XXHighEnd

Hi y'all.

I finally found some time to do something that I've been meaning to do for a while now...

I wanted to see how close a digital capture on my Tascam recorder could sound compared to the original CD rip after it had been played back through the audio PC. So I set up the following chain:

Mach II audio PC -> USB-to-spdif converter -> BNC cable -> Tascam recorder digital input

I won't go through all the USB cables and various permutations that I tried (with, without Intona, etc), but here are my findings:

The closest digital capture to the original CD rip was with Clarixa cables and the Intona. I tried many vastly more expensive USB cables, and none got as close as Clarixa-Intona-Clarixa combo.

So how close is this digital capture to the original CD rip? Well, why not take a listen for yourselves? Here are the downloads:

1. Original CD rip:
https://drive.google.com/open?id=0B0PU5LO5jVjfUHozS3NENkhCLWM

2. Digital capture:
https://drive.google.com/open?id=0B0PU5LO5jVjfOXJGZ2xQZUMxMEk

Not absolutely identical, but VERY close. I think this is really impressive. Think about it... A file sits on my music server in my basement. It gets sent via ethernet to the audio PC over many meters distance. XX does its magic (RAM-OS, 'Copy to XX folder', MinOS, Unattended, etc) and sends it via a USB port to the USB cable. Having gone through the Intona, it reaches the USB-to-spdif converter. Once converted, it gets sent out via a BNC cable to the recorder's digital input, and is captured on an SDHC card in the recorder. All this, and the captured file remains so close in SQ to the original CD rip!

For me, it's a clear indication of how good the Mach II, Clarixa and Intona combination are.

I also captured the output from the following chain:

Mach II audio PC -> USB-to-spdif converter -> BNC cable -> DAC -> Tascam recorder analogue input

You're welcome to take a listen to this too...

3. Analogue capture:
https://drive.google.com/open?id=0B0PU5LO5jVjfTFBrZG5vRXE0UDQ

Hmmm... This is also incredibly close in SQ to the original CD rip. Again, not identical - there are differences, but they're pretty small. It's not that easy to differentiate between this file and the digital capture.

Now, do you want to hear the 'killer'? I used the analogue output of a 25 year-old 16-bit DAC for this!

Sure, we've made a lot of progress in the audio PC field (pioneered to a very large extent by the likes of Peter), but can we really say the same about DACs? Listening to the performance of this old-timer DAC, I'm not so sure...

Mani.

Hey Mani,


Haha, yes, quite killing. OK, not super bad perhaps, but the strings sound strange, the S'es harsh and the hall is non-existent. Flagiolettes and all are cut.
And then to think that this was my "verdict" without listening to the other two. :derisive:
Of course I ate a plate of placeboes in advance, but ...

I then listening to the CD Rip (with right away the familiar sound). This made me realize how inaccurate the first I listened to was. After that I played the Digital Output one and found no differences to notice. But notice I played all only once and for maybe just over 1 minute only.

We always agree over everything (for such tests) so we should agree about this too. But also notice that it can't be about a small very short listening test anyway if it were about D/A converters; after an hour you may be annoyed about all sounding the same or be disturbed about some flavour or are tired because of things being too harsh (that would be the case for soure with this example, which btw is too gentle for real testing).

Now I am waiting for the pitfall I fell in; some unpleasant surprise. :evil:
haha

Thanks !
Peter


PS: The conclusion could also be that the A/D does not add much. Right ?
Wrong (I misjudged the process).

Hey Peter, quick reply because I need to go to the gym... No, absolutely no surprises. All is exactly as I described.

I'll have more to say later, but for now I'll just remind you that file 3 went through a 25 year-old DAC and a delta-sigma ADC. It'd be very difficult to attribute the negatives that you're hearing to just the DAC. Of course, easily done if I do the same with a few more DACs, but keep the ADC the same - it's contribution to the sound should become apparent.

More later...

Mani.

OK !
(these things are always fun)

Yes, I agree totally with your description.

When I play these 3 files back through the 25 year-old DAC, I hear the differences you're describing - so, this DAC has the resolution to show these differences up easily. Also, when I was taking the captures, I was monitoring through the headphone socket on the ADC. The signal coming in from DAC was much closer to the original file than that captured by the ADC.

The 25 year-old DAC is a Sugden AU51 with a TDA1541A S2 Double Crown chip. It also uses the old 4x OS SAA7220P/B filter, which I suspect is contributing to the sonic signature you can hear. Yes, it certainly has a sonic signature, but I actually find it very pleasant - very musical and very easy on the ear.

I would certainly like to 'prove' how good this Sugden DAC is. In order to do this, I need to be able to capture its analogue output accurately. The ADC in the Tascam is clearly not up to the job, but its spdif digital input seems pretty transparent. So I'll be on the lookout for a nice 16-bit linear PCM ADC with a BNC spdif output. Not many of these around, but I've identified a few old DAT recorders that may do the trick.

Once I've sorted this out, I'll be doing a full 'Old School vs. New Kid' DAC comparison. There will be a number of rounds comparing DACs that cost a similar amount today (similar used price of old DAC vs. new price of modern DAC). The DACs I'll be comparing will be the following:

Round 1 (portable, ~$500): Hifiman HM603 (TDA1543) vs. Ponoplayer (ESS9018) vs. iFi Nano iDSD (DSD1793)

Round 2 (~$500): Sugden SDA-1 (TDA1541A S1 Single Crown) vs. iFi Micro iDSD (DSD1793)

Round 3 (~$1200): Sugden Au51 (TDA1541A S2 Double Crown) vs. Altmann Attraction (TDA1543, Paul Hynes PSU) vs. Chord 2Qute (26k tap WTA)

I am deliberately not including the NOS1a in the 'official' comparisons, though of course I will be comparing for my own benefit (hopefully after the B'ASS situation is resolved). The NOS1a will be the DAC that I play all the captured files back on to identify the performance of the DACs.

I already have all the 'old school' DACs here (been doing a lot of searching on eBay these last few weeks!). Just have to get hold of an iFi Micro and the Chord 2Qute, and I should be ready to go. (But still need to sort the ADC out first...)

Why am I doing this? To promote old DACs? Nah... it's because I'm genuinely interested in knowing the outcome and I think it'll be a lot of fun!

Mani.

Now that we've established that I can capture the digital output of the Mach II PC with very little loss (remember, the digital capture sounds almost identical to the original CD rip), I'd now like to capture the output whilst playing around with various XX settings, to see (hear!) what affect these settings actually have. I should be able to determine which settings are the most accurate. Working through all the various XX settings permutations will be fun  :wacko2:

Even if I do find the 'most accurate' settings, these would only apply to the particular USB-to-spdif converter I'm using, which is only 16/44.1 capable.

Once (if?) I achieve a close-to-100% ADC capture with the Tascam (I'm working on it) I'll be able to do the same with the analogue output of the NOS1a. I'm really excited about this because we may be able to put all the "I prefer this OS", or "I prefer a higher SFS", etc thinking to bed.

I wonder... is there really an ideal set of XX parameters???

Mani.

A slight change to my list. It'll now be:

Round 1 (portable, ~$500): Hifiman HM603 (TDA1543) vs. Ponoplayer (ESS9018) vs. iFi Nano iDSD (DSD1793)

Round 2 (~$1200): Sugden Au51 (TDA1541A S2 Double Crown) vs. Altmann Attraction (TDA1543, Paul Hynes PSU) vs. Chord 2Qute (26k tap WTA, Paul Hynes PSU)

I have all these DACs now. I'm still trying to optimise the ADC though. I think I'll just have to splash out on a new ADC... probably go for a Forrsell MADC-2. Shame they're all delta-sigmas now... :(

Once I'm done, I'll probably post my thoughts on CA. But I'll state right now that the Chord 2Qute is a seriously good DAC irrespective of how little it costs - I easily prefer it to the Schiit Yggy I had here a few months ago (sorry Brian!). But is it accurate? We'll see!

More later...

Mani.

I'm deliberately not starting a new thread for this post because I don't want to make a big deal about it - I'm fed up with people on CA thinking that I'm a secret agent for Phasure! But I thought a few of you might be interested nevertheless...

I've just compared my microRendu (powered by an Uptone UltraCap LPS-1 PSU) to my Mach II. I know many of you are busy, so I'll just give you my thoughts, and link the files for when you have time to take a listen yourselves.

1. Original CD rip (Julia Hülsmann Trio, 'Zahlen Bitte')

A typically good ECM recording. The cymbals are a bit too forward in this recording for my liking, but they have great delicacy and colour - I visualise them as being very thin, delicate and golden in colour. I wish there were more 'kick' from the rest of the drum kit. The piano is OK, but not as sweet-sounding as I'd like. And the bass is a bit too vague (until the solo near the end). But a nice track to my ears.

https://drive.google.com/open?id=0B0PU5LO5jVjfQmcyNWxZMFNPSUE

2. microRendu -> USB-to-spdif converter -> BNC cable -> Tascam recorder digital input

Flatter than the original CD rip - the dynamics are not there. But the colour/tone is pretty similar. I think this is about as good as you're going to get before getting into a fully optimised audio PC like the Mach II. For the cost of the microRendu/LPS-1, it's very good... but certainly not perfect. I'm sure the microRendu crowd on CA wouldn't agree...

https://drive.google.com/open?id=0B0PU5LO5jVjfNWF5YVl1WTVCSG8

3. Mach II -> USB-to-spdif converter -> BNC cable -> Tascam recorder digital input

OK, the dynamics are back. Very, very close to the original CD rip, though I feel there is a slight 'hardening' of the sound. The cymbals now sound 'cheaper', i.e. made of a thicker and inferior metal. I have SFS=4.00 and Q1-5=0 (xQ1=1). It might be possible to 'tune' the sound to become closer to the original by playing around with the XX settings - something that I'll do going forward, if I have the time and inclination. It also occurs to me that maybe this is 14393 showing some of its attributes?

https://drive.google.com/open?id=0B0PU5LO5jVjfcDNQVFllRjFuZ0k

I've given you my views. I'd really appreciate it if some of you could compare especially file 1 to file 3 and let me know what you think. Any ideas what I could do in XX to bridge the small gap that currently exists between 1 and 3?

Mani.

Hi Mani,

I will give this a try in a couple of hours.
But by what means did you listen to the original CD Rip ?

Thanks,
Peter

OK, I guess I need some psychological help here ...

I don't know what the gag or pitfall is, but with the first version I thought "oh boy, finally cymbals to compare but now all lack dynamics. It really disturbed me. Also the colour of the cymbals was not right. The tick in the beginning on them is too loud compared to the too thin remainder.
I noticed that especially at the end things were not right at all (was it hitting on the rim ? very close by that on the skin ?).

With the second there was much more body to the cymbals but the punch (sort of) in the drums was so much lacking that when the bass started I seemed to have missed the lower keyed drums (compared to the first version). Colour seemed too dirty of everything and halfway the piano started to disturb me (buzzes a bit through the room).

With the third all was suddenly OK ? Dynamics were there, cymbals had the right colour, the bass seemed to be more life-like and the end sounded dynamic as should.

Because I thought I might have gotten used to it all I replayed the last 3rd of each track, but I my initial judgement remains.


I'm sure I failed on you, Mani !
Peter

I've been comparing the 3 files mainly with Mach II -> NOS1a.


There really can be no failure here. Irrespective of what you think of the SQ of file 1, all we need to establish is whether file 2 or file 3 sounds the closest to file 1. File 1 could have been a recording from AM radio (although I admit it would be almost impossible to distinguish between the files in this case).

Interestingly, I think the playing back through the NOS1a makes it more difficult to distinguish between the 3 files. If I use the USB-to-spdif into another DAC (e.g. the Altmann, which I have here), the differences seem easier to pick up, to my ears. This is NOT because this setup is better than the NOS1a - definitely NOT. No, it's more because the NOS1a seems to make all 3 files sound actually quite good. Through the Altmann, file 2 sounds quite flat compared to the others, but through the NOS1a, it sounds OK - not as dynamic as the other files, but OK. It's almost like the NOS1a brings out the best in the file, flawed as it is.

I'm not sure if I'm making any sense...


I mentioned earlier that to me, the cymbals on file 3 sound different - I said 'cheaper', and made of thicker and inferior metal. But in retrospect, I can totally understand how they could actually sound more real than the very, very delicate - almost paper-thin - cymbals on file 1. They seem to shimmer more once struck, so you don't just get the initial strike, but some body afterwards also.

But irrespective of our preferences, it means that the Mach II is changing the sound. It's a very pleasant change, but it's still a change.

Overall, I still maintain that file 3 sounds closer than file 2 to the original file 1.

My conclusions:

1. Both the microRendu/LPS-1 and the Mach II are 'near-perfect' 'transports', if setup correctly. (I can only claim this to be the case with the particular USB-to-spdif converter I used, limited to a 16/44.1 output.)

2. There ain't much wrong with USB for audio purposes, if done right (at 16/44.1, at least).

3. There ain't much wrong with spdif, if done right (at 16/44.1, at least).

Happy to hear other thoughts...

Mani.

What playback software was used to capture the files in 2 and 3? Will give them a listen this evening.

Hi Brian,

File 2: Roon -> HQPlayer -> bit-perfect 16/44.1 output
File 3: XXHighEnd -> bit-perfect 16/44.1 output

Mani.

Haha...

The drummer is Heinrich Köbberling. I couldn't find which drums he actually used on this recording, but he endorses Canopus Drums. Have a look at the first video here: http://canopusdrums.com/en/concept/. The cymbals are exactly the way I visualised them when listening to the original CD rip.

Mani.

Mani,

To be 100% sure : Is that original CD Rip the original file, or did it go throught some ADC process ?

Thanks,
Peter

Is this the first video ?
https://www.youtube.com/watch?v=za96T19N3XM&feature=youtu.be
I think it is, but ...

It's the original CD rip, absolutely 100%. No ADC whatsoever. All I did was change the name. But here it is again in case the name change screwed something up  :grin::

https://drive.google.com/open?id=0B0PU5LO5jVjfMjJkNFJ5M1RxSnc

Mani.

No, scroll down towards the bottom of http://canopusdrums.com/en/concept/ to 'RECENT VIDEOS'. It's the first video with Hyuk Jang playing.

Mani.

Thanks.

Say I have these two albums from The Julia Hülsmann Trio :
- Good Morning Midnight
- The End of a Summer

Would you accidentally have one of these and which you ripped by the same means as the one uploaded ? (assumed you ripped that yourself)
If so, please upload one track (I suggest 09-Riverman from Good Mornign Midnight or 02-Konbanwa from The End of a Summer).
If you have both (in same rip fahsion as your uploaded test track), I'd prefer The End of a Summer album.

Cheers,
Peter

Well, that is then super smart of you, because I could not get a vision of them at all. I was only highly disturbed of a strange recording method with the drums up front (and yes, later I saw you writing about them being forward) and poor microphones or whatever that made them sound wrongly all over. Also they are not played at all like in the video, so ... smart.
And if you also get some real sense of the sound by means of that video, even more smart ! haha
Mind you, that video is about showcasing the different snare drums. Maybe you can hear the difference between them ?
:) :)

Anyway, I know you said yourself it wasn't the best, but maybe it doesn't really work out for comparisons if things are not he best. Mind you, this should all be objective comparison, but if things are wrong in the base, it becomes subjective (we can't compare to reality and only can have it sound to our own best taste now).
If you have one of the albums I pointed out, you'll know what I mean (OK, you already did).

Regards,
Peter

Unfortunately, I don't have rips of either of those albums. But I've attached the rip info for Zahlen Bitte - I'm very confident it's a good/accuarte rip.

But does it even matter how the CD was ripped or how it sounds... for the purposes of this test? I mean, all we need to do is to determine how closely files 2 and 3 match file 1, irrespective of whether we 'like' the sound of file 1.

What am I missing in your thinking?

Mani.

I totally disagree. We don't have to compare to reality at all. All you have to do is compare to file 1.

If I went to Paris and took a photo of the Mono Lisa with my phone, I could figure out how accurate my photo was compared to the painting. Would you then say to me that that's not a fair comparison because I don't have Mono Lisa (the woman) actually standing in front of me to see what she really looked like? Or that Da Vinci wasn't a very good artist? Or that the materials he used make it difficult to compare? These would be irrelevant... no? The reference would be the painting and nothing else. The reference here is file 1.

But I do agree that the higher the 'qaulity' of file 1, the easier it should be to compare. In which case, I think ECM should be a good choice.

Mani.

Hey Peter, just send me 02-Konbanwa from The End of a Summer and I'll repeat the 'test'...

Mani.

Mani, you did not get my message ...

You seem to be able to compare to anything, no matter how flawed that anything is. I can't even listen to it; I get too disturbed.

Now combine with my personal judgement that 3 sounds better than the original.
READ :
All what is not good to begin with can be improved by fairly random means. Example : Say that the original sounds way too bright. All it takes is a pile of noise (say ADC) to take out some of the brightness and it will "be" better.

What it comes down to (I'm afraid) is that you will sit down and listen to that Imprint album with a slight remark of the drums being a bit forward, while I will shut it down in 30 seconds because everything is wrong with it (note my all so often remarks "does not make sense"). Same counts for the bass. So yes, you too made a remark about that, but apparently you can listen to that because you (I'd say) are less interested in reality. I hear nothing but strange bass sounds and don't know where to go with it.

If you like I can upload a track of The end of a Summer (also ECM). I suppose you will be shocked.
Or not. I don't know ...

The comparison with the Mona Lisa goes different IMHO :
You hand me your iPhone and show me the Mona Lisa as well the Last Supper and ask me which is the better painting (without me knowing either original).
This does not make sense.

Peter

Ah, OK. Haha.
Coming up ...

https://drive.google.com/open?id=0B82Qx9VwoovtUGExZWNrVEZfbDA

There it is.

Btw Mani, I was not really suggesting to repeat the test with a better example, but I sure will contribute again and if I again choose for #3 to be the better one, so be it !
But in my view strange things need to happen, because this one (The End of a Summer - Konbanwa) does not disturb me to begin with and although anything undoubtedly can be improved, it should not happen by means of an additional ADC process. I am confident we agree about that.
Obviously it bugs me that I chose #3 while it should have been #1 (which I even knew in advance, but that distance I can take (if that's English)).

Thanks,
Peter

Nah, totally wrong. I take you to the Louvre and then put my iPhone photo next to the painting and ask you how closely it matches the painting. You say, "I can't do it. I don't like her smile... it disturbs me too much - it's too false. Ciska never smiles like that." I would then slap you across the face (nicely) and say, "Get a grip of yourself. Whether you like the smile or not is totally irrelevant to this exercise."

But if you really can't even listen to file 1, then I agree, it's not going work.

Mani.

Thanks. Will repeat test a little later on.


[Italics mine.]

OK, this is the 2nd time you've mentioned an ADC in this thread. There is no ADC, because there is no 'A' to convert - it's 'D' all the way from the NAS drive to the SD card in the digital capturing device.

I hope this is absolutely clear.

Mani.

I agree. But it is you who introduces the smile example to get your right.
What it is really about is that you NFC your iPhone next to a printer (OK, no NFC on iPhones ;)), print it, put it to the wall, and ask me what I think about the painting.
My answer : nothing, because this is not a painting.
You could also ask me abot the picture. Then I might say : this is not a picture because it is a painting.
Ok, you say, then what about the composition. My response could be : without the real lighting as applied in the Louvre - hard to tell.

Moral : we force ourselves into the subjective because it doesn't start out with reality.

So we're both in Paris now ...
we walk down those stairs and, go to the left (if my memory serves me well) and you ask me what I think of this painting. My response should be : superb in my view. No clue how such a thing is done, but then my skills are elsewhere. Btw, I don't like paintings. I don't have any in my home, except for one made by the grilfriend of a college when we were 20, which she made of the Notre Dame, and which painting I lent and never gave back. I love that painting just because it is a painting.
I have many photo's of the Notre Dame as well, but none of them ended up on the wall.

Did I make sense a bit ?

Let's put an end to this because we both made out points, I'd say. But it reminds me of this :
People in countries not accessible to audio auditions (like from amplifiers or speakers) seriously ask for YouTube recordings, judge by means of that and next buy (or not). They do it like this because it won't go otherwise.
I can not imagine this. You ?
Still it is reality and thus makes sense to them.
We are all different and reside in different contexts.

Peter

Hey Peter, I never asked anyone which of the 3 files they preferred. All I asked was which of files 2 or 3 sounds closest to file 1. Nothing more, nothing less. I really thought that that was the beauty of the test - much less subjective than the regular listening tests.

But what you hear in file 3, I do too. It is different from file 1. What is interesting for me is, why? So, here are areas that might be contributing to the differences between file 1 and file 3:

1. the hardware (right from NAS drive to Mach II USB output)
2. W10 14393
3. XX configuration
4. the USB-to-spdif converter
5. the BNC cable
6. Tascam spdif receiver
7. Tacam SD card hardware

4-7 would seem the obvious culprits (quite a lot to go wrong), BUT... they were identical setups for file 2 and file 3. So if there is a flavour being added by 4-7, that flavour should be present in both file 2 and file 3, but not in file 1. I can't hear any such thing. To my ears, file 2 and file 3 are as different from each other as they are from file 1.

I still suspect it's W10 14393 doing something [very pleasant] to the sound... Maybe we'll be able to figure something out as we go forward?

Mani.

Yes, let's.

Mani.

Not really !
Here I thought this about it :


I stroke that through because of my conclusion that no ADC should be anywhere. But :


And


And


So ... this was all a kind of lead in to today's activities, but ...
Of course I believe you if you now say that in this process no ADC is in order, but to state that all is perfectly clear ?
no.
Mind you, I stroke through my own sentence because there should be no ADC anywhere (also see Title of the topic). And you actually debunked my stroking through. So I have been confused all the way, although I never believed any ADC should be in order.
So for today's activities it is very clear now. But for the general setup (this Topic in general) sure not ...

:)

Possibly you are referring to the ADC only because it is a device (en entity) and it is cable of recording digital data just the same ?

Peter

PS: First thing this morning I made a fairly large post in this topic, all about this question(ing). I ditched it because I was too confused.
So not being really clear is an understatement. Haha.

Totally agree.
This is why (instead of the large post) I finally aksed wheter that was the vanilla rip.


Sure. But I can't help myself in changing the subject to the "why" of the #1 being so different then.
Let's not forget (and you seem not to focus on it) that when I listen to #1, I must be listening to the same path of your creation of #3. So all what is in my mind is that I am listening to "double" the #1 path when listening to #3.

Anyway, we have consensus in getting somewhere now.
That is, what we should hunt (and this seems not microRendu stuff if you ask me). We miss something ...

Peter

OK, here's the mistake I made. I should indeed have started a new thread for this because it has nothing to do with testing different USB cables, or testing different DACs. It is about testing two different digital transports (the microRendu vs. the Mach II).


Yes, I'm very aware of this.

This is the way I look at it. If the path adds, say, 5% 'distortion', then when listening to file 1, there would be 5% distortion and with file 3, 10% distortion (assuming they're perfectly additive). With 1% path distortion, file 3 has 2% on listening. With 0.1%, file 3 has 0.2%. Etc, etc.

But if the path is truly transparent (0% distortion), files 1 and 3 would sound totally identical.

So the question I was trying to answer is, which digital transport (the microRendu or the Mach II) gets closest to the holy grail of true transparency?

Mani.

With the #2 (microRendu) it is very obvious that the "colouring" (but translate to give all more body) is caused by distortion. It sounds more "ugly" (I recall that you said similar originally, but already yesterday I could not find that back). Say that the refinement gets lost in the process.
Sadly, with #3 something similar must be going on, but nothing at all points towards distortion. It coincidentally must make right what lacks in the original.
But see ? this is nothing for a thesus. There's no logic in it, nothing. It must thus be regarded as coincidental.

If anything, I would say that the original is smeared and more body is the effect of that - though not for real. So I'd like to think that the body remains the same but that the top end is smeared and thus stops to be profound.

Now we could start thinking whether something like the Mach II is doing such a thing in the first place. So never mind it should not be possible, everything matters and this does too;
we have more bass/body, we have less unrealistic highs, we have more backness (this latter certainly is so in my view) and do that once again and all doubles.
We need two cascaded Mach II's.

By now I think it (indeed) would be best to start out with something that is "better" than the originally uploaded test track. So what I'm hoping for is that I can hear how #3 degrades compared to #1 instead of perceivedbly improves.
That we on a side note (ahum) will also still be able to judge what the mR does in the chain, is a bonus. But I hope that the differences will be more profound *and* that nothing will go against my logic - I really can't deal with that, psychologically.

Btw, I now see that with the not posting of my longer post from this morning, this text was eliminated too and that is a shame :

---

This very much looks like me today being quite satisfied about many recordings regarding the reality of the drums, while not so long a go there was one measure only : my own recording of the drumkit. This puzzles me. So my own recording has been at 100% right form the start (say 5 years ago) and played back through NOS1 and all from that time, the remainder of recordings (and throughout all my many albums) being at 20-30% or so, while today I wonder whether there's much difference between my own recording and many other albums. One thing is clear : playback (SQ) improved significally since, and my own recording was already at 100% back then (so could not improve).

---

?

Hi Mani,

For what it is worth given the conversation that has followed since I downloaded and listened to those three files, I can certainly hear what yourself and Peter are talking about.  The latter two files are quite different dynamically to the original file, and second in particular is very flat and fuzzy, the kind of fuzzy that I would expect from a cheap dac.  Anyway, the conversation in this thread has moved on but I am still interested to see where this all leads.

Cheers,

Anthony

Mani, a question which may appear to be totally irrelevant :

Do you know - or can you see how often those ultra capacitors are beging (re)charged ?
(for others, this is about the mR hence #2 and its by Mani used power supply).

Also, do you take care that the mR PSU is out of the mains when you do all these tests (except for when the mR itself in is under test of course) ? ... this is crucial ...

Thanks,
Peter

I'll be posting the output from Peter's CD rip a bit later on (once I've done it). Should be interesting.

Mani.

It's on the CA site somewhere - will have a look later.


Aha! No I didn't. The Mach II was switched off for the mR test, but the mR was still on for the Mach II test. This time, I will make certain only one is on at the time it is being tested.

Mani.

OK, here's what John Swenson wrote on CA:


So, a very, very low frequency.

Mani.

Yes, but it is not about that in the end (I implied it though). It is about the charging time. IIRC the maximum capacity is 2.5A ? Then it could be read something like this :

4 (seconds) / 2.5(A vs 1A) = 1.6 seconds. So for its max capacity it switches each 1.6 seconds, which also is the maximum capacity (per bank) and thus this should imply continuous charging.
I know, I know, this won't be 100% correct because the capacity as such needs not to be equal to the charging capacity, but assumed all has been made in some balance, it will be close to equal.
So :

This means that for a device which draws 2.5A (at 12V I suppose, but this is not relevant) a continous load will be on the mains, which is fine. But with, say 250mA (could be the mR) there's an on/off switching each 10 seconds, where the loading (charging) takes 1.6 seconds again and which will go at the highest current (because the bank will want to charge fast).
I don't regard this to be good because of this changing. But never mind. Point should be (assumed I am a bit correct) that when it is used for real (like when feeding the mR) it will be OK as it is because it is there for a reason. But if it is not used (no music plays but all is under power anyway) then it shouldn't give the "opportunity" to disturb.

Side note (or not) : When I design stuff (or try to :)) this is foremost about capacitors, which are now inside of the device and while there for their good various reasons, they disturb just the same (because they charge at "frequency"). This is also why the Mach II has only a small capacitance on its (linear) PSU instead of what you might expect.
Ultracaps are just capacitors hence they charge about as fast (but last like a battery). So in this case (LPS-1) they are exernal and will not be protected by a buffer which normally comprises of ... capacitors (say that these to some degree normally protect from the mains to be 1:1 discharged by the high current spikes a device (like power amp) may imply.
Of course nobody looks at it that way, and this is only looked from the 180 degree other angle : you need capacitors because the mains is not fast enough and also is too far away (which would be true). But also think what it takes for your *other* devices when a very large cap (and ultracaps can be seen as that) need to be charged and which goes by infinitely higher spikes than what your amplifier will ever ask for. Thus, amp asks 10 times for current before the cap recharges, which thus goes by 10 times the current the amp asks for. And might this be about 1:1 then the (buffer cap) design is not right. So ... law.

That was all. But interesting in itself I think. And oh, this is nothing against ultracapacitor power supplies and most certainly not against the LPS-1; my remarks are just general and my own ideas about matters.
Peter

Don't you assume that there is some equivalency in sound between Roon-HQ/Player and fully tweaked/RAM-OS optimized XXHighend player? If there is no equivalency (I'll bet the family farm there is No Way Roon/HQ sounds the same as XX on your Mach II) how can a valid comparison be made? Wouldn't it be better to play all (microRendu and Mach II) with identical HQ player settings?

OK, here we go with the Konbanwa track...

1. Peter's original file (flac):

https://drive.google.com/open?id=0B82Qx9VwoovtUGExZWNrVEZfbDA

2. microRendu digital out (wav):

https://drive.google.com/open?id=0B0PU5LO5jVjfNGxheGZFVXFEd1U

https://drive.google.com/open?id=0B0PU5LO5jVjfUUFxcGs3TDFlclU

3. Mach II digital out (wav):

https://drive.google.com/open?id=0B0PU5LO5jVjfdVBwTWlYVDE3ZGs

https://drive.google.com/open?id=0B0PU5LO5jVjfU1NkMXhWZTVUd1k

I won't say anything at this point, other than the shortcomings of digital path are much easier to identify with this track.

So, which file sounds the closest to the original - file 2 or file 3?

Mani.

I'm sure you're right. But the point isn't which hardware is better, the point is which digital transport is better, and that includes the software. I've chosen what is considered the best software in each case.

So what we're really comparing is the 'best' microRendu setup vs. the 'best' Mach II setup. I think this is the most interesting comparison, certainly to me.

The whole purpose of this exercise, for me, is to determine whether I can live with the microRendu setup in my office system, or whether I need to set another Mach II up in my office. At the moment, it seems the latter is going to win hands down.

And yet, over at CA, people are convinced that the microRendu/LPS-1 is absolutely as good as it gets...

Mani.

Okay, I thought you were just trying to determine which transport was better. Once you throw in the software one is left not knowing whether the decrease in SQ with the Rendu is to to the hardware or the software. Would be great if a way could be figured to load XXHighEnd on the Rendu. Or, if you just wanted to judge the transport and not the software, to run HQ player on both devices.

Brian, heads up. You posted something with sense even I understand. :swoon:

Not that I am going to say very different things, but if we had to compare XXHighEnd to HQPlayer then it would be clear how much of apples and oranges that is. Thus, I did not think of it really, but the to me clear distortion I hear, is that perhaps from HQPlayer ?
Why not. At least that seems 10x more plausible to me than a digital interface which turns out to molest.

No no, I am of course not saying that digital interfaces can't matter a thing, but while that is hard to grasp (how it can matter) it is totally obvious how XXHighEnd vs HQPlayet just will. It is made for that explicitly, so to speak.

To balance things a bit, I don't think that what we hear from #2 is HQPlayer, *unless* that's converted to DSD (no idea whether it can on the mR).

Peter

If I wanted to compare XXHighEnd to HQPlayer, I would simply load HQPlayer onto the Mach II and do that. Why bring the mR into it?

As I said, my interest is in the 'best' mR playback vs. the 'best' Mach II playback, to determine which to go for in my office system. For the latter, I think Unattended is an absolute necessity.

Mani.

Just to reiterate, I'm outputting straight 16/44.1, with no DSP whatsoever in both HQP and XX. Even then, I know they will sound totally different, even if they were both on the same machine.

Mani.

Oh happy Day!  :)

What? No thoughts yet on whether file 2 or file 3 is the closest to the original???

Mani.

Uhm, I am still working over here. :)
After that the system has to warm up for some 30 minutes.

Aha. OK. I missed that.
Or forgot aready. :old:

First round - cold system, 1st minute only :

#1. With the bass from yesterday in mind - there's a bass.
#2. But where is that bass here ??
#3. Starts with a tick. Not sre whether this sounds fuller (the fullest of them). But it looks like it.

Again #1 is the least dynamic one. But today I name this "more black".

Now steam up the system for a while ...
Use 15 minutes of Safri Brothers to unleash the drum kit hiding in the system (you'd call that Dance);
Use 15 minutes of Saafi Brothers (:yes:) to wake up the woofers (I call that Ambient).

Now start playing the last two tracks of the The End of the Summer album and let go fluently into its track number 2 of Mani's upload which is the #1 test, followed by the #2 and #3.
All written live as it happens over here.

The album itself (last two tracks) :
In the middle of the first track I receive the message from the catacombes overhere that this is "not music". OK, accepted; too much of modern Jazz I guess.
But I hold on.

When the real #1 begins all kinds of discussions start to happen and all I can do is regard the music background music.
It is okay for that.

#2 is not though, it interferes too much.
But why the heck is the bass (well, lower mid) is the disturbing factor ?

#3, strangely or maybe better : unexpectedly, is the best background music. It simply is (also) the best musically exhibiting. It is just flowing ....

Makes sense ?
Logically not. But is happens to be so.

Peter

I redid files 2 and 3 - I found a way of improving the performance of the Tascam. To my ears, the new files sound more musical. So here they are...

1. Peter's original file (flac):

https://drive.google.com/open?id=0B82Qx9VwoovtUGExZWNrVEZfbDA

2. microRendu digital out (wav):

https://drive.google.com/open?id=0B0PU5LO5jVjfUUFxcGs3TDFlclU

3. Mach II digital out (wav):

https://drive.google.com/open?id=0B0PU5LO5jVjfU1NkMXhWZTVUd1k

Mani.

Well, I should start 'remastering' all my collection with my USB-to-spdif converter and Tascam... and sell my 'remasters' for millions ;)

Over here, #1 still sounds the best... but only just. But with my newly redone files (see above), #3 is now very much in the same ball mark as #1. #2 remains flat and lacking to my ears.

And just in case anyone is concerned - there's absolutely no trickery going on here - everything I've said is exactly as it is:
#1 really is the original
#2 really is the digital output from the microRendu
#3 really is the digital output from the Mach II

I'm surprised more people haven't chimed in...

Mani.

I've been asleep.  Just downloaded the new files and started the system, so I will play one album for warmup and then get into the test.

So, I have listened.

The original CD Rip is clearly the best sounding here, reasonably closely followed by the third through the Mach II where there is some loss of clarity and perhaps a touch of dynamics.  The second file through the Microrendu is flat, thick and unresolved.  The transients are flattened, the soundstage is less deep and it is what I would call a large departure from the CD rip. 

Most notable is the piano rendition of the Microrendu: the touch of the hammer is muted and dull compared to the other two versions, and the notes are quickly truncated and fail to hang due to the lower resolution and smearing of other instruments and notes.

Saying that, I don't think the Mach II is perfect either, but it is much, much closer to the original CD rip, and remember that I am not playing back using a Mach II.

Thanks Anthony. You're findings pretty much echo mine, though I have to say the sound of #3 grows on me the more I listen to it. It's certainly not totally identical to #1, but deviates from it in quite a nice way.

Mani.

The difference could be in the Tascam unit as well...but I prefer the original CD rip.

What a total mind f**k...

I posted the 3 files on Audio Science Review. Ray Dunzl did some analysis on them and found that all 3 are absolutely identical (with a few samples added to #3 - I could hear the clicks but couldn't get rid of them).

And having seen his null results, I'm now having problems hearing any differences between the files!!!

Mani.

I would have been surprised if the files were anything but "bit perfect" so to speak, or "identical"...these sorts of conversions are well sorted in the modern world. 

As to what you hear when your mind has new information to digest then...well...that is quite well documented.

So Anthony, what were we hearing when we said things like:




We've just established that the 3 files were identical (a few added clicks in #3). There simply could not be any sonic differences between them, right? BUT... we heard differences. All in our minds?

Mani.

At least I was lucky to be so ignorant to judge the #3 as the best sounding, even after knowing (finally) that the #1 was the original.

Btw, I did not expect the files to be different. I mean, not one single moment I even considered it. IOW :
Make two of these recording from your best (or worse) setup and, for example, have the SFS different between the two.
Now too, I wouldn't even consider that the files could be different. They just ARE the same.

So don't worry. :)

Peter

For me, it's a reminder that remaining objective is bloody difficult.

Mani.

My main point was that something would be wrong if the files were not the same.  For instance, as Peter said above, we can influence the sound in XXHE by changing this dial and that one and minimising the OS but in all cases the bits remain the same, they are not changed.  How do we change the sound with those dials?  By changing the behaviour of the cpu and other parts of the computer and altering how the transport (MR or Mach II in this case) interacts with the DAC. 

Yes I agreee. XX can only influence the noise getting to the DAC - there is no possible other explanation for why any of the XX settings change the sound.

But back to this 'test'...

Why did you, Peter and me hear a difference between the 3 files... when they were absolutely identical?

Mani.

Why do you hear a difference when you change XX dials ?

Actually there is only one thing that worries me : that I judged #3 to be the better one (over #1).
Oh, I can live with it all right, but if I am correct with that judgement (notice that Anthony already does not agree), I am eager to apply a similar improvement without the hoopla of recording the digital file (output by the Stealth) digitally first.
haha

Regards,
Peter

And this is exactly my point - you were not correct with that judgement! It's an absolute impossibility... The 3 files were identical!!!

You, Anthony and I were hearing differences that were simply not there.

Mani.

I'm going to redo the test, but this time through an DAC and then ADC. The path will be:

microRendu (no DSP) or Mach II (no DSP) -> Chord 2Qute DAC -> Prism AD124 ADC -> Tascam spdif digital in

In this case, even though the mR and the Mach II are producing bit-perfect outputs (as proven by Ray on ASR), we certainly should expect to have a difference in SQ. We'll see...

Oh and BTW, this time I am not going to reveal which captured file is which! Let's get any listening biases out of the way.

Mani.

Are you now serious ?
If so, you may have lost it.

:yes:

(referring to your one but last post)

OK, I did myself. :swoon:

Of course I'm serious!

#2 vs. #1 nulls to zero.
#3 vs. #1 nulls to zero (if clicks ignored).

On conversion in the DAC, there can be no difference between the 3 files. It would be like loading file #1 three times and hearing a difference between them, surely?

Peter, please enlighten me - where am I going wrong with my thinking?

Mani.

No-where so far. The mistake I made today was thinking about the offsets and difference with filtering (you will remember from CA-Dennis). But then I saw your "no DSP" in you later post and it reminded me about the not-upsampling you applied. So there can't be a filter and thus also no difference implied by that PLUS that it is all digital to begin with and no difference is no difference.

Did I hear the same/similar differences in both the files (your example track and mine) or do I suffer from amnesia ?

Peter

Yep! This is what you said about my original files:


So, although these have not been analyzed, we can assume that they are indeed identical too.

Mani.

I assume you supplied the latest incarnation of the files, right ?

Anyway, I am pretty sure they did not compare the files as should. And you know what ? I have massive problems with doing it myself this time. The first time this is so !

Working on it ...

Here we go!

1. Original CD rip (converted to wav, for those not using XX):

https://drive.google.com/open?id=0B0PU5LO5jVjfTUdrMGtRRkZuZkk

2. Digital transport A (microRendu or Mach II):

https://drive.google.com/open?id=0B0PU5LO5jVjfM0paUHFfRVppZHM

https://drive.google.com/open?id=0B0PU5LO5jVjfQ1JYdk9mVWtEbHM

3. Digital transport B (microRendu or Mach II):

https://drive.google.com/open?id=0B0PU5LO5jVjfSGxOVE9BWmR2aWs

https://drive.google.com/open?id=0B0PU5LO5jVjfMEF3eURHVkU1ZWM

[Edit: Actually there was no need to redo these - they were fine in the first place. I've added new links.]

Of course, these 3 files will not sound identical. The original file (#1) is going through the Chord 2Qute DAC and then through a Pass Labs X1 preamp (to convert the Chord's SE output to balanced) and then through the Prism AD124 ADC (balanced input only). But nevertheless, we should be able to determine which digital transport, the mircroRendu or the Mach II, is doing the least harm.

So, which file, #2 or #3, sounds closest to #1?

Mani.

Yes, the files I supplied were of 'your' Konbanwa track.

Here is are the null results: http://audiosciencereview.com/forum/index.php?threads/sonore-microrendu-hardware-tear-down-review.770/ (scroll down towards bottom).

Mani.

Of which you supplied two versions ...

I have seen that thread over there. This is why I say that they did not compare as should. Ok ?

I wanted to be totally consistent and upload 3 files of the same format to ASR, and not 1 flac + 2 wav. Not everyone uses XX, and I didn't know whether other software players would treat flac and wav identically (as I know XX does).

So, I converted your flac to wav with dBpoweramp Converter. I used this wav file to make the digital recordings, #2 and #3. And this is somehow invalid? Again, enlighten me...

Mani.

Don't make yourself crazy. Did I say or even imply somewhere that something is invalid ?

Don't you see that I am comparing myself and that I thus need to know what you supplied those guys ?
Btw, I could have downloaded it from there of course, but I didn't (I already had the lot). And, aligning the files takes an hour or so.

Anyway, start putting yout Tascam, mR or Stealth to eBay. Or all of it. Something is not right and nothing is equal anywhere.

Satisfied ?
I am.
Btw I am not going to listen to the latest test. I have fun as it is. :) :)

But seriously ... for you, start thinking what can be wrong with the latest (is 2nd) version of my track you put up. And notice it can even be FLAC. The differences are always 1 or 2 decimal (on 65536 possible values) - more than 8 million of them.



What I like you to do is make a second take of the mR output. So, just 100% the very same as you did last time with my track and also with the very same source file (if you converted it to WAV first, then again do that os use the WAV if you still have it). Thus, changing nothing compared to the previsous time !
If that resulting one is different from the one I have here (lead-in and -out differences trimmed) then it has to be your Tascam.

Peter

PS: ;) I thought you had learned by now that you shouldn't give random people some files to compare. Because remember, those people are the same who prove that XXHighEnd sounds the same as JRiver. They just don't know how to do it (apparently).

PPS: No I am not deaf. I also can't be placeboed. But you can ! hahaha

I am not sure what the options are, but let's start with these :

1. FLAC encoding/decoding is not lossless. And keep in mind, I just used the literal XXHighEnd method for the decoding (I grabbed the UnicodeTrack file coming from it).

2. The Tascam fails on us.

3. The digital output of the two means (mR and Stealth) are not bit perfect. Or at least one of them.

4. The digital output of XXHighEnd is not bit perfect. Btw, think of it ... you can avoid upsampling all right, but I don't think you can avoid the bit depth going from 16 bit to 32 bit. And then what ? then something must be rounding, because the files are 16 bits all right.
Ha !

Please add other possibilities (anyone is invited).

I did not take the #3 file yet, because I thought that the tick in the beginning could bother me. And it still can, as the other file(s) has been trimmed already and possibly it needs more of that to avoid (skip over) the tick.
Anyway, interpret 4. above carefully because I imply that XXHighEnd's output can not be compared ... but I did not yet. But with the mR the same can be going on.

Peter

I'm sure you picked this up, but I just wanted to be absolutely certain...

There are clicks/ticks on file #3, but not on #2. Their frequency varies with varying SFS. Ray has referred to them as 'stuffed zero samples' - there seem to be 7 in total over the 4min 36 sec track. These have increased the length of #3 by 7 samples over #2, and obviously need to be taken into account in the nulling process.

Sorry if this is totally obvious to you, but I wanted to be sure.

Mani.

I have no idea what the difference between your flac and my wav files of #1 could be. More than 8 million of them? How has this happened in the flac-to-wav conversion?


Will do.

Mani.

OK. I don't think it will be really needed to re-do the mR recording as I suggested; this was before I found the rounding probable problem (though in the same post).

I have been looking at it again, and while I now found other means for better proof to myself, I could also find back the differences I had a hard time with (as mentioned, for the first time).

Thinking further ... it is hard to dictate this all to rounding errors, with also the knowldege that expaning to 32 bits without upsampling, should just add (padd with) zeroes. Also I am not even sure any more how XXHighEnd deals with this these days (or better : 5 years ago, because it must have been that long that I looked into that for the last time).
I also come to this "conclusion" because looking at it in the way I want(ed all the time today) I am reminded by the dithering XP always did and it just looks like that. So ... maybe you forgot to switch off Dithering in HQP ?? (I suppose it can do that at 16/44.1 just the same)
XXHighEnd does not dither.

Below you see the difference file of it. This is always 1 decimal of difference. And yes, earlier on I said "1 or 2" but I think I misread the hexa-decimal data somewhere.

Peter

I will now do the #3 file as well; I think this will be useful for conclusions, especially because I'm on the path of Dither now.

No definitely no dithering in HQPlayer.

Mani.

I set DAC to 16 bits in XX. This is the only setting that would work with my 16/44.1-only capable Legato USB-to-spdif converter. The 'DAC is 24 bits' and 'DAC is 32 bits' buttons became grayed out once I selected 'DAC is 16 bits'. I therefore assume there was no padding going on.

Mani.

Mani, correct. But I just forgot whether the NOS1 plays at that setting. Uhm ... which you did not use of course. :wacko:
So indeed, with that setting it is and remains 16 bits. So, good.

Peter

The results of the comparison between #1 and #3 are also in : looks as the same "culprit". Btw, up to second 21 because there a hiccup occurs and all is really different.

Now compare #2 with #3 ...

Interesting ...

#1 vs #2 and deviations are all donwards (see earlier screenshot).

#1 vs #3, same story.

#2 vs #3 and they are both ways ... See below (dashed line is how it should be when no difference at all).

Now I must think. :smirk:

Ah, luckily I saw that wrongly !

#1 vs #3 is upward.

But I still must think.

Firstly I like to draw conclusions audible wise :

So, #1 is as is; it is the original.

#2 deviates downwards compared to #1.
This kind of puts pits in the audio wave.

#3 deviates upwards compared to #1.
This sort of puts spikes on the audio wave.

I like spikes better than pits.
But I also like spikes better than smooth (which we can dedicate #1).

And, when I would still be correct that #3 sounds better than #1, then "positive dither" (haha) would be the solution. "Negative dither" sounds like a drag.

Btw too bad that no contenders are to be seen anywhere (except for us 3); they could help improving SQ. But alas, everybody already has the best SQ I reckon. :tongue2:

I suppose the number one suspect right now is the Tascam, right? But I also wanted to just throw in another possibility...

I've been using an Intona throughout. There are some very weird things going on with Intonas. For example, many people have reported Intonas not working with certain DACs, or with certain USB cables. Some people report having to use a different length of USB cable to get the Intona to work. Well, I discovered something potentially 'serious' today...

The Intona is massively affected by its proximity to an EMI source. If I move the Intona close to the Mach II's power cable, it will lose connectivity with the DAC. Moving it a few cm away, and the connectivity comes back. This is not because of loose USB connections - they are very tight on the Industrial Intona.

Could the Intona be the culprit?

Mani.

I know you said you didn't need this, but here it is again anyway...

2. Konbanwa (wav file) _ microRendu digital out v2:

https://drive.google.com/open?id=0B0PU5LO5jVjfMlY1dlBvZXBrUTA

Does this still exhibit just the downwards deviations from #1?

Mani.

Edit: I just discovered Tascam's auto record feature today - the recording starts once a signal of -54dB has been detected. Hopefully, this will make alignment easier?

Ah, I was just contemplating to ask for help from you. Good ...

Nah, I think we must focus on some functional process (like Dither is that) because of the always deviating of 1 decimal. Btw, the comparison between #2 and #3 deviate 2 decimal if we are allowed to think like that. So, it is not deviating 2 in one go, but the maximum deviation over the file (OK, the first 21 seconds) is 2.

When hardware errors are in order, the deviation will be anything, randomly.

Indeed, the Tascam seems suspicious now. But why would the one "recording" (#2) deviate the other way from the other (#3) ?

:scratching:

No, it won't. Actually it will get worse because it will start too late (and my trims as of now will be useless).

But I am fine by now. I know the whole file visually by heart already.
Not. :)
But I can do it in a few minutes now if only the track is the same.

Peter

OK, manual start/stop:

2. Konbanwa (wav file) _ microRendu digital out v3:

https://drive.google.com/open?id=0B0PU5LO5jVjfOU5mYnlsMkNRTlk

Mani.

Thanks. After 15 minutes I had to give up ... and now I know the reason.

Hey Peter, when you say, "there's 1 decimal of difference", what exactly does that mean? I mean, how big is this deviation, in other terms I might be able to better relate to?

Mani.

1 decimal is just the 1 you count with and it is the smallest difference the level in audio can make. Mind you, for 16 bits, where the total number possible is 65536 (from full plus to full minus).

So the difference - when occurring - is 1/65536 and would be something like 96dB down.

The difference between the original mR recording and the v3 version ...

Plus and Minus deviation.

No, I don't get it. :scare:

Thanks.

Mani.

I think we can blame the Tascam with a high degree of certainty, no?

Even if the Tascam is to 'blame' for these deviations, surely the error is small enough to allow its use in the DAC/ADC chain that I posted earlier today?

Mani.

With the danger of me getting crazy of it, I double checked the both mR files against the original. Both deviate downwards.

How can the two mR files against each other deviate both upwards (positive) and downwards (negative) ?

There must be something I am missing myself now in the realm of the first file I load is taken for reference, the second uses. Earlier on I did not see that (so possibly I have a post wrong about "downwards" or "upwards") but with a result deviating both ways it seems hard to do something wrong.
But if I can't make the results logically / consistent, something else may be wrong (as well). But what ?

Maybe not ?
Unless I am deaf after all, the difference was very easy to hear. So I'd say that this makes everything illegal.

And whether it is the Tascam ? that seems obvious, but then I first like to understand my findings.

Peter

You first thought the deviations in both #2 and #3 were going downwards. Something made you change your mind about those in #3. Are you sure that whatever it was that made you change your mind was absolutely valid?

Mani.

Yes. So that too will have been caused by which file was loaded first (in the analysis of two). So do that a next time and swap the both and it will show the other way around.
If course I did not know this, as it is the very first time such a deviation to one side appears on my screen.

Btw for the same reason at first I could not see it, because the screen resolution is actually too low to see such a one-side deviation normally (two is fine, and that I am used to (dither). So I had to amplify the differences to make them appear (the analysis telling me 8.9M mismatches to be there anyway).
What I showed (screenshots) is no amplified by 24dB.

Anyway, this does not explain how the difference of the two mR recordings is both ways, up and down. However, it could be explained by the Tascam (or something else) deviating to one side at random (and once it deviates to one side, it keeps on doing that).

Hey ... maybe your recording is not bit perfect "officially". I mean, supposed you do this recording via Windows, then that recording too can be done in Exclusive Mode or Shared Mode. And what I recall is that you can't play and record both in Exclusive Mode. So or example, play back via XXHighEnd (always Exclusive Mode) and record that on the same PC (WASAPI) can't work. Not that you have been doing that (I'd say) but only to get the idea of what can be going on.
Whether that implies what we are seeing ... I can't tell. Anyway dither should deviate both ways.

Maybe I will dream about it and come up tomorrrow with what we're missing.

Regards,
Peter

Thanks for doing all this Peter - I know it can't be of much interest to anyone else, but it certainly is to me.

Sweet dreams ;)

Mani.

I have a camp here and wake up and there are three pages to read in the morning.  I can see that Peter just lost most of his day to this, but I can also see that this might lead somewhere as well if the results can be understood.

Right below we have the original and each of the samples is 5 (first row).
On the second row we have recorded it, and some are lower than the orginal (4). The deviation shows on the 3rd row.

5	5	5	5	5	5
5	4	5	4	4	5
0	-	0	-	-	0

We make another recording which you see on the second row.
On the first row we put the first recording (first row here is the same as the second row above).
The deviation is shown on the third row again.

5	4	5	4	4	5
4	5	4	4	5	4
-	+	-	0	+	-

And we see how the plus and minus appear.

Peter

Haha, of course!

Mani.

So, any thoughts on why #2 vs. #1 is all downwards and #3 vs. #1 is all upwards? (BTW, the guy over at ASR believes both are downwards.)

Another guy at ASR has taken a listen to the files and agrees with my initial subjective assessment of how #2 or #3 differ in sound. He's also listened to #8 and #9 (through DAC/ADC chain) and can hear differences there too.

Mani.

Mani,

#1 vs #3 is downward too. And as I said in one of the last posts yesterday : here or there I can have that judged that wrongly because it depends on which you load first (which becomes implicitly the reference and I did not know that).

But also see below ... at (my) first glance this goes upward. But it really does not ... I am tricked by the long beginning being down of the upper channel.

Sorry for the wider screenshot.

Peter

Mani, what about this ?
So where does your Tascam recording software run ?

Thanks,
Peter

The Tascam is a stand-alone unit with its inbuilt software - no connection to any PC. I record straight to an SD card in the Tascam. I then take the SD card out of the Tascam and transfer the files to my PC.

I have an old CD/DVD player here with an spdif output. Later, I will take a capture and then we can compare to the CD rip.

Mani.

Hey Peter, I had two objectives for doing these tests:
1) Figure out how good my current ADC is - it's a 20 year old unit (older than my previous PMII), and I needed to know whether it's up to the job of digitizing my vinyl
2) Figure out which digital transport to use in my office

I now have the answers to both of these and I'm happy to put this whole endeavour to bed. But let me know if you'd like me to do anything else with the Tascam.

Thanks again for all your effort.

Mani.

Late reply for what its worth nothing new but


I agree with this statement.

I listened last evening and through my setup, preferred original #1.

Found other 2 slightly coloured(as you would expect) preferring #3 MachII over #2 for sound.

Would have to say my sound has never been better with current XX settings. Someone is suggesting on CA, CD is still better. Obviously hasn't got XXhighend.

Robert