My Listening to Natalie
There are 2 things that may make this impossible for me to predict the effect of the post DAC cap: [1] The DAC power supply supercap is not installed in one of the setup [2] The original circuit has HF peak (Mooly's measurement) that may contribute to fatigue.
I listened (and ABXed) B1 versus B3. No contest here. The bass in B3 is so deep and wide, just like some report for the Joe mod (Coris?). I can't be sure if it is not an effect of louder dB level, or the supercap (in case B1 is the original).
But if B3 were louder than B1 then there is no bias that I heard fatigue with B1. But this could be the HF peak (if it is the original).
Knowing that most audiophile listen to music like the Natalie, I can understand if people prefer Joe modification (I would too. But the issue with supercap and HF peak is not clear yet).
Like I had guessed from the beginning, the mod sounds more like NOS DAC. Very good for slow music, but for music that require high level dynamic/transient (Such as the Genesis, tho not a perfect example), OS type is more musical. You can check my listening preference with Mooly's OS versus NOS listening test.
There are 2 things that may make this impossible for me to predict the effect of the post DAC cap: [1] The DAC power supply supercap is not installed in one of the setup [2] The original circuit has HF peak (Mooly's measurement) that may contribute to fatigue.
I listened (and ABXed) B1 versus B3. No contest here. The bass in B3 is so deep and wide, just like some report for the Joe mod (Coris?). I can't be sure if it is not an effect of louder dB level, or the supercap (in case B1 is the original).
But if B3 were louder than B1 then there is no bias that I heard fatigue with B1. But this could be the HF peak (if it is the original).
Knowing that most audiophile listen to music like the Natalie, I can understand if people prefer Joe modification (I would too. But the issue with supercap and HF peak is not clear yet).
Like I had guessed from the beginning, the mod sounds more like NOS DAC. Very good for slow music, but for music that require high level dynamic/transient (Such as the Genesis, tho not a perfect example), OS type is more musical. You can check my listening preference with Mooly's OS versus NOS listening test.
I am a little out of the loop at the moment, but if you are feeding those files back into another delta-sigma DAC, then I think the whole thing has been negated.
Try Moderator 'anatech' since he has a non-ds (or non-sd) DAC to play those files and get his findings. That would make a lot more sense IMO.
If I play those files back, guess what, my DAC has the post-DAC filter and will probably work - may need to feed through JRiver's 64 Para-EQ or else I will now be down an accumulated -4dB @20KHz. Too much in any's book.
In your case, why mod just play standard music files with the DAC with approx 470R x 2 plus cap (whatever values you ended up with in that ballpark, I did see the diagram but only got a few minutes now and cannot check actual values you used) and play them - not a DBT, maybe just listen for pleasure and maybe just like it and yet later do further analysis. Maybe I trust your ears more than others, but what can be the harm. Gotta run.
The player in standard form meets its spec of 20-20kHz -/+ 0.3db I can tell you that the -0.3db bit is right up at 20khz. Up to at least 19kHz is flat with respect to 1kHz. The same applies to my component values used in the mod. That has been proved comprehensively by me using Red Book test CD's checking both normal and recordings with pre-emphasis. So what appears at the line out sockets of the player for mod vs standard is as near as possible to call equal as regards response irregularities.
I'll post my impressions later today.
So as promised 
These were my impressions at the time of listening which I typed out at the time. This is a copy and paste of that note. And as stated at the end, YMMV.
File A1 Genesis is un-modified player.
File A2 Genesis is modified player. Preferred. Vocals in louder sections seemed more solid and better defined.
File B1 Natalie C is modified player.
File B2 Natalie C is un-modified player. Preferred. It just seemed to hang together better but it was a very close call.
File A3 Genesis is player modified to 'final spec'. This was a track of two halves. In the build up to the halfway point I felt the modded player was 'deeper' sounding and presented a more believable soundstage, however the louder sections seemed to be 'reigned in' somewhat. Also felt the vocals hardened a little too.
File B3 Natalie C is player modified to 'final spec'. On balance I preferred the modified player. The vocal image seemed more stable with the instrumental backing backing seeming to have more authority and a 'deeper' kind of sound.
All these differences were evaluated via speakers (B&W 703) and on repeated listening comparing one to another. YMMV
These were my impressions at the time of listening which I typed out at the time. This is a copy and paste of that note. And as stated at the end, YMMV.File A1 Genesis is un-modified player.
File A2 Genesis is modified player. Preferred. Vocals in louder sections seemed more solid and better defined.
File B1 Natalie C is modified player.
File B2 Natalie C is un-modified player. Preferred. It just seemed to hang together better but it was a very close call.
File A3 Genesis is player modified to 'final spec'. This was a track of two halves. In the build up to the halfway point I felt the modded player was 'deeper' sounding and presented a more believable soundstage, however the louder sections seemed to be 'reigned in' somewhat. Also felt the vocals hardened a little too.
File B3 Natalie C is player modified to 'final spec'. On balance I preferred the modified player. The vocal image seemed more stable with the instrumental backing backing seeming to have more authority and a 'deeper' kind of sound.
All these differences were evaluated via speakers (B&W 703) and on repeated listening comparing one to another. YMMV
So as promised
A final note on A1 vs A3, I compared just the region below 100Hz mathematically using FFT's and found substantial differences that I have no explanation for. I did this because just over 100Hz or so the two tracks when aligned are almost exactly the same (+.5dB taken into account) until the higher frequencies. The transition is quite abrupt too.
If Joe or anyone can provide any references to this sub Hz jitter and audio other than the usual anecdotes or "if you can't hear it you're deaf" it would be appreciated. Other than that I've spent enough time on this.
I have no idea on that either. You guys can run rings round me when it comes to looking at data like this. All I can say with certainty is that the files were produced using identical settings on the PC. So honest answer is 'I dunno...'
For those interested in taking this further, which I assume would be Joe, then this needs all needs looking at more carefully. The voltage DAC in a mid-price player of the 90's isn't the ultimate playback source, I think we all appreciate that. All I have done is take the mods at face value and present the results as best I could to give an A-B comparison. To listen to the player in real time is a different ball game because with just one player there could be a gap of a day or two between being able to strip it down and redo or undo any changes.
So the batton is now passed on to those wanting to delve deeper.
Not quite sure I was looking on a linear frequency scale the transition was quite abrupt at just over 100Hz but the highs behaved as I expected a gradual mismatch due to the filtering/phase shifts that are present in the circuit. A3 is almost exactly 9 samples delayed from A1, all my math was done at 64bit float.
I listened to A1 & B1, since the 3rd track is said to be different in level i did not listen to it. A really valid test would involve listening to the analog output (with 2 players feed by a computer toallow easy switching). I have not been following the thread since people started posting their results
1/ The output of the mod is convolved with Karl's ADC, and with my DAC. Since we are likely dealing with subtle information it could easily be lost
2/ I did get the reference file from Karl for A. This Genisis track has none of the kind of information that has been attributed to the mod revealing. So i was expecting no difference. I detected none. This is a bad track to pick for trying to pick out the kind of differences described from anecdotal info.
3/ i did not get a ref for B, but i detected a very slight bit of greater 3d-ness in the image and a bit more body in the song.
dave
1/ The output of the mod is convolved with Karl's ADC, and with my DAC. Since we are likely dealing with subtle information it could easily be lost
2/ I did get the reference file from Karl for A. This Genisis track has none of the kind of information that has been attributed to the mod revealing. So i was expecting no difference. I detected none. This is a bad track to pick for trying to pick out the kind of differences described from anecdotal info.
3/ i did not get a ref for B, but i detected a very slight bit of greater 3d-ness in the image and a bit more body in the song.
dave
Well maybe later this weekend I'll have time to post some pictures of what I found.
- Status
- Not open for further replies.