48 kHz. Much better now, and of course great Roon support (the 5 CCAs were the reason I finally subscribed to Roon btw).

Listening impressions

on main system

- Linn Katalyst to Giya G3 vs CCA optical to Linn Katalyst to Giya G3: no difference that I can hear.
- Linn vs CCA analog to Linn on the G3: notable difference, Linn is better for everything (resolution, perceived DR, image)

on office system, Marantz HD-AMP1 to Kef LS50, also a difference, CCA analog doesn't cut it.

on other systems, NAD, Marantz/Yamaha "audiophile" on Tannoys or Klipsch, very hard to tell. Zero difference on Bose 901 Type VI ;) Good enough IMHO.

Basically, over the years, my subjective philosophy is that

- speakers are by very, very far what matters most (provided the amp can drive them).
- amps have a class "character" more than obvious quality differences, I often switch amps depending on the mood. Have had a few mismatches though such as a relatively expensive tube amp (was just a test, not a purchase) completely lacking bass on the Giyas.
- never could hear a difference with cables.
- DACs have a "character" too, although it not as obvious as amps. CCA analog is probably the worst DAC I have taken the time to compare (I confess I have never bought any Sc***t :) ) but I wouldn't know it if I hadn't done back to back comparisons on what I think are very resolving speakers and it depends heavily on the recording. Plus, it's a matter of taste I guess.

My Giya dealer went ballistic when I told him I loved the CCAs as "access points" - poof, there goes the business of selling me 5 Streaming DACs. I most certainly must have bad ears...

Used "Chesky's Ultimate Demonstration Disk" for testing specific aspects, but now gone from tidal :(

But again, I immensely appreciate Amir and many contributor's tremendous expertise here! Thanks.
 

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,658
Likes
5,276
Perhaps it is worth pointing out that the high (or low) dynamic range option on the CCA has nothing to do with what you may think of as dynamic range. As someone demonstrated on Hydrogen Audio, there is no compression in the 'low' setting - it simply has a lower output voltage. On some gear this may be very useful to avoid clipping of the input stage of the amplifier. A modest case in point is my Tivoli Audio model Two, that has a quite sensititive AUX input. In this and other cases it is better to avoid the HDR setting.
 

tlr

Member
Joined
Jan 4, 2019
Messages
42
Likes
65
I'm kind of tempted to stuff one of these in a JDS Atom, hard wired to the aux input.
 
Joined
Feb 8, 2019
Messages
22
Likes
20
Location
Australia
I wouldn't underestimate the important of using the supplied Chromecast Audio power supply. When I tested it in 2016, I did my first run using a computer to supply power. SNR was -85dBA with 16 bit audio. So I redid it using the included PS. It jumped to -97dBA. When I used 24 bit signals, it scored better than -104dBA.

That said, I'm in Australia so the PS may not necessarily be identical (although I do think that it's multivoltage).
 

AndrovichIV

Active Member
Forum Donor
Joined
May 25, 2019
Messages
158
Likes
218
has anyone tested this against a Modi 2 Uber? I know that thing measures horrible. I'm trying to decide if it's worth selling the Modi and just using the analog output of the Chromecast.
 
Top Bottom