John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
Member
Joined 2002
Paid Member
I suspect vinyl will outlast CD, most of us could build a victrola like turntable but not a CD player. I had to search a bit for my wife’s last laptop to find a DVD drive model and windows 10 now comes on a USB stick, the pig won’t fit on a DVD.
My new car can at least play wave files from a stick but I do miss my 68 Olds convertible especially the back seat.
 
You are obviously frustrated that you are almost alone in your opinion on the Oohashi experiment.

No, being alone in my opinion is not frustrating, reading criticism that is based on flat wrong premises but nevertheless repeated despite any correction effort is a bit.
Otoh, as I am still interested in dynamics of human interaction generally and in its internet variant, nearly everything provides additional information.

Well, partly you were right that I didn't read the full report because when input is of dubious quality, the outcome of the experiment will have its flaw's also.
In other words, Rubish In - Rubish Out.

Who could disagree, but the bottom line is that no perfect experiment exists (most probably) and so we can't use the shere existence of flaws/ambiguity/unclarity as reason to dismiss any result per se; instead we have to evaluate what is reasonable and what will likely skew the results.

<snip>
Had they spent just as much time at the input, the experiment would have gained quite a bit in trustworthiness.

Speculation presented as fact?! ;) (see below)

<snip>
However Ooashi used 1.92 Mhz, causing his noise spectrum to shift even further to the left.
Conclusion is that with the used 1.92Mhz DSD, noise at 100Khz will be at around -40dB.
But looking at his Gamelan spectra, Image 4, I see no noise at 100Khz or below at all, how's that possible ?
That can only be when these spectra were made before digitizing and therefore do not represent the spectra as offered to the test candidates IMO.

Thanks for the argument. I shared that reservation -not right from the beginning but after I saw a couple of years later the measurements you've linked - and read the publications again (including Yamasaki's contribution to the invention of the 1 bit process, I'm shure it was a team effort, but afaik his ideas) and both parties (Yamasaki) and the other authors emphasized the advantage of the pre-emphasis and de-emphasis process used wrt to the reduction of noise. As at least Yamasaki is IMO a trustworthy source I tend to accept it, but of course I'd prefer to have a full set of measurement of the whole chain they were using.

Especially today when the internet provides unlimited space for supplemental informations that can't be pusblished in a journal for various reasons.

These filters have been analogue filters with 170dB ? and 80 dB/oct slopes. What was their FR and GD, not documented !!
He mentioned a flat response to 100Khz, but not where this was measured, at the output of the D/A or via a microphone at the position of the listener.
But in either case the FR of the filters and their GD is missing.
On top of that, FR from the speakers could very well be flat with both LPF and HPF in action, but that does not at all mean a flat response from both chains individually around crossover. That's why all these FR's and GD's are important.

Now we are talking!
Again he measurements aren't surely detailed enough and "flat" wrt the speakers might be a bit too optimistic when considering the response curve the authors showed in their convention paper from 1991; the speaker system used later seems to be more refined, but the of course the question remains in which respect that could have skewed the results?
IMO it's the same wrt GD.

Surely an interference effect occurs in the crossover region given the short wavelength in that frequency range, which is still one of my concerns.
But looking at the whole picture given by all the test results in their experiment it seems unlikely that any interference was the reason.

And then my firm objection that when listening to the LPF version, the HPF chain was switched of and v.v......

Let's see, presenting the HFC alone (via the supertweeter path) provides a control and in that condition any GD concern is moot or would you argue differently?
Why should the "LF" path be active too in this condition?

And if the "LF" path is active to provide a baseline for this "normal" condition, the impact of the filter GD would show up, or do you think differently?
Why should the "HF" path be active too in this condition and what should it carry?

..... brings me back to the point that I have already mentioned a few times: when comparing things, reduce the amount of variables to the minimum.

While it generally is a good consideration, any argument (haven't I asked that before?) why these experimental setup would skew the results is still missing.

Playing these files over the same chain without switching Speakers on and off, would have resulted in a much more creditable experiment.
One speaker system, whether partly or complete active going to 100Khz should have been used all the time without any switching in between.

Would be an easier experimental process but still does not invalidate the other approach, unless there are arguments.
I've to search, but have a faint memory that the experimenters in one of the follow ups later used 192/24, but a full range covering the whole audio band and in addition the ultrasonics up to 100 kHz?

There once was a fullrange prototype using plasma for everything but .... :)
 

I think Bill was referring to this...

Due to the offset between the cartridge's axis (which is approximately tangential to the disc) and the tonearm's pivot, the force applied (through friction) by the rotating disc to the cartridge tends to draw the tonearm toward the center of the record and distort the balance of the sound and of the wear suffered by the stylus and the vinyl groove.

No mention of direction of radial travel. Also the experiment is easy, put a stylus down on a blank LP.
 
Member
Joined 2002
Paid Member
Talking about preferred bit depth in recordings.
This recording (*) has a hefty 70 dB difference btn lowest and highest music level and if it was recorded without clipping it would have at least 3dB more.
Right at the beginning, during the first second, the fagotto changes from F to B#. Adjust the playback volume so that you can just perceive this change in note. This is the lowest music level. Now, listen the whole musical track at this volume setting.
What do you think, with the achievable dynamic range of the playback audio chain, the ambient noise level at your listening space and your highest allowed playback music level , would a higher bit depth recording be potentially advantageous?

(*) EAC wav rip: 2nd CD, 12th track from DECCA 467 314-2 (2001,DDD)
Dropbox - 12 2. Alles Vergangliche (Chorus mysticus).wav - Simplify your life

George
 
vlPkCIx.png
 
Status
Not open for further replies.