• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

PlayClassics TRT v2.0 master file giveaway for ASR members

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#61
Albéniz -the right hand is a bit heavy in parts - 1-3khz area a bit objectionable to me at times. Has some hardness in it.
Hmm... not hearing that so much this time around. Just hearing him pounding the keys and the piano making its sounds now.

Perhaps I had the level just a little higher when that judgement occurred to me, or I got used to it, as that was an initial impression that stuck for a while. It also came from initial listening to the Albéniz demo track, and not the entire piece.
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#62
LZ peak - 99.4 in the first couple of minutes, 101.3 somewhere later in the piece since it is still playing and I look at the display again
Humm, That's seems pretty high Ray. Do you think that's fairly representative of what the mics heard?
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#63
Could be. A grand piano is a studly device.

Come on over and see what you think. I find it to be a comfortably loudish level.

Just now, a single hand-clap 5 feet off to the side and a little behind the mic gave 113.1 dB peak. A loud (but not shouting) "hey!" gave 94.6.

And that is peak, not averaged.



upload_2016-8-30_22-19-14.png


Ever used a peak reading meter?
 
Last edited:

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#64
I have a Radio Shack SPL but have never used it in the presents of live acoustic instruments.
I'm just asking questions and learning here as I have no direct experience from which to draw.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#65
Doesn't read peaks, assuming it's the analog or its digital replacement that has been around forever. I have two of the analog type.

The original SPL post - 86dB slow Z (unweighted) maximum, take off a couple of dB for C weighting, maybe you'd read 82dB in the loud parts.



---

Albenez #3:

Ah, good guess. I pulled my meter, battery still good, slow, C weight, about 82 on the loud stuff.

On the fast setting, in the 90's on the loud moments, but I suspect a little ballistic overshoot, as the REW meter reeads 89.7 on LZFmax.
 
Last edited:

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#66
Doesn't read peaks, assuming it's the analog or its digital replacement that has been around forever. I have two of the analog type.

The original SPL post - 86dB slow Z (unweighted) maximum, take off a couple of dB for C weighting, maybe you'd read 82dB in the loud parts.

Yep I got the original analog one on the left, had it since around 1990
So setting the response to Fast doesn't actually equal level peaks even though it can't hold em?
I'm so confused. LOL
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#67
Slow and Fast are averaging times.

"Fast, Slow and Impulse
Most sound level meters have two conventional time weightings, F = Fast and S = Slow with time constants of 125 ms and 1000 ms respectively.
Some also have Impulse Time Weighting which is a quasi-peak detection characteristic with rapid rise time (35 ms) and a much slower 1.5 second decay.
F : Fast = 125 ms up and down,
S : Slow = 1 second up and down,
I : Impulse = 35 ms while the signal level is increasing or 1,500 ms while the signal level is decreasing.
Back to the days of analogue meters, Time Weightings were introduced to give the operator chance to 'follow' the rapid meter fluctuations by eye.

Peak, P-P, True Peak, Lpeak, Lpk, etc.
Peak : to measure the True Peak values of impulsive sound levels a meter must be equipped with a Peak Detector."

http://www.acoustic-glossary.co.uk/time-weighting.htm

http://support.radioshack.com/support_audio/doc72/72441.pdf
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#68
Yep got the original one in the box even. LOL
There's also a neat little pdf with some measurements of the different models.
http://diyaudioprojects.com/Testeq/RadioShack_33-2050/RadioShack-Sound-Level-Meters.pdf
Crude in any case, just fine for my speaker/subwoofer measurements in the past.
Going to have to break down and buy a UMIK-1 in any case. Been trying to use the RS meter as a mic for REW and it just won't work for me. Even using CSL's calibration file REW has had me pulling my hair out just trying to do a basic response measurement :mad:
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#69
Humm, That's seems pretty high Ray. Do you think that's fairly representative of what the mics heard?
I've been looking and found 109dB peak thrown around, but it took a while to find a more scholarly example, instead of a vendor.

upload_2016-8-31_2-29-35.png
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#70

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#71
You don't have a peak meter...

Albeniz #3:

On my RS meter, 80dB range, C weighting, slow, it maxed about 85 during the first two minutes.

REW claimed LCSmax of 87.9, with a peak of 101.9 during the same time.
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,586
Likes
2,978
Location
Central Fl
#72
You don't have a peak meter...

Albeniz #3:

On my RS meter, 80dB range, C weighting, slow, it maxed about 85 during the first two minutes.

REW claimed LCSmax of 87.9, with a peak of 101.9 during the same time.
OK, thanks. Guess my attempt to pick a number I thought would be close, wasn't. LOL
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#73
Last edited:
Joined
Aug 19, 2016
Messages
44
Likes
9
#74
I've played all the material.

I think you've done a good job, per your criteria.
Thank you Ray, your comments are very much in line with what the performers are doing. I think it is working all right :)

The purpose of our fixed recording setup is to produce a "consistent" sound across all recordings. Then you can easily recognize the ability of the artist depending on what he is able to do with it. All these recordings are done in exactly the same way, nevertheless they all sound different.

Albéniz -the right hand is a bit heavy in parts - 1-3khz area a bit objectionable to me at times. Has some hardness in it.
Debussy - better overall balance, not as loud as the Albéniz, so the problem may be the player (or my playback) and not the recording.
You talked about the "right hand high notes" on the Albéniz. Luis voiced it that way, we just picked up what he was doing. When you listen to Cabrera playing Debussy you do not hear that at all. You might like one thing better than the other but you can be sure that whatever it is you are hearing they are purposely playing it that way.

The sound quality on these recordings is a constant, so when you listen to different recordings you are truly appreciating the different qualities of different performers, the same way you would if you were to listen to all those performers playing live on the same auditorium. To put it another way, what we have done is develope a system that takes the engineer/producer variable out of the way so you can get to appreciate the sound of the performers without any makeup (for good or for bad)

I found the Flamenco vocalist a bit odd-sounding, though it may simply be the style, with which I am less familiar. The guitar came through very well.
If you look at the flamenco sketch you will see that the flamenco singer was sitting much closer to the mics that the guitar was. That may have a lot to do with what you are hearing on this sample.

Flamenco is usually recorded using close miking techniques. They are used to having the engineer balance their performance at the mixing table. But we do not do that. Here the artists are responsible for the balance of the performance. Just like they would be if on an acoustic live performance without any amplification. They were here for just one hour trying things out. They ended up making the recording this way to try to compensate for the fact that her voice was not so big and the guitarist sound was big. Had they been here long enough, I am sure they would have been able to balance themselves to be able to sit next to each other on the stage. That would have probably gotten rid of that "odd-sounding" quality of the voice.

I cannot say anything about the rock. I can only be thankful because they were willing to try this experiment. I think rock in general does benefit from the intervention of engineers and producers on the making of the sound. Anyway, I would be willing to try it again if there was a band that dared to do it :)
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#76
You talked about the "right hand high notes" on the Albéniz. Luis voiced it that way, we just picked up what he was doing.
I agree.

Undiluted solo piano ffff may not have been featured often in my typical listening repertoire.


 
Joined
Aug 19, 2016
Messages
44
Likes
9
#77
ok, I have just read though all the posts I missed....

We always use the same gain on all our recordings. (except for the drums and rock)

The albums and the flamenco all have real relative levels. So once you find a realistic master volume level you should be able to listen to all the albums and the flamenco without having to touch the volume.

In other words, whatever volume you choose, leave it there and switch back and forward between piano, guitar, voice... if you do not feel the need to adjust the volume for any of those then you can be pretty sure you are listening at a realistic volume.

Anyway, It might be a good idea for me to develop a test tone that you could use to calibrate your system (should you want to) That way you would be sure that the volume you are listening to corresponds to the real volume of the performance.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,048
Likes
23,034
Location
Seattle Area
#78
Anyway, It might be a good idea for me to develop a test tone that you could use to calibrate your system (should you want to) That way you would be sure that the volume you are listening to corresponds to the real volume of the performance.
That would be great. What would be fantastic :), would be some kind of frequency response of the recording room. With those two we can come much closer to the live recording.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,010
Likes
3,269
Location
Riverview, FL
#79
The albums and the flamenco all have real relative levels. So once you find a realistic master volume level you should be able to listen to all the albums and the flamenco without having to touch the volume.
I didn't change volume while listening to your tracks.

In other words, whatever volume you choose, leave it there and switch back and forward between piano, guitar, voice... if you do not feel the need to adjust the volume for any of those then you can be pretty sure you are listening at a realistic volume.
That's how I felt.

Anyway, It might be a good idea for me to develop a test tone that you could use to calibrate your system (should you want to) That way you would be sure that the volume you are listening to corresponds to the real volume of the performance.
When first pondering 24 bit long ago I had it in my head that the levels on disc would (since they could) correspond to the "live" levels, all the way up to 144dB, which would usually leave a whole lot of headroom, but that's not how it was implemented

What would be fantastic :), would be some kind of frequency response of the recording room.
What would be your idea on how to do that? I draw a blank.


----------------

Mario is Mr Truth in Recording

Here's my current method for Truth in Playback (at least, true to the source, which is all I get when I doo da best I can).



My "calibration" procedure for CD (and other digital) playback:

First, calibrate the REW source monitor:

Record a full scale tone onto CD.

Open an instance of REW.

Playback via CD player's analog outputs (no DRC applied) and calibrate REW to that analog electrical signal, 96dB.

I pick 1000.1 hz, an FFT bin frequency (per REW), and since it was common in my telecom life.

upload_2016-8-31_13-57-13.png


Using the REW tone generator, play a full scale digital tone and verify 96dB is seen when using the normal path through the central system DAC.

It's not, 97.5dB is seen. Take note and call it "close enough".

Here's the Note: The full scale digital monitor reference is 1.5dB hotter than the analog reference. Something to investigate later.

upload_2016-8-31_14-10-32.png


Open another instance of REW, for use with the measurement mic.

Put on ear protection.

Play the full scale digital signal, adjust volume for 96dB at the mic, without DRC.

Note the volume setting on the preamp. it is 58 / 151

Turn on DRC. REW reading is 89.7dB. DRC needs a little headroom to do its magic.

Adjust volume back to 96dB at the mic.

Note the volume setting on the preamp. It is 77 / 151

upload_2016-8-31_14-20-31.png


Play CD music on one speaker only.

Monitor peak RTA levels for source and destianation, and check levels.

Top is in-room, bottom analog source from CD. I like to think the comparison comes real close. Good enough for now. Good enough for my deaf ears. Probably about as good as anyone elses.

upload_2016-8-31_14-33-4.png


Call it "calibrated".

It's not perfect. But the system outputs at the level of the recording, and is loud, but not too loud.

When both speakers are playing, then the in-room trace will be higher than a single channel of the CD monitor, by up to 6dB. The peak trace will vary a bit too.

Anyway, that my Better than Nothing playback calibration method here at Neverland East, when the idea is to calibrate the playback to the levels of the source.

My objective is accuracy to source, this is how I go about confirming it.

Repeat for the right speaker, maybe. Or just go on to both speakers:

This shows the additive room level, and some differences in the congruence of the room trace (both channels) to the monitor (left only).

upload_2016-8-31_14-45-26.png


Sometimes I put both left and right on monitor, and sometimes create a mono signal, just for personal amusement.

After all that, which is really pretty much a one-shot deal, I twist the volume knob to wherever I want it, generally about 10dB lower, but this is my "calibrated" loud level, for loud listening, and for comparison and verification of in-room sound to source.

I know it's not perfect. What is?

Any questions?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,048
Likes
23,034
Location
Seattle Area
#80
What would be your idea on how to do that? I draw a blank.
I am assuming there is a system where the talent approves the final work. It is that system+room that I like to see profiled.
 
Top Bottom