John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
I agree that some seem to be completely content with MP3 though I can't listen to that format myself, it makes me cringe. And many of us agree that there are good and bad recordings on cd, no matter what you can do to your equipment, there are just clear differences in quality from one recording to another, but that was always the same in any format, even vinyl has always been that way. We can't control the recording process or the recording engineer and have to select the music that we like and those examples that come up to our personal quality levels.

For you Frank,
I know you will chime in here so I am going to beat you to it. I know you will say that you can take a bad cd and make it sound better on some iteration of a system that you have at one time put together, but I say there are times when I don't care what system some cd's are played on they are just poorly mastered and you cannot fix that. Solder all the connections you want, a poor recording is just that and most people would agree that something just sounds wrong with certain material, it is not system related.
Not quite quick enough! :)

I have no problem with accepting that some material is badly mastered, badly transferred, badly recorded originally. However, most so-called high end, or pro systems then proceed to make a total dog's breakfast of the material, it's impossible to listen to. I'm interested in the core qualities, that which is often easily identified on a simple kitchen radio, and I have a means of retaining that in the listening, without the further "colourations" of a supposedly superior playback system.

I have zero interest in the often heard valve "treatment", where a spoonful of honey is mixed in with the roast meast, the pasta, the fish soup - and sometimes appearing to have a sliding, limited dynamic range filter, discarding a great deal of low level information.

What may seem strange is that I have little problem with MP3 - the real quality issues arise at the time of playback, and can be circumvented ...
 
diyAudio Member RIP
Joined 2005
Both statements are incorrect.

Detectability does not require that the percept be "loudness", just that it is "different". Detecting a level difference is about 0.75 to 1 dB. This is true over a wide variety of conditions (there is an exception at lower overall levels with certain bandwidth - but for our purposes that is a rare occurence). For the "balance shift" the jnd is still about the same (0.75 to 1 dB).

This is a robust finding, in spite of what someone may have written in a wiki contribution.
Cite?
 
Recording at 24/192 and reducing to 16/44.1 results in loss of data. So even if 16/44.1 is perfect for reproduction :) mastering errors can affect the final product.
SY will shudder in disgust, but I agree with him 100% on RB being transparent. I have gone through the exercise of taking a true hi-res recording, from one of the sites that spouts their virtues, checked that there was real information above 22kHz, downsampled to 16/44.1, upsampled that again to the original hi-res format, and diff'ed the two hi-res versions. There is nothing there except a few short bursts of ultra-sonic - real sound, not noise - put that through speakers on maximum, ear pressed against the cone -- nothing there ... ;)
 
Frank,
If only MP3 wasn't a lossy recording method perhaps it would be satisfactory, but I just always here that things just sound lean, there is just something missing, especially in the extreme ends of the frequency band. Perhaps it is just time to poke out my ears and give it up! I don't know why with all the storage space we have these days in most any medium that we aren't all using a loss-less method like Wave or Flak? I mean the cost of a 2TB hard drive is nothing these days and even a memory stick or SD card will hold so much today.
 

This is all basic stuff. Have a look at textbook by W Yost et al or BCJ Moore.

Before the arguing goes too far, let me cut it off by confessing something. I am actually trained and have worked in this area for over 30 years. I am not trying to be condescending, it is just that I don't like having to "defend" my statements against something "published" in a wiki article.

I am sure you would feel the same way about things stated regarding your own field of expertise.
 
Yes, MP3 is lossy, it's a form of distortion. However, I would never listen seriously to MP3 being decoded in real time, I would always upsample to a higher resolution, even hi-res, and listen to that transcribed file. On my PC this makes a huge difference, cheese and chalk ...

Of course this is a silly solution, because the storage space needed then explodes, so you might as well have got the original, master in the first place
 
I'm not a fan of chesky recordings , prefering Reference recordings , Telarc ....

Not a fan of any of those, RR being fairly decent but not exceptional (perhaps its because I lack an HDCD decoder). Older Telarc (before they adopted DSD) is also OK, not stellar.

Chesky's stuff is way too bland , no life , no music , like food without the good spices and yes i do have most of his stuff ...

Yep, Chesky stuff lacks dynamics I concur. I bought a few disks many years back because I was hoodwinked by the marketing of their ADC - it had better low level linearity than the multibit converters at the time. In those days I believed low-level linearity was the thing to watch out for with digital. I hear (over the grapevine, not from listening to their most recent releases) that Chesky have now ditched their old Bob Adams designed ADC and have gone for a multibit one :)
 
diyAudio Member RIP
Joined 2005
Both statements are incorrect.

Detectability does not require that the percept be "loudness", just that it is "different". Detecting a level difference is about 0.75 to 1 dB. This is true over a wide variety of conditions (there is an exception at lower overall levels with certain bandwidth - but for our purposes that is a rare occurence). For the "balance shift" the jnd is still about the same (0.75 to 1 dB).

This is a robust finding, in spite of what someone may have written in a wiki contribution.
Well how would you like a more authoritative source? (plus you are still not getting my point about the effect of small level differences on preferences, but let's pass over that for the moment).

From the Springer Handbook of Acoustics, ed. Rossing, ISBN 978-387304465, 2007; 18.2.2 Audio and Electroacoustics, The Psychoacoustics of Audio and Electroacoustics, Amplitude (Loudness), pg. 748: "The amplitude resolution of the ear is generally taken to be about 0.25dB under best-case conditions, although for some situations it is considered to be slightly larger, on the order of 0.5 - 1.0dB."

(Section written by Mark F. Davis)

But again --- this is not what SY is saying if I understand him. We are not asking a listener to say which loudspeakers are louder, but which are preferred.
 
Last edited:
Well how would you like a more authoritative source? (plus you are still not getting my point about the effect of small level differences on preferences, but let's pass over that for the moment).

From the Springer Handbook of Acoustics, ed. Rossing, ISBN 978-387304465, 2007; 18.2.2 Audio and Electroacoustics, The Psychoacoustics of Audio and Electroacoustics, Amplitude (Loudness), pg. 748: "The amplitude resolution of the ear is generally taken to be about 0.25dB under best-case conditions, although for some situations it is considered to be slightly larger, on the order of 0.5 - 1.0dB."

(Section written by Mark F. Davis)

But again --- this is not what SY is saying if I understand him. We are not asking a listener to say which loudspeakers are louder, but which are preferred.

I understand what you are saying and I disagree with your logic (about the implied relationship between preference and detectability etc). However, I do agree that it probably does not matter to the original argument. So let's leave it at that.

As far as the rest of the stuff ..... No I am not going to degrade myself and be baited into listing the articles and chapters that I have published. If it matters, and maybe it does not, what you have quoted above is a lousy (and not very believable) summary of the literature. Believe me, no one in the field would use those sorts of numbers.

Enough!
 
Both statements are incorrect.

Detectability does not require that the percept be "loudness", just that it is "different". Detecting a level difference is about 0.75 to 1 dB. This is true over a wide variety of conditions (there is an exception at lower overall levels with certain bandwidth - but for our purposes that is a rare occurence). For the "balance shift" the jnd is still about the same (0.75 to 1 dB).

This is a robust finding, in spite of what someone may have written in a wiki contribution.

odd position to take, this gets referenced at lot here - have you any refutations citing it?

ABX Amplitude vs. Frequency Matching Criteria

Clark, David L., "High-Resolution Subjective Testing Using a Double-Blind Comparator", Journal of the Audio Engineering Society, Vol. 30 No. 5, May 1982, pp. 330-338.

he certainly missed a chance for a retraction: AES E-Library Ten years of A/B/X Testing

and of course perceptual audio codec developers think its necessary too - they are very experienced in ABX, anything interfering with listening resolution at statistical limits
 
Frank,
How do you up-sample something that has been removed? I understand using a compression scheme but that doesn't have to be a lossy format, unless I am misunderstanding the difference? Sort of like saying a small Bose speaker has high fidelity, what you don't hear you don't miss!
You can import an MP3 file into an audio editor, like Audacity. Internally, it's then treated as a very high resolution version of an uncompressed, all the "lossy" gaps filled in so to speak, music file at the sample rate of the MP3 original, say 44.1 or 48. Then one has full facility to alter the sample rate to however high you want to go, and you can export, save to file, the result as a true, say, 24/96 music file.

Edit: Remember, lossy just means that some low level detail is discarded that theoretically you shouldn't hear as happening. Neither the dynamics, nor the frequency response is tampered with if the original encoder was half decently used ...
 
Last edited:
JCX: Yes, these numbers do get mentioned frequently around here.

I have already listed a couple of standard introductory texts that one can look at. It will require some reading to understand how these things are measured and what the numbers mean. It is not as simple as one might initially think. I will stop there
 
Last edited:
diyAudio Member RIP
Joined 2005
I understand what you are saying and I disagree with your logic (about the implied relationship between preference and detectability etc). However, I do agree that it probably does not matter to the original argument. So let's leave it at that.

As far as the rest of the stuff ..... No I am not going to degrade myself and be baited into listing the articles and chapters that I have published. If it matters, and maybe it does not, what you have quoted above is a lousy (and not very believable) summary of the literature. Believe me, no one in the field would use those sorts of numbers.

Enough!
I am astonished at your arrogance! You parachute in to deny, and imply that citing your own work would be degrading?? That we are trying to bait you? What an incredibly condescending attitude!

But to each his own.

Brad Wood
 
parasound 1500A amplifier problem switch or electric current?

Hi fellow Audiophiles, I havent been on here forever.Also I have never really posted either so please forgive my ineptitude or rudeness of hijacking this post.Its not on purpose.I seen the legendary John Curl name& the date. i am pleading semi ignorance but it is a emergency lol.The story follows I recently developed problems with my Parasound 1500A amp.I never had any before.I have never abused her and only gave her the best care& always was not only impressed with the build quality but also the sound.Anyways it was working fine .That is until I bought a new house that has the older kind of electric wiring that was never being updated.Also may have put the high current power cord with 1 of my other amps?The problem is when I turn it on I only get the orange AC on and nothing else.No clicking or standby or anything green? I am upset because I its to hot forany of my tube gear which I take out in the winter and most of my other ss gear is not even close to her in sound quality..Just wanted to end with a thanks for any help offered.I need it I am broke because of this money pit house lol regards Glen
 
I am astonished at your arrogance! You parachute in to deny, and imply that citing your own work would be degrading?? That we are trying to bait you? What an incredibly condescending attitude!
......
Brad Wood

Okay, please explain to my simple brain how you can have the following. A signal that is preferred to another yet at the same time it cannot be detected as being different. After you have explained that, feel free to call me condescending, arrogant and whatever else you like.

In the end it does not matter. If you think the threshold is 0.2 dB and I state that it is 4-5 times larger, you will not be convinced.

Why bother?
 
Status
Not open for further replies.