-
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!
Focusrite Clarett 2Pre USB - Tear down, bit of internal analysis and few in-house measurements
oversorrow
Member
The Clarett+ 2Pre review from JK is out.
ADC and DAC chips both have changed to Cirrus Logic's flagship products.
Pray for AKM btw.
ADC and DAC chips both have changed to Cirrus Logic's flagship products.
Pray for AKM btw.
OP
- Thread Starter
- #23
Why, please? Can you elaborate? The time-start from your YT link is related to how to use Clarett software and the lack of Line/Instrument physical buttons, same as previous version, so it's what Focusrite was choosing for Clarett design).Pray for AKM btw.
oversorrow
Member
The channel balance of the headphone output is pretty bad. Could it be like that across the Clarett+ 2Pre line, or just an issue with Julian's specific interface?The Clarett+ 2Pre review from JK is out.
ADC and DAC chips both have changed to Cirrus Logic's flagship products.
Pray for AKM btw.
- Joined
- Apr 18, 2021
- Messages
- 13
- Likes
- 69
Pretty sure it's only my particular unit. In the video I shortly mention it at 13:48. It seems the potentiometer is defective and I expect the channel balance of other units to be more in line with my previously measure Clarett 2Pre USB.
I've had the Thunderbolt Clarett 2 Pre (not plus) for about 3 years now. From the start, I was quite disappointed with the outputs when the DAC is engaged - they seem to produce better audio when monitoring analogue input, but playing back digital audio is messy, possibly audibly distorted. I started out using them for all sound on my PC, but ended up sending all non-ASIO output from the PC through a Node 2i or Mojo instead, which while not exactly the best DACs in the world, do sound way better.
Reading this thread today, I realise that disappointing performance I heard is pretty much how I'd imagine that graph of high THD+N up to 10kHz to sound, so maybe there are a bunch of duff units about and those who got them are missing out on the quality other people are raving about.
The inputs are much more performant though - especially with the 'air' feature that helps appreciably with my dull voice and adds some sparkle to sax and acoustic guitar. That said, perhaps this feature is simply a small counter to poor environment or timbre of my cheaper instruments - my Nord Stage 3 usually seems better off without it.
Overall, if the output quality really does vary this much, I probably wouldn't go for Focusrite again or particularly recommend them. It's a very nice looking, well-featured unit with good feel of build, but good playback is pretty essential, and there have been other problems too - rather poor support/documentation for the elderly Thunderbolt interface, and popping issues early on that might have been down to their driver at the time.
As upgrade fever strikes, my eye is now on the RME ADI-2 Pro FS, which should smash everything I have for playback and recording. I could keep the Clarett for when low latency is crucial (monitoring with VSTs in the chain), and for extra inputs with air and phantom power, with the ADI-2 handling everything else.
Reading this thread today, I realise that disappointing performance I heard is pretty much how I'd imagine that graph of high THD+N up to 10kHz to sound, so maybe there are a bunch of duff units about and those who got them are missing out on the quality other people are raving about.
The inputs are much more performant though - especially with the 'air' feature that helps appreciably with my dull voice and adds some sparkle to sax and acoustic guitar. That said, perhaps this feature is simply a small counter to poor environment or timbre of my cheaper instruments - my Nord Stage 3 usually seems better off without it.
Overall, if the output quality really does vary this much, I probably wouldn't go for Focusrite again or particularly recommend them. It's a very nice looking, well-featured unit with good feel of build, but good playback is pretty essential, and there have been other problems too - rather poor support/documentation for the elderly Thunderbolt interface, and popping issues early on that might have been down to their driver at the time.
As upgrade fever strikes, my eye is now on the RME ADI-2 Pro FS, which should smash everything I have for playback and recording. I could keep the Clarett for when low latency is crucial (monitoring with VSTs in the chain), and for extra inputs with air and phantom power, with the ADI-2 handling everything else.
OP
- Thread Starter
- #29
Not sure I fully understand your issue, is it related to Thunderbolt device? I'm also not sure what's wrong with the 10 kHz measurements too.
SINAD is 105 dB, so it's quite OK for most folks and distortions will definitely not be audible.
SINAD is 105 dB, so it's quite OK for most folks and distortions will definitely not be audible.
Sorry, a late night post and I think I was confused about what you "were uncomfortable with" which, together with my own experiences, led me to all-too-hastily misinterpret a graph. Re-reading, you were looking at the headphone jack after there was a discrepancy mentioned on another post and found it behaves well for you, so neither a problem, nor the thing I was talking about. Please correct me if I'm still misunderstanding your post I linked, I find a lot of posts on this site presume a lot more familiarity than I have.
My main issue was just that all playback sounded like it had some noticeable problem, seeming like noise throughout most of my hearing spectrum. I don't know for sure if that's the DAC implementation or something else, but it monitors analogue input fine, only having this issue with digital. I have no fancy electronic measuring equipment to test further.
It's probably not limited to the Thunderbolt model unless the Thunderbolt and USB versions are different beyond just the interface. I haven't seen a teardown of the TB one to be sure, but I can't see why they'd do anything else differently.
Sounds like my issues could be unique and unexpected. But it's out of warranty now. Having not had any opportunity to try much gear out I just assumed for years that this was simply the flawed sound quality you should expect on an audio interface at this price.
OP
- Thread Starter
- #31
Maybe your unit was having some hardware issues or, why not, software issues (related to both firmware or driver). You can also try using another laptop/computer and recheck, but best thing to do in such cases would be using ARTA or REW or Multitone or RMAA and investigate what the issue is about, then you can update this thread (or create a new one) with your findings.
Awesome! I kept reading Amir's reviews and seeing him mention how much his analyser cost I thought this stuff was the reserve of people with much more to spend than I have. The fact you can just download some software and get measuring makes me wonder why so few audiophiles and even professional reviewers don't bother to check these things...
Anyway, I have had some fun getting familiar with ARTA and checking my Clarett in the looping-back-to-itself configuration (using a decent Mogami cable so that shouldn't be a limitation). Here's what I found with the monitor dial at half-way (around where I'd typically be listening):
I'm a noob at this but it looks like there's a ton of distortion on odd harmonics. To determine whether this was coming from the input side or the output side, I tried connecting my Chord Mojo to the same input:
The huge peaks of harmonic distortion are gone, so I think there's something quite wrong with my Clarett's output (be it the DAC or preamp).
I've adjusted input gain on the Clarett to get RMS = -1.5 dBFS simply because I saw it in yours. However my noise floor seems higher than yours in both these tests. Perhaps it's another fault with my unit, but let me know if this might suggest I've tested incorrectly.
Again, thanks for recommending the software. I'm now tempted to test all my devices! But unfortunately I don't have quality adapters to plug things into the Clarett.
I don't think I have another ADC of sufficient quality to test with. I have a supposedly OK-quality consumer sound card in my old PC, but probably not worth wiring up. Similarly my Behringer UCA 202 is pretty badly distorted for analog - I bought it just for cheap USB Toslink output.
But I had a lengthy play around with each gain control (in hardware and software) and found an optimal case.
I've been listening for a while with the output around this configuration and I'm pretty happy with the sound. THD+N beats my Mojo's, as I'd expect from others' measurements. Most of the problems I saw occur if any of the gain is turned up too high, or in some cases too low.
This is a refurbished Clarett direct from Focusrite so perhaps there were some defects missed or unaddressed. In support of this, I did notice while playing at the weekend that one input channel has a sudden burst of high pitch whine when above about 3/4 gain. It's not clipping, it's present even with no input, and the noise floor gets very audible too. Still, knowing accurately where it falls down means I can get the best out of it and ARTA's been really helpful for pinpointing problems I wouldn't immediately understand or hear in music.
restorer-john
Grand Contributor
That's straight up clipping. Odd harmonics 3/5/7/9 etc.
Good spot. Yeah, that's where the worst of the distortion comes from. It was actually the software settings causing the clipping. I might've realised that straight away 15 years ago, back when I knew the Fourier series of a square wave is just the odd harmonics...
paulraphael
Active Member
Looks like there's a new version—same box with new guts. Reported specs are a bit better. Apparently result of fire at AKM factory. Unfortunately the same incomplete reporting on headphone amp performance. Would like to see if this one does significantly better with low impedance / low efficiency cans.
Similar threads
