Room measurement, what to learn?

[ToTo Man]

Tony,

I spent the day re-measuring my Lancasters before swapping them out for my Lockwoods, so I've uploaded the results in case you're interested in comparing to yours. Please note, all of the following graphs relate to my Lancasters, I'll be measuring my Lockwoods tomorrow.

Graph 1A - Lancasters measured near-field (1 metre) with mic centred on compression horn, 1/24 smoothing:

Graph 1B - Lancasters measured near-field (1 metre) with mic centred on compression horn, 1/3 smoothing:

Graph 2A - Lancasters measured far-field (from listening spot) with mic 1.2m above floor, 1/24 smoothing:

Graph 2B - Lancasters measured far-field (from listening spot) with mic 1.2m above floor, 1/3 smoothing:

Graph 3A & 3B - Just to prove that my supertweeters are of benefit, Lancasters measured far-field (from listening spot) with and without ST-100 supertweeters, mic 1.2m above floor, 1/3 smoothing:

 

[BE718]

Hope this doesn't offend, but ditch the super tweeter, you can clearly see it puts a sting into the upper frequencies. It is not well integrated with the existing tweeter.

Super tweeters audible effect is because they interfere with the normal audible range. It's perfectly normal and correct for the
measurement to dip at high frequencies.

 

[Tenson]

Can't agree with that, the supertweeter looks both well integrated and having a benefit. The test is what happens off-axis.

 

[ToTo Man]

Can't agree with that, the supertweeter looks both well integrated and having a benefit. The test is what happens off-axis.

Just to confirm, the final tests (3A & 3B) with and without supertweeter were done far-field at my listening spot, i.e. off-axis (I only have a few degrees of toe-in on my speakers). The supertweeters were not active in tests 1A, 1B, 2A and 2B.

 

[BE718]

Can't agree with that, the supertweeter looks both well integrated and having a benefit. The test is what happens off-axis.

I agree that off axis is very relevant, however you won't find many in room measurements at the listening position show a hump at those frequencies. They will be tailing off.

More to the point, why would you expect a bolt on tweeter to integrate properly with any particular speaker?

There will be all sorts of issues, crossover point, slope and level. Is the ST the correct distance from the existing tweeter? Lobing characteristics at xo?

You heAr a difference (not necessarily an improvement) due to the ST effect in the normal audible band. This is unlikely to beneficial or correctly integrated.

 

[Tony L]

Very interesting. I'll redo mine at 1m at some point to see how we compare. My close measurements were really close. Possibly too close:

My aim was to remove as much of the room as possible, but I suspect I'm so close a difference of a mm or so will screw things up given the directivity of the compression driver. I did it this way as I felt that by locking the mic stand down firmly and placing it precisely against the bottom of the cab I'd get the same each side, but I have a feeling it wasn't a great idea.

 

[ToTo Man]

More to the point, why would you expect a bolt on tweeter to integrate properly with any particular speaker?

There will be all sorts of issues, crossover point, slope and level. Is the ST the correct distance from the existing tweeter?

Tannoy ST-100 Supertweeters, designed to work with current Prestige and vintage Tannoy DC drivers, so not exactly bolt-on! It has a choice of crossover frequencies and SPLs so that you can match it to the DC driver you are using it with.

And yes, I have carefully followed Tannoy instructions to align the ST with the compression driver at the rear of the throat of the DC driver (though I admit I may be a couple of mm's out).

 

[zippy]

quote: "Umik -1 is cheap as chips "

You must have more spare cash than me - I respectfully suggest that only an enthusiast would spend £80+ on a microphone that may only get used a few times.

No disputing that they're good though, and that their integration with REW makes measuring a lot easier.

 

[BE718]

Tannoy ST-100 Supertweeters, designed to work with current Prestige and vintage Tannoy DC drivers, so not exactly bolt-on! It has a choice of crossover frequencies and SPLs so that you can match it to the DC driver you are using it with.

And yes, I have carefully followed Tannoy instructions to align the ST with the compression driver at the rear of the throat of the DC driver (though I admit I may be a couple of mm's out).

Thanks, I stand corrected, my presumption was a third party driver, apologies. However this doesn't change my opinion of it further increasing hf output which already has a fairly significant hump.

Btw was the mic pointed at the speakers of at the ceiling?

Also can you post the ST plots with less smoothing, say 1/24?

Thanks

 

[BE718]

quote: "Umik -1 is cheap as chips "

You must have more spare cash than me - I respectfully suggest that only an enthusiast would spend £80+ on a microphone that may only get used a few times.

No disputing that they're good though, and that their integration with REW makes measuring a lot easier.

Well considering that many will spend hundreds or tmaybe thousands replacing kit/ speakers when the problem is the room then I would say the information it provides for the price is an absolute bargain.

 

[Tony L]

ToTo Man, just to clarify were your supertweeters on or off for the first 1m measurement? (I'm assuming off).

 

[Bemused]

Very interesting. I'll redo mine at 1m at some point to see how we compare. My close measurements were really close. Possibly too close:

My aim was to remove as much of the room as possible, but I suspect I'm so close a difference of a mm or so will screw things up given the directivity of the compression driver. I did it this way as I felt that by locking the mic stand down firmly and placing it precisely against the bottom of the cab I'd get the same each side, but I have a feeling it wasn't a great idea.

I can see why you placed the mic stand where you did but thats quite a chunk of steel work real close to the driver. Would be interesting to see results with the boom right out and the stand out of the way.

 

[ToTo Man]

Thanks, I stand corrected, my presumption was a third party driver, apologies. However this doesn't change my opinion of it further increasing hf output which already has a fairly significant hump.

Btw was the mic pointed at the speakers of at the ceiling?

Also can you post the ST plots with less smoothing, say 1/24?

Thanks

Mic was pointed horizontally at speakers. Lots of ping pong debate on various forums as to whether the ECM8000 performs more like a direct or diffuse field mic without a definitive answer (majority think direct field).

Here's the 1/24 smoothing graphs as requested. I'm not sure how much accuracy you can place on these measurements to be honest, especially the stereo reading, due to mic limitations and/or comb-filtering effects:

PS - I suspect my ultrasonic hearing is not as sensitive as it was when in my teens as it definitely does not sound toppy to me when the supertweeters are on. I currently have them set at the maximum SPL (95dB/1w) and minimum crossover frequency (18kHz) so I could easily pad them down if need be.

 

[ToTo Man]

ToTo Man, just to clarify were your supertweeters on or off for the first 1m measurement? (I'm assuming off).

Correct, supertweeters were off for all measurements except 3A & 3B.

 

[davidsrsb]

I can see why you placed the mic stand where you did but thats quite a chunk of steel work real close to the driver. Would be interesting to see results with the boom right out and the stand out of the way.

This is why measuring microphones use a long wand, the same diameter as the microphone element. That too close stand will be giving serious reflection errors

 

[Tony L]

It hadn't even occured the me I'd get a reflection from the stand itself, but that makes sense and I have therefore learnt something! I'll do the close measurement again from a metre and with the boom extended.

 

[Bemused]

This is why measuring microphones use a long wand, the same diameter as the microphone element. That too close stand will be giving serious reflection errors

Yes PD recomended me the Panasonic electrect a few years ago and I picked some up for pennies but switched to the UMIK1 as it integrates so easily with REW. Glued mine on the end of a nylon tube.

 

[Tony L]

Here's a 1m measurement with the mic right out on the boom and no stand angles that would reflect:

One metre is a long way in a room as small as mine so I'd expect this to have a lot of room artefacts. I also can't be 100% certain that the mic position was perfect as it is very awkward to measure things like this to mm accuracy with regards to axis etc (Tannoys beam almost as badly as Quad 57s), but I'd say it was within a cm or two. I also used the same amp channel for both measurements, i.e. swapped the speaker lead thus ruling out any volume balance or preamp tube issues. It is pretty obvious the pair matching is to usual vintage Tannoy standards, i.e. all but non-existant (this is my third and by far best pair of Monitor Golds!) All the measurements to date show anything up to a 5db treble lift at certain points on the right. I can't hear the main spike due to hearing damage/age loss, in fact I can hear very little if anything over about 10-12k these days. Even so after visiting the RNCM a couple of times recently for some classical concerts I do think they may be just a little bright so next time I can be bothered undoing all the bolts to get the back doors off I'll knock the roll-off control back a notch. which should tame the spike at the very top (which as I say is almost certainly inaudible to me). The level and roll-off controls are all set to 'level' at present.

The odd thing with this is I suspect the imperfection in balance actually helps me so it is probably unwise to try and address it. I have with all my speakers tried each of the pair in the left and right position and selected the ones that give me the most central stereo image. Even the JR149s with their brand spanking new and factory paired falcon units and rebuilt crossovers have the odd 2db dip and peak, so picking the right one for each side does make a difference if your hearing is flawed. I know I have a slight notch in my right ear at about 5-6Khz which does impact stereo imaging, so even a slight speaker imbalance here can be a disaster/beneficial depending on which side of the room you stick it!

PS I bet I am far from alone in the above, the only difference being I have gone to some lengths to test it, e.g. I have spent some time with headphones listening to how well I perceive different frequencies. Fabre Acoustical's Sound Meter app for iPhone has a great signal generator where you can sweep a sin-wave anywhere you want and also adjust the balance. I also have a Minimoog Voyager which is a far more expensive way of achieving the same thing! It is very revealing/depressing to get it to a weak area and then swing the balance slider so you think you have a central image and then realise you are say 25% out! Getting old really sucks!

 

[Robert]

We probably worry too much about perfect balance on the channels, both in terms of SPL and response. A live performance will vary considerably in terms of balance depending on where you sit, but it rarely concerns or spoils enjoyment.

My ESLs always sound a tad brighter on the left side thanks to more reflective surfaces on that side of the room, and something as simple as sitting on the right hand seat of the sofa with an expanse of shiny leather to the rear of my left ear.....

So many seemingly trivial things in the room can alter balance, so no point getting too OCD over it IMO.

Tony can you show a 1/3 smoothed version of the above plot?
We don't hear all of those peaks and dips and a smoothed response more closely replicates how 'speakers actually sound in-room.

 

[davidsrsb]

Much cleaner ~10kHz now Tony, which is where stand reflections happen.
I wouldn't worry about the >10kHz peak as beaming is reducing the room average up there