Yeah, I've posted tests with piano and voice and gotten the "I can't tell from that" comments. Please.... 🙄
I might do that next week. Having lived with a supposed much better player (Marantz Pearl-Lite) for the last 4 years it will be an interesting 'swap'.
I've often used music I wouldn't normally listen to as test pieces. When a system 'clicks' it can actually become more bearable... if that makes sense.
I found this display shows the difference best. This is a plot of 20*log10(|a1|)-20*log10(|a3|) from the first 5sec (220500 samples) so each bin is exactly .2Hz. No windowing or stretching to a log frequency plot which I think obscures the effect. The vertical axis is dB and I think the ~.5dB baseline is visible while the verticle is linear Hz*5 so 500 is 100Hz. I could be that one track has a small amount of distortion in the lows so when compared in dB there could be a large ratio, I don't know. There are stretches with almost perfect match. Moving things around and changing the external noise pickup between tracks could also matter certainly the stuff below 20Hz or so would be mostly LF noise and possibly the A/D has enough 1/f noise to account for all this.
Just food for thought.
Just food for thought.
Attachments
What it is doesn't much matter, it just needs to be recorded well.
Maybe some Nights from the Alhambra? Many Jazz titles would qualify. Well done folk or ethnic music.
Something with a good 3D image/soundstage and little use of effects.
None of that thou eliminates the problem of convolving the DUT with the ADC and the user's DAC.
dave
The question you're not asking is, "Why isn't Joe doing this?" His claims, his business, his responsibility for validation.
Because he was asked by diyA members to post implementation details.
It is a simple mod, he thinks it works, but makes no real claims. Now diyers can try it if they want and make up their own minds.
Even if it is just a placebo and it makes listeners happy that is a good thing.
I don't really care, but i do care that you have turned a simple "let's try it and see, here is how" into a witch hunt.
dave
No, but this is a golden opportunity to show validation via measurement and/or ears-only no-peeking listening tests. I prefer to be optimistic. 😀
What it is DO matters. Such records you describe will favor the modification. I prefer good music than good recording. Something from Dire Strait is liked by many audiophiles. Telegraph Road, Private Investigation, will favor no modification I believe. Or the AC/DC.
Thanks Scott, that's what I was thinking. A/D noise. I often see this very low end noise in soundcard loop tests.
Below is what I found in the beginning 5 seconds. Analysis done in Audacity @ 65K, exported and plotted in HOLMImpulse to compare. 1/12 octave smoothing.
Attachments
I probably could come up with something along those lines but its not the sort of stuff I listen to these days.
Also, isn't this getting round to the old notion of a system that sounds just brilliant with some types of music or a select few discs, and that the fact that others sound no good is down to the recordings. I made the transition away from that many years ago.
In fact I've just pulled the Shostakovich Preludes and Fugues* that I am (make that was 😀) listening to and popped the Genesis disc in. Nothing wrong with what I'm hearing here although its not the Sony that's spinning it.
* Can't imagine those going down well as a test track.
The point I'm trying to make is that a successful system has to deliver the goods pretty much irrespective of what you throw at it.
That explains far more than many might care to believe (imo).
Lol, even I have that album.
And no dorking around with settings in between, also. Just spin up the music (record/cd/hdd/electrons in a ssd...take your pick) and enjoy.
Only in that it allows enuff information to actually allow any veracity of the mod toshow itself. Otherwise you are like the drunk who is looking for his keys under the lamppost even thou he knows that is not where he lost them. If what the antecdotes suggest the change does are not able to be expressed in the music you listen too, then the mod is of no use to you, but that has no bearing on deciding whether it works or not.
dave
Comical I should come across this blog on Nature about ways scientists can fool themselves. While the article's direction is towards its readership, the message is far more general.
Mind your P's and Q's. Mind also how you determine your P's and Q's.
How scientists fool themselves ? and how they can stop : Nature News & Comment
Mind your P's and Q's. Mind also how you determine your P's and Q's.
How scientists fool themselves ? and how they can stop : Nature News & Comment
My Listening to Natalie, Ori vs Modded
With the Genesis, I preferred the ori. I choosed A1 (instead of A2) to compare with the mod because initial quick listening I reveal that A1 was more enjoyable. This is most probably because it is 0.5dB louder. During ABX, I was able to easily hear that the ori is more natural, hence my preference, to support the first listening perception (enjoyable).
With Natalie, I preferred the mod, but I didn't compare with the ori, B2. The reason I choosed B1 was because it is more different than B3. B1 was much worse, may be because of the extra RC filter added.
Now I'm comparing B3 with the ori, B2, and I think I know what I'm hearing...
The original is more accurate. The mod added second order distortion that made soundstage wider and sharp peak rounded thus easier on ears.
So, is there any way/plan to measure the harmonics?
With the Genesis, I preferred the ori. I choosed A1 (instead of A2) to compare with the mod because initial quick listening I reveal that A1 was more enjoyable. This is most probably because it is 0.5dB louder. During ABX, I was able to easily hear that the ori is more natural, hence my preference, to support the first listening perception (enjoyable).
With Natalie, I preferred the mod, but I didn't compare with the ori, B2. The reason I choosed B1 was because it is more different than B3. B1 was much worse, may be because of the extra RC filter added.
Now I'm comparing B3 with the ori, B2, and I think I know what I'm hearing...
The original is more accurate. The mod added second order distortion that made soundstage wider and sharp peak rounded thus easier on ears.
So, is there any way/plan to measure the harmonics?
It's not that simple a mod. Actually, designing a fair and conclusive test for it isn't easy.
1/ It only appears simple if you follow exactly the implementations shown in the first posts. Even in the first "scenario", as soon as you change the opamp, the value of the I/V resistors and so on, you can get a different frequency response and all bets are off about the ability to reproduce the alleged effects.
2/ It cannot be said a simple mod when you have 8 different scenarii to account for. Can we even seriously think that it is "one" mod with such technical variations ? Still, that's the main claim at hand, isn't it ? That we get a similar subjective improvment in all 8 cases ?
3/ The last pages of testing are an amusing experiment but somehow miss the point. At best, they would only establish that there is an audible difference in between the particularly modded player and the player in its original state. Even if we dismiss the point raised by the OP himself (that playing the test files through another DAC invalidates the mod), that wouldn't establish that this difference is what the OP (and others) says he (they) hears. We would need cross-scenarii modded/unmodded DBT for that. Sounds like a load of fun. Or not.
edit: 4/ And actually, we'd need to add the variation "corrected FR/uncorrected FR" to the DBT, to account for the auxilliary claim that it isn't the FR deviation itself which is to blame/praise.
Last edited:
- Status
- Not open for further replies.