I'll do what I can and attach my listening notes. Certainly not definitive and I'm not a trained listener.
If there's something fortuitous to come of this, going through the drivers has me off of the generic Microsoft one, which was previously prone to some unfortunate conflicts and squeals. So there's a SQ improvement.
Yes agree short term memory can disappear but speaking for myself, I can today smell that Jax beer when hearing New Orleans blues-jazz. The French Quarter 2AM natural reverb is unforgettable.
So I could not be in the loop on reproducible AB testing.
How would you pick the golden ear?
-
Attachments
Just posted a blog that has bearing on this:
http://www.diyaudio.com/forums/blogs/jan-didden/
To soften the blow, some humor. I found the attached in an email from makemydynaco.com.
Jan
Edit: vacu, great sig line!
http://www.diyaudio.com/forums/blogs/jan-didden/
To soften the blow, some humor. I found the attached in an email from makemydynaco.com.
Jan
Edit: vacu, great sig line!
Attachments
Last edited:
On the first issue I would say that if you are asked to judge a beer competition where jazz will be played, you should probably recuse yourself or at least disclose your memory association to the organizers. However, it's not clear that it alone should disqualify you from judging wine where Mozart will be played, unless you feel otherwise.
Assigning Golden Ears status is a hard question in the sense that we have not defined what the qualifications for that designation should be. I would think that at least normal hearing for a not too old person using conventional measurements would be a preliminary requirement. In other words, no hearing loss and frequency response up to at least maybe 14 -15 KHz. Again, if it's a frequency response type of Golden Ear status, then I assume we would want to see a least a little over 20 kHz on the high end. But for a distortion Golden Ear, that may not be necessary. In the distortion case, it might be someone who at least hears a little lower distortion levels that the current standards consider to be normal, or whatever the lower limit is considered to be. Something like that.
However, often we are not really so much interested in finding one freak of nature type person (a highly unique "Golden Ear"), as we are interested in finding out statistical information about people in the tails of the distribution. And, including statistics about people who have received prolonged training (whatever that is for this case) and engage in regular practice. Mastering engineers might tend to fall in this category, at least as things stand now.
Last edited:
Oh I did not consume the Jax. Could only smell it, the brewery was located within the French Quarter. I drank Galliano on rocks.
I am still eligible for golden ear status?
-
I don't know. It sounded like a pretty strong and persistent association. And I wish I knew exactly what ability deserves to be given "golden" level recognition. Would you consider accepting silver? And even at that, coming from me, you might not get much respect for it in many places.
It works for me.
I did notice that the link "Read the full article with schematics, board layout, stuffing guides etcetera here."
in the "Superregs for your line-level projects" is broken.
Last edited:
I briefly got the you have been banned forever message from diyAudio today.
Okay, no more takers for the tambourine test, it seems.
SPOILER ALERT:
Here is the decoder information.
The 24-bit file is the original recording with no mixing of any kind other than shortening tails a little
The first four 16-bit files were produced in Reaper with the following options:
A EHQ Interpolation (Extreme High Quality) Shaped Dither
B EHQ No Dither
C Good quality interpolation (default) Shaped Dither
D Medium quality interpolation (2nd to lowest, the lowest is linear interpolation) Shaped Dither
The last two 16-bit files were made using Voxengo R8BrainPro. Both used Ultra-Steep mode, Auto clipping avoidance, and flat dither. R8Brain upsamples to the least common multiple sample rate of the source and destination file sample rates:
E Linear phase filter (some pre-ringing would be expected)
F Minimal phase filter
Here is a link to a version of the origninal recording with compression and EQ applied. The processing brings up the low level detail to make it more easily heard: https://www.dropbox.com/s/kzece70jmxkhgu2/Tabourine24_Comp+EQ.zip?dl=0 At the end you can hear me walking across the soft carpet to turn off the recorder.
SPOILER ALERT:
Here is the decoder information.
The 24-bit file is the original recording with no mixing of any kind other than shortening tails a little
The first four 16-bit files were produced in Reaper with the following options:
A EHQ Interpolation (Extreme High Quality) Shaped Dither
B EHQ No Dither
C Good quality interpolation (default) Shaped Dither
D Medium quality interpolation (2nd to lowest, the lowest is linear interpolation) Shaped Dither
The last two 16-bit files were made using Voxengo R8BrainPro. Both used Ultra-Steep mode, Auto clipping avoidance, and flat dither. R8Brain upsamples to the least common multiple sample rate of the source and destination file sample rates:
E Linear phase filter (some pre-ringing would be expected)
F Minimal phase filter
Here is a link to a version of the origninal recording with compression and EQ applied. The processing brings up the low level detail to make it more easily heard: https://www.dropbox.com/s/kzece70jmxkhgu2/Tabourine24_Comp+EQ.zip?dl=0 At the end you can hear me walking across the soft carpet to turn off the recorder.
Last edited:
Nothin's wrong with an opamp circuit that a good discrete won't fix...
Troll bait fer sure, bur I've been on diyaudio since 2004, and my opamp submissions have been few and far between. Make whatever conclusions you want to make from that.
If you want to spend the money and buy the chips, you can get good results (or not, given skill level) from opamp circuits.I prefer to take the challenge further, and haven't been disappointed. Bye-bye, and happy new year.....
Troll bait fer sure, bur I've been on diyaudio since 2004, and my opamp submissions have been few and far between. Make whatever conclusions you want to make from that.
If you want to spend the money and buy the chips, you can get good results (or not, given skill level) from opamp circuits.I prefer to take the challenge further, and haven't been disappointed. Bye-bye, and happy new year.....
Some further commentary on the Tambourine test.
Here is some description of Reaper's shaped dither:
"There are lots of discussions and arguments out there about whether or not to dither and when to do it and why. This post isn't about that -- it's just to explain technically what Reaper does if you click the buttons.
The dither and noise shaping are independent (you can use either with or without the other).
The dither is 1.5 bits which is somewhat on the light/do-no-harm side. With noise shaping off, the dither PDF is triangular; with noise shaping on, the dither is high-pass filtered.
The noise shaping is a 9th order filter, which is similar to pow-R type 3.
Unlike pow-R (or almost any other method), the noise shaping coefficients are calculated on the fly for any sample rate analytically, based on a very smooth equal loudness contour.
The render dialog should only give you the option to dither when the target bit depth and format is appropriate. The calculation is done post-master-fader, so you don't need to do anything special except click the buttons.
There are a few JS dither plugins included with Reaper. These are for technical use or backwards compatibility. You don't need to use them. Definitely don't use any dither plugin if you're using the native dither options when rendering."
For R8Brain, it applies flat dither only when an output format with reduced bit depth as compared to the source is selected. The preferred process some people recommend is to use R8Brain for SRC but not bit depth reduction. Then use the shaped dither in Voxengo Elephant, followed by bit depth reduction in Reaper or similar, without any further dither selected.
Here is some description of Reaper's shaped dither:
"There are lots of discussions and arguments out there about whether or not to dither and when to do it and why. This post isn't about that -- it's just to explain technically what Reaper does if you click the buttons.
The dither and noise shaping are independent (you can use either with or without the other).
The dither is 1.5 bits which is somewhat on the light/do-no-harm side. With noise shaping off, the dither PDF is triangular; with noise shaping on, the dither is high-pass filtered.
The noise shaping is a 9th order filter, which is similar to pow-R type 3.
Unlike pow-R (or almost any other method), the noise shaping coefficients are calculated on the fly for any sample rate analytically, based on a very smooth equal loudness contour.
The render dialog should only give you the option to dither when the target bit depth and format is appropriate. The calculation is done post-master-fader, so you don't need to do anything special except click the buttons.
There are a few JS dither plugins included with Reaper. These are for technical use or backwards compatibility. You don't need to use them. Definitely don't use any dither plugin if you're using the native dither options when rendering."
For R8Brain, it applies flat dither only when an output format with reduced bit depth as compared to the source is selected. The preferred process some people recommend is to use R8Brain for SRC but not bit depth reduction. Then use the shaped dither in Voxengo Elephant, followed by bit depth reduction in Reaper or similar, without any further dither selected.
One of the more revealing tests for me showing opamp differences ( and discrete jfet op amp ) is Greg Lake's (rip) Father Christmas. After the acoustic intro there is collection of Christmas bells which are wonderfully detailed on the best reproducers or harsh and blurred on the worst of them. No measuremnt required, I can hear it....
Thanks for the results Mark. I guess that statistically the impressions I posted don't really mean much as its possible to put some spin on the results.
Its a real pity that more didn't post their impressions... however.
I did feel E stood out as being lacking in resolution in some way and that is one you mention as having some expected pre-ringing. Would that be the audible effect of that. I don't know
Thanks.
Its a real pity that more didn't post their impressions... however.
I did feel E stood out as being lacking in resolution in some way and that is one you mention as having some expected pre-ringing. Would that be the audible effect of that. I don't know
Thanks.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.