Measuring Some Vintage Speakers

[bconline]

Nice to see that graph. I've hear people say these speakers with the 035ti tweeter go bad and sound harsh because damping material in the tweeter disintegrates. But this measurement doesn't show any obvious problem in the treble.

Actually, the measurement I posted was from the better of the two L100t's the other one has tweeter problem that I have not figured out yet. But instead of harsh, it is coming off as dull. Here's a comparison of the two.

 

[AudioX3]

Great posting and learning experience. The technique you deployed I have seen others do but done outside to take out any room modes. If you had fun it was worth it.

 

[Golf]

You can confirm the information by going outside with an open surrounding. NO wall or artifacts that would cause an effect on the measurements.

Nevertheless, there will be some kind of floor. Some kind of.

 

[thewas]

Nice work and don't worry, since you have done the measurements with REW you have also the distortion ones automatically and you can in few seconds add a gate which will reduce the room impact, I would recommend something like between 5-10 ms.
Would be great if you could post those here one day, no hurry or stress.

 

[LugsyTL47]

Great post, thank you for the effort. With results posted on single graph, reinforces what I’ve always known, that I've saved a lot of time by just enjoying my Yamaha’s. Thrown in everyone’s unique room, a couple drinks and music of choice…good times all around.

 

[bconline]

Thank you for doing all this and sharing! I'm always very curious about the performance of vintage speakers.

Because you have plenty of space around the speakers and measured at 1m, you should actually be able to get a decently high resolution gated (AKA quasi-anechoic) measurements for upper frequencies using the data you've already captured! How clean the data is depends on how much the cabinet the speakers are rested on affects things, but since you pushed the speakers most of the way toward the front, it should not be a big deal.

In any case, the measurements wouldn't change much since the speakers are already fairly far from the walls, but could be interesting to investigate.

If you are willing to share all the measurements in a REW file, I can go ahead and convert these into gated measurements (although it's not complicated).

Also, I don't at all intend for you to do any more work, but if there any specific speakers you want to measure the bass for quasi-anechoically I can help you out with that too. The measurements process is fairly simple to do, and basically just involves placing the microphone very close to the woofers and ports. You wouldn't have to position the speakers in any special type of way. I could do then do the processing.

Wouldn't be *pristine* without a bit more care, but should "close enough".

(I apologize if you already know how to do all this, just offering the few skills I have in this space )

Thank you for the offers! Here's a link a zip file with the .mdat files from REW. I am looking for ways to measure bass better and get rid of the floor bounce problem.

 

[bconline]

Thanks for publishign this.
Quite some work...



Looking at this one is interesting, IMO, since it shows immediately common "incidents" which are more likely due to room/position.
Like the peaks at 55Hz, 70Hz and 95Hz and the consequent dip around 85Hz.

This could be even more obvious if the plot levels were aligned on an average on, say, 1kHz-2kHz, as an example.

I'm not sure how to interpret this, but here's the RMS average of the 14 measurements. Does this confirm room modes you mention?

 

[bconline]

bconline and napilopez, please show also distortion dB from REW! THD plus 2nd to 5th harmonics - and noise floor. Looks like your amp had same voltage/gain setting

Actually distortion dB and fundamental is enough, it uses variable gating I guess, and presents on-axis spl very well in same pic!

How did you mount small vs. towers, was mic height matched to tweeter height or what? Vertical response variation is guite bad, you know!

Notes:
- to record noise floor, it's box must be tagged before making the measurement!
- my collection of measurements are not standardized, and cover 15 years and several versions of REW

Pic of JBL LSR305


View attachment 360311

Here, I think, is a noise floor and THD chart for the Revel F206. I don't know why the THD graph is gray.

For the mic position - I placed the smaller speakers on the cabinet in the first picture, which was 27" high. The mic was adjusted to center on tweeter height for each speaker.

 

[Juhazi]

Thank you for the offers! Here's a link a zip file with the .mdat files from REW. I am looking for ways to measure bass better and get rid of the floor bounce problem.

Seems like your noise floor is too high to evaluate distortion. Something in the room resonating at 120Hz? Some gadget making noise and distortion at 1kHz?

Anyway nice effort and fun to follow! Please spend some time to tune your technique - do lots of tests with various changes - one at time!

Mic location and speaker distance should remain fixed. Use a variable stand to adjust speaker height. A turntable plate is easy to make! I suggest mic height 1,0m and distance 2m (1m is too near for floorstanders). Spl at 2m should be 90dB at 1 kHz, or like that if you want to keep gain fixed. That way we can see sensitivity too (relative, not absolute)

 

[Hart]

This is really interesting, there are so many vintage speakers I would be curious about, great job!

 

[AdamG]

These are in a room responses at 1m on axis at tweeter height.... How do these curves correlate with listening experience, in this or other setups? Are these measurement really useful at all?

As a refresher maybe have a look at the second and third linked teaching Video. The measurements were taken as explained in the 1st post of this thread. It appears that the OP is doing this as a fun experiment and a means to compare the FR sweeps from new production speakers against some older vintage speakers. Mostly fun using a scientific approach.

Master List of Instructional Teaching Video presentations (How to read and interpret/understand all the Graphs and Charts!

Collection of Amir’s instructional Videos for reference, refresher and new members: (“Behind the Curtain @ ASR University”) Welcome New and returning Members of ASR. Our Host Amir has put together a collection of comprehensive instructional videos that might help new and returning Members...

www.audiosciencereview.com

 

[AdamG]

Yes i have spent hundreds of hours doing similar job, and so have learnt that are totally useless...

Really but you spent hundreds of hours anyway.

Gated measurements in order to get a semi anechoic response free of most room artifacts would be a huge improvement, used by most diyers...

We can’t wait to see the results of your project.

 

[AdamG]

I thought this group would be interested in how some vintage speakers compare to each other under the same measuring conditions. I have a collection of vintage speakers and simple measuring setup. No Klippel machine for me! I measured the frequency response of each speaker with:

• Laptop PC with Creative soundcard
• REW software
• Dayton UMM-6 microphone, with calibration loaded in REW
• Random Denon amplifier
• 1 meter distance, on the tweeter axis
• 20x24 garage with hard floor and 9.5' ceiling
• Small amount of damping on the floor

All of the measurement suffer from a floor bounce dip at ~80-90hz depending on the height of the woofer and mic. Sorry about that.

Here's the setup.
View attachment 360204

Here's the collection.
View attachment 360206

Here's all the graphs together. I'll post each frequency response curve separately.
View attachment 360208

Awesome effort and thanks for sharing your measurements with us. Bravo Zulu Sir and Happy Easter.

 

[bconline]

Thanks for sharing, but just to clarify, could you please specify whether the speakers were all measured at the same height or not?

For that matter, please also indicate whether you applied smoothing and what SPL you used. Thanks!

The mic position was tweeter height for each speaker. I mounted the monitors and stand-mount speaker on a 27" cabinet (see picture in the original post).

The graphs I posted use 1/6 octave smoothing. I have the original .mdat files and could compute a different smoothing. My post #66 has all the files. I don't know the actual SPL level. It was pretty low, may be ~75db.

 

[Toni Mas]

Really but you spent hundreds of hours anyway.

We can’t wait to see the results of your project.

Will take my time. No hurry!

 

[soypeter]

Nice to see a fair amount of measurements. Good job @bconline !

 

[bconline]

It appears that the OP is doing this as a fun experiment and a means to compare the FR sweeps from new production speakers against some older vintage speakers. Mostly fun using a scientific approach.

Exactly! This is just for fun. I don't have what it takes to do the rigorous testing that Amir and others do. I started this measuring project becuase I wanted to compare two legendary studio monitors from the '80's - the Yamaha NS1000M and JBL 4311. Here's that graph, with the Yammy in Red and JBL in Teal.

 

[Curvature]

So in conclusion with regards to the information provided, there is a wealth of information and PROOF that just because you've taken hundreds
of hours of measurements you still need to know what to look for. Off axis is one of Danny's (at GRs) favorite gripes. What avid listener listens 15, 30 &
45 degrees off axis? I don't know of a single listener unless they are playing with AVR/AVPs and several speakers in a surround system. So we are
clear. Surround sound is not the norm for listening to music. SO off axis is just a measurement most good stereo speakers don't require

Right, because sound is like a laser beam. Goes straight and does not deviate. Walls? Whatever.

 

[bconline]

In the modern v.s. vintage, here's two big guys. Revel F206 in Red versus JBL L100T in Purple.

 

[Rick Sykora]

The mic position was tweeter height for each speaker. I mounted the monitors and stand-mount speaker on a 27" cabinet (see picture in the original post).

The graphs I posted use 1/6 octave smoothing. I have the original .mdat files and could compute a different smoothing. My post #66 has all the files. I don't know the actual SPL level. It was pretty low, may be ~75db.

First, you want the tweeter height to stay fixed at about half the ceiling height. Otherwise, your floor bounce is changing with each speaker you measure. Also, this makes for easier, more consistent measurements between speaker sweeps. From your mdats, your gate should be set to 10-11 msec.

If you do not have a SPL meter to calibrate a level, you should set a level that is fairly high (by ear). If you do not or environment noise is high, you will not get above the noise floor and some measurements will be unreliable (as per the distortion discussion with @Juhazi). Once you find a reasonable level, should keep it the same for each speaker or cannot even do relative SPL comparisons between the speakers.

Your supplied mdats are showing 1/12 smoothing, 1/6 is a bit much. If you do the gating, should be able to do without additional smoothing, but otherwise 1/12 or 1/24 is preferred.

Did you calibrate your soundcard? If not, your measurements will be less accurate (but still can be relatively compared).

Hope this helps! Hope this does not dampen your fun too much, my wife is the fun one in our family.