So Matt, would you mind to tell me which one do you prefer? The "smoother" one, or the "sharper" one?
No, I don't mind I was listening on my Grado headphones, which as you know aren't very accurate . Unable to easily connect to hi-fi, may try later to see if my preference changes. So far I prefer the smoother version, there is plenty of strident brass which doesn't hurt from being toned down a bit, also I believe my tinnitus makes me a bit over sensitive to the higher harmonics.
I was listening on my Grado headphones, which as you know aren't very accurate . Unable to easily connect to hi-fi, may try later to see if my preference changes. So far I prefer the smoother version, there is plenty of strident brass which doesn't hurt from being toned down a bit, also I believe my tinnitus makes me a bit over sensitive to the higher harmonics.
Ok, second listen tonight and same findings......preference for WS2 although I am not in love with either.
WS1 is subjectively quite distorted especially on complex/loud passages, WS2 is 'clean' but could be more musical which means could be cleaner/clearer.
WS2 has stronger low bass and better dynamics overall but not totally 'transparent'.......needs more life.
WS1 is fatiguing type dirty, sounding like a kitchen radio.....can be used in background but never enjoyable.
I am interested to see what you did wrong lol.
Dan.
WS1 is subjectively quite distorted especially on complex/loud passages, WS2 is 'clean' but could be more musical which means could be cleaner/clearer.
WS2 has stronger low bass and better dynamics overall but not totally 'transparent'.......needs more life.
WS1 is fatiguing type dirty, sounding like a kitchen radio.....can be used in background but never enjoyable.
I am interested to see what you did wrong lol.
Dan.
Thanks for your comments, Dan. We have to wait till Nov 18 with a disclosure, the poll will expire then. I understand we shall probably not collect much more votes than we already have, but it would not be wise to disclose everything immediately.
Would you like to try ABX? I know it is both boring and exhausting, however ......
Would you like to try ABX? I know it is both boring and exhausting, however ......
I have my preliminary test report on both preamps already prepared. It might be interesting, IMO, and not only for me.
It is highly probable that we shall not get much more test impressions and comments here, so I have decided to disclose the data to keep some attention. It is of course possible to continue in poll votes. I am also posting test reports of the components and complete paths used.
---------------------------------------------------------------
Listening test of 2 preamps – worldstring test
I have prepared this test for the reason that I was quite sure that I can hear a difference between 2 preamplifiers that should fulfill all necessary parameters for audio link level preamps. It is then difficult to make a verification – because the complete audio chain cannot be effectively transferred into the virtual web space ;-). So I made the only possible thing – inserted the 2 preamps into a DAC – ADC loop and made 2 sample recordings at same settings and same levels, these 2 recordings were posted on-line. Technical description of the DAC – ADC loop and 2 preamps under test follows.
DAC – ADC path (loop)
The samples were recorded at 96kHz/24bit format. The DAC-ADC loop has maximum ADC input level of 1Vrms (1dBV). Minimum of THD(1kHz) is 0.00034% at 0.24V – below this point distortion is buried in noise. At 1V, THD(1kHz) is 0.0026%. THD is quite flat with frequency, as will be seen from the posted plot. Both input and output are only single-ended, provided with RCA connectors.
Worldstring1 sample
was recorded through a preamplifier that I call “MSYS” and is intended as an analog interface for a soundcard, for measuring purposes. It is not normally used as an audio preamplifier. It has a balanced input, consisting of two 2SK170BL JFETs, followed by AD797 opamp. Then it has a gain control, stepped attenuator, DACT Danish product. This is followed by ADA4898 opamp to make a single ended output. It also has a balanced output, which was not used in this test. Because the DAC output is only single-ended, the parameters of the balanced input were not fully utilized and the distortion parameters were thus degraded a bit. However, some parameters remained quite remarkable, IMO.
Frequency response 20Hz – 40kHz / better than +/-0.02dB
Frequency response 10Hz – 1MHz / better than 0/-1.6dB
Rise time Tr(10%-90%) 0.342 us
Output noise -111dBV(A), complete DAC-preamp-ADC loop
THD(1kHz) 0.0005% at 0.1V, complete DAC-preamp-ADC loop
THD(1kHz) 0.005% at 0.7V, complete DAC-preamp-ADC loop
THD(10kHz) 0.006% at 0.7V, complete DAC-preamp-ADC loop
Input impedance 100 kohm
Output impedance 100 ohm
Worldstring2 sample
was recorded through a preamplifier that I call “BALPRE” and is intended as a single-ended to balanced converter. It starts with a blue Alps 2x10k pot, followed by OPA134 opamp, this followed by a DRV134 circuit with balanced floating output. OUT- was grounded and OUT+ used as a single-ended output from this preamp.
Frequency response 20Hz – 40kHz / better than 0/-0.25dB
Frequency response 10Hz – 1MHz / better than 0/-21dB
Rise time Tr(10%-90%) 2.90 us
Output noise -100dBV(A), complete DAC-preamp-ADC loop
THD(1kHz) 0.0004% at 0.15V, complete DAC-preamp-ADC loop
THD(1kHz) 0.003% at 0.7V, complete DAC-preamp-ADC loop
THD(10kHz) 0.0035% at 0.7V, complete DAC-preamp-ADC loop
Input impedance 10 kohm
Output impedance 100 ohm
High frequency noise was measured as well, to see if any of the preamps is sending HF garbage to ADC. Noise spectra are flat in both cases and worldstring1 preamplifier noise is again about 10dB lower than noise of worldstring2 preamplifier, similar result as for audio band noise.
To conclude, worldstring1 preamp has much wider bandwidth and about 10dB lower noise than worldstring2 preamp. Distortion is slightly higher because of sub-optimal use of the input stage (SE input signal to a balanced input stage). From the comments here in the forum, it seems that worldstring2 is evaluated as subjectively better, based on listening. This is in agreement with my long term experience – very transparent audio components with highest dynamic range are usually not preferred by listening, especially if the music sound material is not technically perfect.
Thanks everyone who has participated in the test and wrote comments.
---------------------------------------------------------------
Listening test of 2 preamps – worldstring test
I have prepared this test for the reason that I was quite sure that I can hear a difference between 2 preamplifiers that should fulfill all necessary parameters for audio link level preamps. It is then difficult to make a verification – because the complete audio chain cannot be effectively transferred into the virtual web space ;-). So I made the only possible thing – inserted the 2 preamps into a DAC – ADC loop and made 2 sample recordings at same settings and same levels, these 2 recordings were posted on-line. Technical description of the DAC – ADC loop and 2 preamps under test follows.
DAC – ADC path (loop)
The samples were recorded at 96kHz/24bit format. The DAC-ADC loop has maximum ADC input level of 1Vrms (1dBV). Minimum of THD(1kHz) is 0.00034% at 0.24V – below this point distortion is buried in noise. At 1V, THD(1kHz) is 0.0026%. THD is quite flat with frequency, as will be seen from the posted plot. Both input and output are only single-ended, provided with RCA connectors.
Worldstring1 sample
was recorded through a preamplifier that I call “MSYS” and is intended as an analog interface for a soundcard, for measuring purposes. It is not normally used as an audio preamplifier. It has a balanced input, consisting of two 2SK170BL JFETs, followed by AD797 opamp. Then it has a gain control, stepped attenuator, DACT Danish product. This is followed by ADA4898 opamp to make a single ended output. It also has a balanced output, which was not used in this test. Because the DAC output is only single-ended, the parameters of the balanced input were not fully utilized and the distortion parameters were thus degraded a bit. However, some parameters remained quite remarkable, IMO.
Frequency response 20Hz – 40kHz / better than +/-0.02dB
Frequency response 10Hz – 1MHz / better than 0/-1.6dB
Rise time Tr(10%-90%) 0.342 us
Output noise -111dBV(A), complete DAC-preamp-ADC loop
THD(1kHz) 0.0005% at 0.1V, complete DAC-preamp-ADC loop
THD(1kHz) 0.005% at 0.7V, complete DAC-preamp-ADC loop
THD(10kHz) 0.006% at 0.7V, complete DAC-preamp-ADC loop
Input impedance 100 kohm
Output impedance 100 ohm
Worldstring2 sample
was recorded through a preamplifier that I call “BALPRE” and is intended as a single-ended to balanced converter. It starts with a blue Alps 2x10k pot, followed by OPA134 opamp, this followed by a DRV134 circuit with balanced floating output. OUT- was grounded and OUT+ used as a single-ended output from this preamp.
Frequency response 20Hz – 40kHz / better than 0/-0.25dB
Frequency response 10Hz – 1MHz / better than 0/-21dB
Rise time Tr(10%-90%) 2.90 us
Output noise -100dBV(A), complete DAC-preamp-ADC loop
THD(1kHz) 0.0004% at 0.15V, complete DAC-preamp-ADC loop
THD(1kHz) 0.003% at 0.7V, complete DAC-preamp-ADC loop
THD(10kHz) 0.0035% at 0.7V, complete DAC-preamp-ADC loop
Input impedance 10 kohm
Output impedance 100 ohm
High frequency noise was measured as well, to see if any of the preamps is sending HF garbage to ADC. Noise spectra are flat in both cases and worldstring1 preamplifier noise is again about 10dB lower than noise of worldstring2 preamplifier, similar result as for audio band noise.
To conclude, worldstring1 preamp has much wider bandwidth and about 10dB lower noise than worldstring2 preamp. Distortion is slightly higher because of sub-optimal use of the input stage (SE input signal to a balanced input stage). From the comments here in the forum, it seems that worldstring2 is evaluated as subjectively better, based on listening. This is in agreement with my long term experience – very transparent audio components with highest dynamic range are usually not preferred by listening, especially if the music sound material is not technically perfect.
Thanks everyone who has participated in the test and wrote comments.
However slight, the IMD hump and the low output bias of the AD797 issues add a couple of footnotes to the conclusion. Perhaps a worthwhile investigation for another time.
AD797 is loaded with 10 kohm here. Would you kindly post a link to a trusted test that shows the "IMD hump"? Not mere comments of forum members, please, and no DAC application. Can it be seen in Samuel Groner's AD797 detailed measurement?
Does that also take feedback resistor loading into account? If so then it has to supply 0.1mA for 1V peak, 20% of bias. How would it sound with > 40k loading? Could be interesting but not that much for immediate attention.
These ok? Not sure about trusted.
https://www.audiosciencereview.com/forum/index.php?attachments/topping-dx7s-dac-intermodulation-distortion-measurements-png.10829/
https://www.audiosciencereview.com/forum/index.php?attachments/1546492665323-png.19807/
I thought you use DAC as source. Could be relevant but I have no way to ascertain without verification measurement. I just notice the potential issue after Mark's post, the issue was suspected to be related to common mode rejection of the AD797.
OPA1611/1612 was mentioned as a fix to the hump issue, similar noise performance and higher output bias at half the price.
I do not know any published IMD measurement of AD797 by Samuel Groner. Sorry.
If you want for me to take your comments on AD797 serious, please read Samuel Groner's detailed AD797 measurement first
http://www.nanovolt.ch/resources/ic_opamps/pdf/opamp_distortion.pdf
and compare it to other opamps. Look at loading conditions.
The opamp is a part of stand-alone preamp in my test, not of a DAC.
Which IMD you want for me to check? 60Hz+7kHz, 250Hz+8kHz, 19+20kHz?
I find DAC IMD hump irrelevant and most of your comment based on audiophile beliefs, not on trusted and verified sources and measurements.
Last edited:
You are kidding right? AFAIK that file shows no IMD measurement on AD797.
I have no wish for you to do any more measurement on my count. Op amps are mildly interesting for me anyway, not much of a difference in sonic to spend a lot of effort.
Don't be so dismissive to unverified measurement. IMD of LME49990 showed similar jagged line on the datasheet itself, albeit at a much lower level.
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.