Went back to black belt. Bigger sound stage and more bass. |
Tried a red belt from North Hill Audio (from ebay, cheaper than white belt, not affiliated). Good numbers initially:
mean freq: 1000.2 Hz
filtered freq:
-0.04% / +0.05%
-0.4 Hz / +0.5 Hz
Also tested my black belt which has posted the best numbers so far:
1002.9 Hz
-0.09% / +0.02%
-0.9 Hz / +0.2 Hz
It is a lot warmer today (by 20+ deg F) since I last ran these tests. Maybe the black belt stretched? Maybe it is more sensitive to temp? Will listen some to the red belt and re-test again. Both seem to be doing better than the factory white belt I got.
One more test run with the red belt:
1000.0 Hz
-0.05% / +0.04%
-0.5 Hz / +0.4 Hz |
To set your mind at ease look at this thread in which various TT's were tested with the platter speed app, The best was the Nakamichi, which centres the record, most of the big variation is eccentric spindle hole.
http://www.stereo.net.au/forums/index.php?/topic/49224-turntable-speed/page-2?hl=%2Bplatter+%2Bspeed#entry941503 |
"The WOW measurement according to DIN IEC 386 is quite accurate in measuring relatively slow speed fluctuations in a range of <10Hz (WOW). Faster variations (flutter) are a problem to measure with a phone because audio can be sampled at a maximum rate of 44.1kHz which is limiting the measurement of fast fluctuations. Please put the device on a solid base for measurement in order to avoid shaking it which causes some minimal doppler effects and decreases the accuracy." |
Noticeable difference on solo piano. After reading more, it sounds like there
is some variance with belts and test numbers. So, finding a belt that
produces the test tone most accurately is the right one for me. Just got my
red belt in the mail. Have not tried it yet. |
Yes, nicely laid back with more definition between the notes plus a little more warmth.
I guess Rega took the shortcut with the white belt upgrade, like other turntable manufacturers (pro-ject), thinking a faster running table sounds better. I disagree... |
Those numbers with the black belt look real good now. Could you hear a difference in the music with the white belt vs. the black belt? |
Ok, did some more tests today. Found some more things about the iphone app and the differences between the Rega white and black belts.
First, the iphone app. Sometimes if you let the test tone run out to the end of the record or lift the needle while the app is running it can cause a large frequency spike or dip, which throws off your overall numbers. I think this is why my numbers were weird (thanks @Tonywinsc for noting this) showing the table running a little fast but the negative deviation being the largest, which didn't make sense. All the graphs for these weird numbers showed a large dip at the end of the measurement session. Since there is no stop measurement button on the app, I've found changing the preamp input is the best way to stop the app without any spike or dip.
So, with the white belt and changing the input source on the preamp after a good sample duration, I got the following numbers:
overall freq: 1001.8 Hz
filtered deviation:
-0.02% / +0.08%
-0.2 Hz / +0.8 Hz
And with the black belt today, I got:
overall freq: 1000.6 Hz
filtered deviation:
-0.02% / +0.03%
-0.2 Hz / +0.3 Hz
So, after reading about other experiencing larger deviations with the white belt, it seems I am seeing the same thing plus it makes the table run faster.
I'm keeping the black belt on for now. I have a $20 red belt on order from the UK after reading good things about it. Will post those test results when I get it.
Anyone want to buy a Rega white belt with not many hours on it (about 6 months of average play)? ;) |
Some interesting reading here:
http://www.vinylengine.com/turntable_forum/viewtopic.php?f=32&t=65143
I still think i will be sending back my TTPSU for verification of settings to spec... |
The results are abnormal. I wonder if the belt might be slipping a little or damaged. Try cleaning and even replacing the belt first. Then if no better, seek repair. Does the plot look onesided as well? |
Numbers are basically the same with the phone on the table. Lowpass filtered is:
-0.11% / +0.03%
Should I have the TTPSU checked by Rega as it is still under warranty? Or, could it be the test record?
The GT Reference RP6 subplatter I am using came with a much nicer bearing and spindle that rides on the bearing. It also came with lubricant that is not as thick as the Rega factory oil. Should I try the Rega factory oil? I could not find wow / flutter specs for this table in the US manual. Anyone know what they are for a RP6 with the TTPSU? |
Yes, I am holding the phone. I will try it stationary next time. |
PS. A record with the 3150Hz tone will be more accurate; if you could get one. |
What is puzzling is such one sided numbers. Are you holding your iPhone in your hand when recording maybe? Place your iPhone on a table while recording. The raw is the actual frequency swing as the record rotates which includes eccentricity of the record hole. The program filters out this periodic side to side motion of the record since that is occuring at 0.556Hz. The filtered number is your actual turntable performance. The +0.05% looks good; but the -0.1% is troubling and I am wondering if the platter is really slowing down occasionally by that much. |
You might see how it performs after its been running for 6 hours. |
Second run of over 50 seconds was much better:
mean freq: 1001.4 Hz
raw freq
max deviation (relative): -.75% / +.68%
max deviation (absolute): -7.5 Hz / +6.8 Hz
lowpass-filtered freq
max deviation (relative): -.10% / +.05%
max deviation (absolute): -1.0 Hz / +.5 Hz |
Correction, speed is fluctuating between 33.27 and 33.43.
Here are some more app numbers from the 35 second sampling:
mean freq: 1001.3 Hz
raw freq
max deviation (relative): -2.76% / +.63%
max deviation (absolute): -27.6 Hz / +6.3 Hz
lowpass-filtered freq
max deviation (relative): -.16% / +.04%
max deviation (absolute): -1.6 Hz / +.4 Hz
What is the difference between the raw and lowpass-filtered frequencies? What is the net wow / flutter measurement of my table using this app?
Thanks everyone! |
Ok, some new info here. I figured out why the Platter Speed app was not registering. I didn't realize that side 1 of the Clearaudio burn in LP is 45 rpm. Once I set the TTPSU to 45 rmp, the app registered the new data, which follows:
- speed varied between the -.2% slow mark and +.3% fast mark during the 1 kHz test in the first section on the first screen of the app
- unfiltered frequency varied between 1006 (with a few peaks at 1008) and 996 Hz over 35 seconds on the second screen of the app
- filtered frequency was almost completely flat at 1002 Hz over 35 seconds on the second screen of the app
So, if I'm reading this correctly, the filtered looked very good and it looks like my speed is fluctuating between 33.39 and 33.43 every second.
Much better? |
I found the Feickert 7" (my copy) tone to have a lot of inherent wow, it's off center or something. For example, if you are at 33.33 you'll see the bar go all the left/right edges of the green section.
The UTR 3150hz has much less sway back and forth in the Platterspeed's meter. It will do a much tighter measurement which looks more like the demo if you have run that in the app. |
Apologies apparently are in order. That amount -/+ 10 Hz is really 20 Hz because it is going from 1000 to 1010 to 990 and in between. I left off the % multiplier which changes it signicantly. 20/1000=.02x100=2%
Still, listening will give a good picture. Excerpt from NAB is correct. Ive been trying to increase a 1Khz tone with the signal generator (True RTA) but I can't move it in 1 hz steps to see how noticable it is, and my old analog signal generator has no way to accurately do the same as well.
Again, my apologies. |
@Gvasale and @Atmasphere, what is the formula you are using for these calculations? |
That's about 0.1% isn't it? maybe sounds a little on the high side... |
If I'm not mistaken that level of performance translates to .02% away from what a casual internet search showed to the best to be, a VPI TT with .01%
Anything that bests .1% exceeds NAB specifieations.
"Wow and Flutter Factor (Re~ording)~
1.15 It shall be standard that the average deviation (measured over the range 0.5-200 cps)
from the mean speed of the recording turntable, when making the recording, shall not exceed
0.04% of the mean speed. The average deviation above shall be measured by a meter the dynamics
of which shall be the same as those of the VU meter as specified in ASA Standard C16.5-
1961.
Wow and Flutter Factor (Repr~ducing)~
1.20 It shall be standard that the average deviation from the mean speed of the reproducing
turntable when reproducing shall not exceed 0.1% of the mean speed."
I can't remember anything that would suggest the NAB would want to drive listeners away from either Am, FM radio because they were sloppy or uncaring with audio performance.
Because today, many turntables routinely perform in the range of .03% or less can't possibly mean your head might spin when listening or that performace in the .03% to .01% isn't top notch. |
Should I try another platter to see if my current factory Rega one is defective? |
That level will make a difference in clarity, decay, imaging, even if you notice no obvious speed instability.
I've seen a lot worse on a lot more expensive turntables, brand new. That does not make it right.....or good......but it's not unusual. |
My phone locks onto a 1kHz tone with the app. I have the latest IOS, unfortunately. Go to info and see if you have the latest version of the app. Under "Prerequisites" It should say that it now works with a 1000Hz tone. It is not as accurate vs. using the 3150Hz tone.
Wow&Flutter is the magnitude of speed variation during platter rotation. That means the frequency will vary proportionately if playing a constant frequency tone such as 1kHz or 3150Hz. It is the product of motor speed variation, eccentricities of the record, platter and motor pulley plus roundness of the platter and motor pulley, if using a belt drive turntable. It is typically considered to be a periodic variation. The iPhone app is based upon that assumption and generates a filtered value to eliminate the influence of test record's eccentricity on the final value. So if the platter is mounted eccentric to it's bearing center, this app will mask that issue as well. Simply watching your platter and/or record for side to side motion during rotation can confirm how well they are centered. If you want to really feed your nervosa, borrow a dial indicator from a machine shop and measure the out of roundness and eccentricity of your platter. I'm not going to do it. No I'm not. No sir. Please don't make me... |
Play some music with piano on it. Not necessarily rock&roll.
Look for something that is moderate or slower in tempo.
You should be able to hear whether a variation of the degree you desctibe is noticable or objectionable.
Report back to us & tell us: "Wow, does this have flutter"
Hopefully it is a non issue. |
