😱 I am going even further off topic, Ken stop me if you want to concentrate on the core of the thread, but the discussion might be worth it....
so:
I have these super duper analog minimalistic life recordings (haha also some from a studio from the Netherlands...) and I have them on "CD" (ripped), High Res (192/24), LP and maser tape first copy (38cm/s 2-track) playing on a Revox PR99 - all from the same recording session! just different formats...
They all sound different in terms of dynamics, spatial sound, stage, micro dynamics, texture of voices and instruments etc etc
Now comes the personal thing.... Lampie likes "CD Dynamics" I like the spatial and natural sound from tape and vinyl (yes, I go to live concerts very regularly, sitting first row, I know how voices and instruments sound in real)
again just my 2 cents. It is (still) a free world and thank God I did not see a sever cancel culture in the Audio world (yet 🙄 )
so:
I have these super duper analog minimalistic life recordings (haha also some from a studio from the Netherlands...) and I have them on "CD" (ripped), High Res (192/24), LP and maser tape first copy (38cm/s 2-track) playing on a Revox PR99 - all from the same recording session! just different formats...
They all sound different in terms of dynamics, spatial sound, stage, micro dynamics, texture of voices and instruments etc etc
Now comes the personal thing.... Lampie likes "CD Dynamics" I like the spatial and natural sound from tape and vinyl (yes, I go to live concerts very regularly, sitting first row, I know how voices and instruments sound in real)
again just my 2 cents. It is (still) a free world and thank God I did not see a sever cancel culture in the Audio world (yet 🙄 )
Optical phono cartridges can get closer to some of the good things about CD that magnetic cartridges can't quite do, particularly if the rest of the reproduction system is very good.
Subjectively comparing an very good dac with a high end optical phono setup using electrostat speakers in a treated room shows that phono can be more true to the original musical performance than digital in some ways, and digital can be more true than phono in other ways.
There is still more work to do on digital it seems.
Subjectively comparing an very good dac with a high end optical phono setup using electrostat speakers in a treated room shows that phono can be more true to the original musical performance than digital in some ways, and digital can be more true than phono in other ways.
There is still more work to do on digital it seems.
Last edited:
Then make you own recordings and compare (as i do ) !
I use a HD recorder with a separate PCM A/D converter (customized). Revox G36 2 Track tape recorder etc. Tube pro mixing console (top of the bill) and different mic's to play with.
Even the lowly portable Mini Disc recorder will do a better job then any analog tape !
For playback i use electrostatic loudspeakers driven with dedicated high voltage amplifiers (OTL) for outstanding dynamics and control.
The entire set is balanced from source to speakers (the nature of ESL).
I use a HD recorder with a separate PCM A/D converter (customized). Revox G36 2 Track tape recorder etc. Tube pro mixing console (top of the bill) and different mic's to play with.
Even the lowly portable Mini Disc recorder will do a better job then any analog tape !
For playback i use electrostatic loudspeakers driven with dedicated high voltage amplifiers (OTL) for outstanding dynamics and control.
The entire set is balanced from source to speakers (the nature of ESL).
Attachments
Last edited:
There is professional recording gear here too, along with a variety of musical instruments. Imperfections in the chain can be found everywhere.
Don't have a disk lathe though.
Don't have a disk lathe though.
Have you checked this track?
Gregorio Paniagua - De pastoribus | Mathematica dies irae | Crepuscularis | Sine Nomine (...) - YouTube
Gregorio Paniagua - De pastoribus | Mathematica dies irae | Crepuscularis | Sine Nomine (...) - YouTube
and here we are, if all of what we talk about here in the dac chain is not made before in the adc chain of the reccording, is it not too late ? All are not Chesky Records...
As far as the performance of CD versus Vinyl is concerned, CD wins. Where Vinyl typically seems to have a subjective advantage is in exactly those same areas which NOS CD seems to provide. Low long-term listening-fatigue, 3-dimensionality, a hard to describe naturalness. Which is why my greatest hope is that our conclusions in this thread will show us how to combine the performance of CD with the naturalness of Vinyl.
I meant is there no filters and oversampling choice in the adc chains in the studios and complex reccordings process, all numeric, brick walls, noise gate and so on ! Quite complex than studio engineer job. The best and the worst exists, nothing you can save after in the dac chain I assume ? OK off topic.
@dddac
"I am going even further off topic, Ken stop me if you want to concentrate on the core of the thread, but the discussion might be worth it...."
Doede,
No problem. I feel that brief digressions help to keep everyone engaged.
"I am going even further off topic, Ken stop me if you want to concentrate on the core of the thread, but the discussion might be worth it...."
Doede,
No problem. I feel that brief digressions help to keep everyone engaged.
"my greatest hope is that our conclusions in this thread will show us how to combine the performance of CD with the naturalness of Vinyl"
You are invited !!!
You are invited !!!
As hobbyists, we can only affect what we have the power to affect, and we don't have the power to affect the front (ADC) end of the chain. The good news is that we ARE able to affect an interesting difference in the sound character of the back (DAC) end of the chain, as OS versus NOS shows.
This treated room you keep bringing up, do you know what the frequency response and decay time are like?
You can reduce ringing of the entire chain at the expense of treble, or suppress the aliases that the halfband filters used during recording haven't suppressed at the expense of bandwidth.
THE NEXT SUSPECT CATEGORY
Without further significant objection, the suspect list will now appear as shown at the bottom. I suggest that we move forward and address the largest remaining category, B) RECONSTRUCTION/IMAGE-BAND HANDLING. As I consider this category, it seems that the primary distinction among the items there are, 1) whether the image-bands are suppressed, and 2) if they are suppressed, how is that suppression accomplished?
Therefore, I will suggest that whether the DAC operates with, or without, a task specific reconstruction filter somewhere, the image-bands still are very significantly suppressed by the human ear, and to a lesser degree initially, it turns out, by the loudspeaker's tweeters. Those two stages of ultrasonic suppression are always in effect, regardless of any reconstruction filter elsewhere in the system. The data reported on the high-frequency response of the human ears varies some, and also largely depends on an individual listener's age and gender. That said, I quickly located a chart showing that a male, aged 40-49, has a 20KHz hearing response greater than -50dB down from their hearing reponse at 1KHz.
High Frequency Hearing Loss - an overview | ScienceDirect Topics.
A quick review of typical tweeter datasheets reveals that while, they too, offer image-band suppression, it is relatively much less (often, not showing double-digit dB levels of suppression until 30KHz) than is provided by the ear itself. In net, what this strongly suggests to me is that the presence of the image-bands themselves is not what is responsible for the subjective difference between OS and NOS sound. I'm certain, however, that there are many other studies reporting various other findings, so, feel free to share any significantly different data which you may find.
Please share your reasoning and your conclusions (if you have any) on the question of whether or not the presence of the image-bands themselves (not yet, on the method by which they are being suppressed) is of major significance.
B) RECONSTRUCTION/IMAGE-BAND HANDLING
===============================================
2) Lack of an FIR interpolation-filter, freeing the DAC from certain processing 'artifacts' , such as:
a) time-domain signal echoes produced within Equiripple on-chip FIR filters.
b) impulse response ringing (pre or post)
c) half-band filters plainly violating Nyquist
d) are prone to clip on peak sample normalized recordings - the intersample overshoot issue.
e) purely analog image-band suppression inherently sounds different than digital suppression?
3) Phase-modulation of the baseband signal due to insufficiently suppressed image-bands. In other words, because the signal waveform is not fully reconstructed according to the sampling theorem requirements.
4) The unsuppressed image-bands are, somehow, producing audible IM products directly within the ear.
C) ALTERED JITTER IMPACT
=============================================
5) Different jitter impact due to fewer D/A conversion cycles per second.
6) Reduced supply and ground noise due to slower clock rates.
D) SAMPLE-PERIOD RELATED QUANTIZATION ERRORS
==============================================
7) Converter settling-time becomes a smaller percentage of each conversion period as the conversion rate is made slower.
8) Harmonic-distortion may be sample-rate dependent. (per, Markw4)
Without further significant objection, the suspect list will now appear as shown at the bottom. I suggest that we move forward and address the largest remaining category, B) RECONSTRUCTION/IMAGE-BAND HANDLING. As I consider this category, it seems that the primary distinction among the items there are, 1) whether the image-bands are suppressed, and 2) if they are suppressed, how is that suppression accomplished?
Therefore, I will suggest that whether the DAC operates with, or without, a task specific reconstruction filter somewhere, the image-bands still are very significantly suppressed by the human ear, and to a lesser degree initially, it turns out, by the loudspeaker's tweeters. Those two stages of ultrasonic suppression are always in effect, regardless of any reconstruction filter elsewhere in the system. The data reported on the high-frequency response of the human ears varies some, and also largely depends on an individual listener's age and gender. That said, I quickly located a chart showing that a male, aged 40-49, has a 20KHz hearing response greater than -50dB down from their hearing reponse at 1KHz.
High Frequency Hearing Loss - an overview | ScienceDirect Topics.
A quick review of typical tweeter datasheets reveals that while, they too, offer image-band suppression, it is relatively much less (often, not showing double-digit dB levels of suppression until 30KHz) than is provided by the ear itself. In net, what this strongly suggests to me is that the presence of the image-bands themselves is not what is responsible for the subjective difference between OS and NOS sound. I'm certain, however, that there are many other studies reporting various other findings, so, feel free to share any significantly different data which you may find.
Please share your reasoning and your conclusions (if you have any) on the question of whether or not the presence of the image-bands themselves (not yet, on the method by which they are being suppressed) is of major significance.
B) RECONSTRUCTION/IMAGE-BAND HANDLING
===============================================
2) Lack of an FIR interpolation-filter, freeing the DAC from certain processing 'artifacts' , such as:
a) time-domain signal echoes produced within Equiripple on-chip FIR filters.
b) impulse response ringing (pre or post)
c) half-band filters plainly violating Nyquist
d) are prone to clip on peak sample normalized recordings - the intersample overshoot issue.
e) purely analog image-band suppression inherently sounds different than digital suppression?
3) Phase-modulation of the baseband signal due to insufficiently suppressed image-bands. In other words, because the signal waveform is not fully reconstructed according to the sampling theorem requirements.
4) The unsuppressed image-bands are, somehow, producing audible IM products directly within the ear.
C) ALTERED JITTER IMPACT
=============================================
5) Different jitter impact due to fewer D/A conversion cycles per second.
6) Reduced supply and ground noise due to slower clock rates.
D) SAMPLE-PERIOD RELATED QUANTIZATION ERRORS
==============================================
7) Converter settling-time becomes a smaller percentage of each conversion period as the conversion rate is made slower.
8) Harmonic-distortion may be sample-rate dependent. (per, Markw4)
I am not very big specialist in this topic, but i would like to mention some things which are obvious for me.
comparing OS vs NOS DAC should be done only adding oversampling(specific shape of OS filter) and not changing the rest of the system. The same chip should be used. If you compare R2R to sigma delta many things are different. Noise, THD, jitter tolerance. So the comparison is meaningless. If we want to compare something specific we have to isolate things as the science does.
In case when two cases are identical below 20kHz it also doesn't mean the sound will have the same impact due to hypersonic effect.
comparing OS vs NOS DAC should be done only adding oversampling(specific shape of OS filter) and not changing the rest of the system. The same chip should be used. If you compare R2R to sigma delta many things are different. Noise, THD, jitter tolerance. So the comparison is meaningless. If we want to compare something specific we have to isolate things as the science does.
In case when two cases are identical below 20kHz it also doesn't mean the sound will have the same impact due to hypersonic effect.
If there is a FR change due to the OS vs NOS, do you think we shall compensate for that when doing comparison? I mean make the 2 different setups end up with the same frequency response? (+/- 0,01 dB?)
If so, how an we do that compensation totally transparent?
//
If so, how an we do that compensation totally transparent?
//
I'm planning to disconnect the tweeters for that. Conveniently, they are cut at 5kHz.
Last edited:
Regarding these interesting slides, are people here who are equipped with an oscilloscope and some DAC chips able to look at the raw waveform coming from the DAC and assessing the exponential settling and/or slew-rate limiting as described on page 9 of the slides? There might be something interesting to see there.