A Subjective Blind Comparison of 2in to 3.5in drivers - Round 5

Select the driver that sounds best to you.

  • A

    Votes: 10 32.3%
  • B

    Votes: 0 0.0%
  • C

    Votes: 1 3.2%
  • D

    Votes: 3 9.7%
  • E

    Votes: 5 16.1%
  • F

    Votes: 6 19.4%
  • G

    Votes: 6 19.4%

  • Total voters
    31
  • Poll closed .
think I've a 1541 in a Shek dac - fwiw, transformer coupled 4398 are smooth and sometimes a pick over active outputs - final outcome might depend somewhat on the amp/cans/speakers. I've got a real old AN with tube output and tube shunt regulation which can sound good - Emotiva's first DAC with discrete output stage crappy volume control works well with a SE EL34UL amp and Klipsch Heresy I with "soft" KGB caps - mix and match for best result :D

What do you use today with active filtering before the amps ?
 
Founder of XSA-Labs
Joined 2012
Paid Member
I have had this rather modest DAC for some time now, and just plugged it into the SPDIF coax out of my CD player rather than the headphone outs I have always used for the tests, and then plug the RCA outs of the DAC into the miniDSP inputs. It definitely sounds better - more clear, more dynamic, quite noticeable. Neat what a $10 DAC can do. I have had this CD player - a Philips 5 disc carousel since 1994 and never even thought about using its native digital out. This can be really useful for getting sound out of the newer game consoles or cable boxes that all seem to sport optical Toslink outs.

CA Digital Optical Toslink/SPDIF Coax to Analog L/R RCA Audio Converter Adapter ith USB cable and fiber cable-in HDMI from Consumer Electronics on Aliexpress.com | Alibaba Group

Quite the value for $10. Even comes with Toslink fiber, and a power adapter. I will have to open it up to see what chip is inside.

If I just switched to a battery to power it, may sound even better...
 
Last edited:
No thanks for me ! Do you know what is Inside ? ;) why not with battery : just putt with a 6V lead battery : 2 x one Farad 5.5 V cap in serie // WITH a 1500 uF VERY LOW esr with low drop reg :), but I surmise bad oap and caps at outputt so certainly a waste of time than going towards a battery !

Waste of time for you as now your are with active filtering :)


So for the clips I don't advise my frenc music :

https://www.youtube.com/watch?v=GU6psn30Pyk

https://www.youtube.com/watch?v=m8ZcAhe03hw

https://www.youtube.com/watch?v=XPy7fhvBG6A

https://www.youtube.com/watch?v=6z1ssy_W_ZQ

where are my pills !
 
Last edited:
DAC and converter talk

Cornelis Spronk, Eldam, xrk971, freddi,

Please don't take this as critism but as a quality progress suggestion instead, although it will involve you believe in good objective data else it doesn't matter, and are ready to local do a little work yourself to get objective data revealed.

The thing that triggers me write this post is because of post 184 and what ever nice cheap or expensive DAC or converter or digital amps one purchase, think we need to analyze objective data to see if its any flaws or opposite its good data that is responsible we feel or declare this device sounds better than another device. Think for a long time many even cheap silicone perform excellent but its layout or filters and time to market means a lot, and therefor just because the product have a famous silicone or brand or good review somewhere in the world or is expensive is not a reassure it perform well as other product with same silicone.

Therefor suggest you benchmark your devices local because we can nowadays : ) to see product quality and if to trust sound local now is really improved. Below are three different DACs some are good all over some are good at a particular sample rate, some run good at WDM drivers some need WASAPI drive and some ASIO to be good. Below benchmark program is free REW but its important to zoom in and not run normal wide ranges as in datasheets because it hides too much. ARTA and Rightmark Audio Analyzer is alternatives but free RAA can not zoom in to high resolution without a license. Think it would be use full to benchmark all devices in chain because any device flaws would end up in our speaker measurements, and if knowing of any it could be corrected at another device either upstream or downstream as a system correction, or the device could be exchanged to a better one.

Below wood of traces think show if i just believe new device is better all over because have a famous chip or is more expensive is not to trust. One device of the three perform excellent all sample rates with WDM/WASAPI/ASIO interface and two of them only at 176,4/192kHz sample rate and sometimes very different if using WDM/WASAPI/ASIO interface. Regarding why to zoom in see in 2dB vertical scale frq response the green one that wiggles a lot, had it been perfect it should have been a thin strait trace as some of the others is, and when running a 50dB scale or more it actual looks false just showing a thin trace.

How to benchmark xrk971 CD player if its best with build in analog DAC or the $10 external one is not as easy and wonder if Barleywater or others have some good suggestions, but one could burn a CD with different square wave wav-files and listen with a RTA, it would look close to the SR shown below. Various test files could be created in free Audacity and one of the test CD's for Dayton Audio's Omni-mic package is free for download at their site.
 

Attachments

  • FR.png
    FR.png
    152.2 KB · Views: 490
  • IR.png
    IR.png
    94.5 KB · Views: 375
  • SR.png
    SR.png
    90 KB · Views: 371
Last edited:
I avoid the whole DAC issue by not having one.

How are you playing your music then? Vinyl? Tape?

A source like a PC or a notebook also has a DAC (built-in or added) AKA the "sound card".

But for this discussion, I guess we are referring to an additional device with USB/Toslink/Coax interface which will do Digital to analog conversion.
 
Last edited:
Bah, all of this is just a casual conversation to have a rest after the great efforst of the poll !:) ! After all we arn't hear the dac in the mp3 reccorded sessions but we are able to hear Xover changes as X find itself with Barely.

And X has also others mic I assume to check if it's better for him or not !

Well it's better to play via a modern USB to spidf via a pc instead of a big jittered cd player and simple DAC ! (I didn't understand than X needed to play disc but just files from pc for the speakers reccordings sessions.

If the CD player spidf for playback can not be avoided then the Inano from IfI is a good bet for the bucks : AMR technology embeded will supress many jitter (so has a coaxial spidf input), the chip is 32 bits, the volume plot is in the numeric domain not in the signal path ! It's battery feeded !

Buy and forgett ( storys of dacs and DIY : good Q/P !) !
If you are totaly active with PC, better to go for devices with I2S front end from a DSP stufs towars external dacs (curryman at mini dsp for cheap and good enough if money is a concern) to go above simple 96 Khz than spidf provides only with coax (192 kHz with TOSlink). Of course this DSP stuffs will be feeded by USB to avoid spidf limitations about upsampled frequencies !

End of the story !

(But as already said by me and BYRTT :layout around a dac chip stays important !).

One can ask advices for cheap improvements tehnics to studios guys like Welterys or Mitchba members ?

A the end X does not need advice to listen to music for himself as he is active with MiniDSP embeded DAC chips !

So I say again : was just a casual conversation around DAC and tralala for the fun and for exchanges point of views !

regards
 
Have a UCA222 which is nearly same as UCA202 X suggested, think its really good performer for CD reproduction at low cost. A 16bit 44,1/48kHz device so if running advanced DSP and one wants more bandwidth to correct nasty's above 20kHz in a IRR filter to tame some peaks/dips for a full ranger or tweeter driver it has limit.

Picture 1 UCA222 running 44,1 and 48kHz looks good when compared to AP192 at 44,1kHz, FR plot is set high resolution 0,1dB per division as in post 186.
Picture 2 and 3 IR/SR the same three.
Picture 4 UCA222 HD where it loose compared to AP192 in picture 5 running Windows sound system, which again loose to same AP192 set as ASIO driver in picture 6.
Picture 7 is low distortion AP192 at 192kHz without analog circuit in a digital loopback via SPDIF, IR/SR in this digital loop is pure textbook perfect.
 

Attachments

  • AP192 44,1kHz ASIO.png
    AP192 44,1kHz ASIO.png
    52.3 KB · Views: 52
  • AP192 44,1kHz Windows sound system.png
    AP192 44,1kHz Windows sound system.png
    53.7 KB · Views: 261
  • HD UCA222 44,1kHz.png
    HD UCA222 44,1kHz.png
    52.3 KB · Views: 362
  • SR.png
    SR.png
    41.3 KB · Views: 359
  • IR.png
    IR.png
    38.5 KB · Views: 367
  • FR.png
    FR.png
    58.7 KB · Views: 533
  • AP192 SPDIF 192kHz ASIO.png
    AP192 SPDIF 192kHz ASIO.png
    61.3 KB · Views: 62
You have a better luck to have a good hifi by knowing how to make a speaker and design its filter for your room with a bad dac than having a good dac with a bad designed speaker for this same room ! All being equal with money..... Hey !

The Hypex DSP is not so bad if you can Stream FIR from your PC.... Versatil on input/output !

Godzilla, we have already told you you're too big for a house !
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Have a UCA222 which is nearly same as UCA202 X suggested, think its really good performer for CD reproduction at low cost. A 16bit 44,1/48kHz device so if running advanced DSP and one wants more bandwidth to correct nasty's above 20kHz in a IRR filter to tame some peaks/dips for a full ranger or tweeter driver it has limit.

Picture 1 UCA222 running 44,1 and 48kHz looks good when compared to AP192 at 44,1kHz, FR plot is set high resolution 0,1dB per division as in post 186.
Picture 2 and 3 IR/SR the same three.
Picture 4 UCA222 HD where it loose compared to AP192 in picture 5 running Windows sound system, which again loose to same AP192 set as ASIO driver in picture 6.
Picture 7 is low distortion AP192 at 192kHz without analog circuit in a digital loopback via SPDIF, IR/SR in this digital loop is pure textbook perfect.

Thanks for the electrical experiments once again. These are great. I need to try looping back my UCa202 to try this.

Why is 44.1k better than 48k on UCA for step response?

I can check my DAC by using Toslink optical from UCA202 to input of DAC and then feed DAC RCA put back into UCA analog in.

More than a diversion, this discussion is helping me to upgrade the signal path for Round 6.

If simply using a decent DAC can improve sound I'm all for it. I still have my new LM3886 amps that I can try upgrading the signal path too. Need +/-25VAC toroidal trafo for LM3886 power supply.

Or Connexelectronic SMPS 32v dual rail power supply.
 
Last edited:
Thanks for the electrical experiments once again. These are great. I need to try looping back my UCa202 to try this.

Why is 44.1k better than 48k on UCA for step response?

I can check my DAC by using Toslink optical from UCA202 to input of DAC and then feed DAC RCA put back into UCA analog in.....

Yes be interesting try loop your UCA202 alone itself by hardwire analog I/O then after that try SPDIF out to the other DAC and back to UCA input and see if any improvement or degrading compared in overlaid plots. You probably know but cant stop saying it because its important to avoid sample rate conversion running windows sound system from Vista and up, remember go into windows sound control panel and change both input and output to the sample rate you intend set in REW to get true benchmark. For the UCA222 on Behringer website they offer a older ASIO driver but admit had never tried it out and maybe its available or same for UCA202, there is possibility it lowers HD by bypassing Windows sound system and another benefit running ASIO is if setting REW to ASIO and changing samplerate inside REW have exclusive access to sound device so one don't need the mess changing Windows sound panel setting.

Regarding UCA222 is better at 44,1kHz verse 48kHz don't hang me if wrong, but think they have relative same roll off filter to cut digital noise and that 48kHz have more bandwidth then seen with 48kHz eyes the roll off is steeper and steeper filters gives higher % attack step compared % point where decay starts, also look at pre verse post ringing is nice symmetri for 44,1kHz and as good as AP192 is, where 48kHz have unsymetri with a lot more post ringing visual both at IR and SR plots.
 
Last edited:
511197d1445960227t-subjective-blind-comparison-2in-3-5in-drivers-round-5-fr.png


The differences in tiny ripple at HF are not the primary source in sound quality differences between DACs. High pass behavior of DAC and following amplification stages is bigger player. Looking at big picture of step response zero crossing, phase response/group delay <100Hz is more telling. In a nutshell, it's getting transient perfect response to low frequencies. Problem starts with high pass filter on electrical side, and continues with high pass behavior of transducers.

BYRTT; in this set of plots:

511151d1445937274t-subjective-blind-comparison-2in-3-5in-drivers-round-5-fr.png


you appear to have some DC coupled DACs? How do these compare in bass realism to UCA222?
 
from RMAA reports, 1541 and 1543 are a couple of dB down by 15KHz. I have a UMC202 which showed an ~ 14KHz spike with the free version of RMAA - Behringer said they could not find it - one of these days I'll search for it with TrueRTA. UCA202 seemed pretty good including its A-D but never used it much as a DAC with a power amp. My Ross Martin has ADA4898-1-2 outputs - fwiw a lot of the time I prefer 4398 with transformer coupling. Emotiva DAC on sale offer nice input switching and volume control.

there was no CRT "on" in the house, lights checked, other devices including an old Xitel INport were clean on the top.
It was tough to get UMC202 to hold with RMAA
3htP25M.jpg
EDORmtA.jpg
 
Last edited:
:mad: I already advertised you X !.....

Ok, I send you Nora Jones for the punition.... Lucky guy :D !


Btw, more seriously, I'm sure looking at the first page of the dymanic range database here we should be finding in our respective database 3 albums of the most dynamics reccording Album list - Dynamic Range Database

For Instance : Happy mondays : I should have it
I saw some Chesky reccords : I haven't but sure someone has one
The Velvet Underground : Peel session
Dire Straits : Money for Nothing : someone here talked about this album iirc
Billie Holyday : Lady in Satin (both voice & classic orchestra while being Jazz ;) )
....

Just having a look at the 2 first pages !

For me, personally I don't always think the highest DR is automatically better. I agree on the wrong turn the loudness wars have taken. But a DR value alone does not tell much about the recorded quality. The Chesky stuff would be a good choice though. He really makes the effort to optimise the recording techniques. I believe Bob Katz has done several of those recordings.
But don't just pick something off of that list and expect good things based on DR numbers alone.
One band that usually impresses me with recorded quality is Steely Dan. It shows pretty good numbers in the DR database as well. Just don't get hung up on DR numbers alone.
 
Hello,

I have Lady in Satin, this reccording is good, although I prefer her Porgy & Bess reccording...

Did we ever hear a good reccording of the Velvett Undergrood which has the palm of sound cave with maybe the lives of Joy Division (and in fact their albums studio as well !) ? Nahhhhhhhhh ! gett off of the list !

Dire straits is not my cup of tea, even if I admit this guy have a good left hand on the guitar, but to have heard this album in the list, I believe it 's a good reccording to my tastes ! Could it be like Nora Jones: always lovely ? Maybe, some reccordings are always sounding good with all systems !

Do you remember of this track " we 're talking about the revolution, etc" by tracy Chapman : it was a so surrealist reccording, even in LP than I bought my Rega Planar III 30 years ago just because it was sounding so good on it !

A noob question, I have a Kef 104/2 with little mtm mids, but are 2 x 2" or 3" in mtm able to move enough air upon 1 K hz with such little FR88 drivers, etc ! ?
 
Last edited: