https://www.audiosciencereview.com/foru ... 131.23138/
I found a creative SB G6 secondhand ,
then only noticed Rock's post about the Fosi DS-2 with dual CS43131 ,
For whats its worth , the Creative drivers are a bit like mud wrestling with a pig,
each one causing a different range of sample/bit rates to become avialable ,
So far the period correct drivers seem to come out best ,
Heres a useful set of tools ,
https://github.com/FORARTfe/HyMPS/blob/ ... ools.md#--
Creative SB G6/CS43131
Re: Creative SB G6/CS43131
I spent a lot of time tooling about with the various driver packages , most dont play nice with the device due to firmware updates , with various sample and bit rates either not working or becoming unavailable depending on the driver used .
The one surprise was using the HP amp out in a balanced config from REW I was able to get +24dbm into 300 ohms , +18dbm into 600 ohms and +12dbm into 1200 ohms with a 0dbfs output , its not low distortion though,
I was able to use the Koord asio (flexasio gui) to amalgamate SSL2 as input and G6 as the output device , I was able to get distortion down to around 0.0012% with careful adjustment .
I tried checking the latency of the G6 ,input to output but I couldnt seem ,with any setting of the controls ,to get less than around 70ms ,
Maybe 70mS delay is of no major consequence to gamers , but for music production its way to late to be of any use at all ...unless you want digital delay.
The encoder based volume control is also absolutely horrible , if you try and move the control fast it stops responding , if you move it at a slow even rate it works ,but your talking a couple of rotations to go from min to max , and the way the control is recessed , its just non intuitive , with no tactile feel at all .
I just wish now I had discovered the FOSI(bear) DS-2 before I spent my 60 euros on the G6 .
Its a gamer dac , the quoted 130db dynamic range and 1mS latency is bullshit as far as I can see, the drivers/software is a window dressing exercise needlessly full of bloat.
Who needs the truth when you have a bunch of gamer nerds turned media moguls in the driving seat , like Musk ,Zucker, Bezos etc
The one surprise was using the HP amp out in a balanced config from REW I was able to get +24dbm into 300 ohms , +18dbm into 600 ohms and +12dbm into 1200 ohms with a 0dbfs output , its not low distortion though,
I was able to use the Koord asio (flexasio gui) to amalgamate SSL2 as input and G6 as the output device , I was able to get distortion down to around 0.0012% with careful adjustment .
I tried checking the latency of the G6 ,input to output but I couldnt seem ,with any setting of the controls ,to get less than around 70ms ,
Maybe 70mS delay is of no major consequence to gamers , but for music production its way to late to be of any use at all ...unless you want digital delay.
The encoder based volume control is also absolutely horrible , if you try and move the control fast it stops responding , if you move it at a slow even rate it works ,but your talking a couple of rotations to go from min to max , and the way the control is recessed , its just non intuitive , with no tactile feel at all .
I just wish now I had discovered the FOSI(bear) DS-2 before I spent my 60 euros on the G6 .
Its a gamer dac , the quoted 130db dynamic range and 1mS latency is bullshit as far as I can see, the drivers/software is a window dressing exercise needlessly full of bloat.
Who needs the truth when you have a bunch of gamer nerds turned media moguls in the driving seat , like Musk ,Zucker, Bezos etc

Re: Creative SB G6/CS43131
I was able to find a set of reliable USB and asio drivers for the G6 , without the need to use the C++ gui /app .
You have to make sure the unit ,which by default is set to virtual 7.1 output to stereo in windows sound panel .
then the full range of sample/bit rates becomes available .
I ran the output at 384khz and its able to output a test signal upto 100khz ,+/-0.25db ,or -1db at 122khz , from line the line output ,
the headphone output already starts to get wiggly at 20khz , and rapidly degrades above that frequency , so not much good for testing ,
The circuit that produces the +/- rails for the op amps in the G6 and SSL2 is very similar ,
One difference i can see is on the G6 they use a sheilded inductor , while on the SSL it seems open core inductors are used ,
this might explain some of the extra induced hf noise under certain conditions in the SSL2 .
You have to make sure the unit ,which by default is set to virtual 7.1 output to stereo in windows sound panel .
then the full range of sample/bit rates becomes available .
I ran the output at 384khz and its able to output a test signal upto 100khz ,+/-0.25db ,or -1db at 122khz , from line the line output ,
the headphone output already starts to get wiggly at 20khz , and rapidly degrades above that frequency , so not much good for testing ,
The circuit that produces the +/- rails for the op amps in the G6 and SSL2 is very similar ,
One difference i can see is on the G6 they use a sheilded inductor , while on the SSL it seems open core inductors are used ,
this might explain some of the extra induced hf noise under certain conditions in the SSL2 .
- mediatechnology
- Posts: 5828
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: Creative SB G6/CS43131
Thanks for posting this.
Would be curious to see the A/D's noise level over a 20 kHz bandwidth.
Would be curious to see the A/D's noise level over a 20 kHz bandwidth.
Re: Creative SB G6/CS43131
Its not great , the input is heavily filtered above 20khz , theres a gigantic step in the noise waveform , noise is apparently low ,but as level increases so do super sonic whistles and harmonics ,
I did the output side frequency measurements with the fluke 187 ,G6 line out L/R configured balanced in the REW generator panel .
I set the output for 0dbm at 1khz then measured 10,100khz and found the -1db point .
I found a neat way to supply my audio devices with from a powerbank to avoid motherboard power.
I have a usb hub which senses power on the incoming lines , so if I plug the powerbank into the hub ,the usb device comes on in standalone mode , If I then connect the USBC end of the hub to the pc it seamlessly hooks in the USB data lines, without interuption to the audio .
I did the output side frequency measurements with the fluke 187 ,G6 line out L/R configured balanced in the REW generator panel .
I set the output for 0dbm at 1khz then measured 10,100khz and found the -1db point .
I found a neat way to supply my audio devices with from a powerbank to avoid motherboard power.
I have a usb hub which senses power on the incoming lines , so if I plug the powerbank into the hub ,the usb device comes on in standalone mode , If I then connect the USBC end of the hub to the pc it seamlessly hooks in the USB data lines, without interuption to the audio .
Last edited by Tubetec on Wed Jan 22, 2025 4:32 pm, edited 1 time in total.
- mediatechnology
- Posts: 5828
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: Creative SB G6/CS43131
I don't use 192K for test and measurement because I've always found it to be noisier than 96K so it's hard for me to compare.
At 96K the Focusrite 2i2 Gen3 measures around -96 dBFS.
At -101 the SB G6 still quite usable.
The PCM4222 EVM is right at -124 dBFS. It has a naked front end and has a great converter so I use it for very low-level noise measurements.
My daily driver is the 2i2.
At 96K the Focusrite 2i2 Gen3 measures around -96 dBFS.
At -101 the SB G6 still quite usable.
The PCM4222 EVM is right at -124 dBFS. It has a naked front end and has a great converter so I use it for very low-level noise measurements.
My daily driver is the 2i2.
Re: Creative SB G6/CS43131
I found a link to the inductors used in the G6 ,
https://4donline.ihs.com/images/VipMast ... 5776-1.pdf
These are indeed sheilded components ,
It seems these would require the correct placement /orientation to work effectively ,
the component land patern might also be different on the SSL pcb.
https://4donline.ihs.com/images/VipMast ... 5776-1.pdf
These are indeed sheilded components ,
It seems these would require the correct placement /orientation to work effectively ,
the component land patern might also be different on the SSL pcb.
- mediatechnology
- Posts: 5828
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Re: Creative SB G6/CS43131
In the upper pic the inductors are Coilcraft. https://www.coilcraft.com/
Your link returns a 403 here.
Your link returns a 403 here.
Re: Creative SB G6/CS43131
I found previously that if I set the SSL2 at 48khz then changed to 192 I got a huge rise in noise in the 30khz-100khz range , evidenced by the bump in the noise floor .
A similar thing happens with the G6 , except I get a step above 20khz in the noise floor you saw in the last picture I posted .
Making sure the audio interface is initialised at the wanted samplerate via the .txt config file in REW and I got the result you see below ,
First pic is noise floor , nice and level all the way out to nearly 90khz ,
second is two frequency response curves , green from headphone out , red from line out - to line input .
A much better result than I got previously ,
Theres something going on here , its not specific to SSL or Soundblaster , its more like some process going on within Windows , perhaps resampling or maybe REW doesnt set the filters correctly when the sample rate is changed , starting REW from the .txt config specifically set up for the sample rate you want seems to avoid the problem .
Wayne , you mentioned getting poor HF noise results on 192khz with the Focusrite in REW ,
Try configuring the device to 192khz , then save the preferences as 192khz.txt , then restart rew , load the config from the text file via preferences , restart REW again and you should see better performance .
Theres a few others out there also getting varying results in REW , Im not sure if this is a problem in Windows or REW or both .
would be nice to figure out why .
John Mulcahy is a member over in Gdiy , maybe someone might point him in the direction of this thread ,
A similar thing happens with the G6 , except I get a step above 20khz in the noise floor you saw in the last picture I posted .
Making sure the audio interface is initialised at the wanted samplerate via the .txt config file in REW and I got the result you see below ,
First pic is noise floor , nice and level all the way out to nearly 90khz ,
second is two frequency response curves , green from headphone out , red from line out - to line input .
A much better result than I got previously ,
Theres something going on here , its not specific to SSL or Soundblaster , its more like some process going on within Windows , perhaps resampling or maybe REW doesnt set the filters correctly when the sample rate is changed , starting REW from the .txt config specifically set up for the sample rate you want seems to avoid the problem .
Wayne , you mentioned getting poor HF noise results on 192khz with the Focusrite in REW ,
Try configuring the device to 192khz , then save the preferences as 192khz.txt , then restart rew , load the config from the text file via preferences , restart REW again and you should see better performance .
Theres a few others out there also getting varying results in REW , Im not sure if this is a problem in Windows or REW or both .
would be nice to figure out why .
John Mulcahy is a member over in Gdiy , maybe someone might point him in the direction of this thread ,