Rode NT1 okay for measuring?

[Mark DuBerry]

I have a very small home recording studio which I acoustically treated last year with DIY rockwool panels according to Ethan Winer's and others' guidelines. The room is 10'x6.5'x8'.

I recently bought a Radio Shack 33-4050 SPL meter in order to measure the rooms acoustics / frequency response using REW. I have since discovered through reading some posts on this forum that this meter is only good for measuring the low frequencies. I presume that is not enough for my room and that I need something that will cope with the whole spectrum(?). As you can tell, I haven't used REW before.

I have a Rode NT 1 which is a good quality cardioid capacitor mic with a frequency response of 20Hz-20kHz. Will this suffice for taking the measurements or do I need to go out and buy a special measuring mic?

Thank you.

[brucek]

Do you have a calibration file or graph of the microphone? If so, you can create a calibration file that will work with REW.

brucek

[Mark DuBerry]

Hi brucek. Thanks for your reply. That's good news but no, I don't have the file or graph. Is that something I can generate myself or do I have to get one from somewhere? How do I go about doing that? I will try and find this out myself by looking at the help files etc. but I would be grateful for any further help if possible.

Mark

[brucek]

Well, a calibration file something is that will come with a microphone to show it's frequency response.
If you use a mic or meter that we have tested on the site, then you can get away with using those calibration files (if you own one of those), with an assumption that they're accurate enough for home use. We have cal files for all the Radio Shack meters, Galaxy CM-140 meter and the ECM8000 microphone. The Radio Shack meters shouldn't be used much beyond 1000Hz as it seems as they aren't accurate past that. The ECM and Galaxy are full range. The ECM requires a preamp though.

If you don't have a cal file for the microphone you own, then you have no way of knowing its accuracy I'm afraid.

brucek

[Mark DuBerry]

Hmm. I'll see if I can locate a cal file through Rode. Alternatively, I may pick up an ECM8000. I want to measure the room mainly to see how effective the treatment I have used is but also to find if there are any problem frequencies needing further attention, particularly in the lower end. I have a lot of bass trapping as it stands. Would you happen to know how necessary it would be to measure beyond the 1000Hz or would the fact that the RS meter will give me a good idea of what's happening in the lower frequencies be a good enough test for my purposes?

Thanks again.

[brucek]

Well, my understanding is that treatment doesn't do much above ~500Hz and does little for modal problems below ~80Hz (where eq is usually employed), so I see no reason not to use the Radio Shack meter. They're suitable for that area..........

brucek

[Mark DuBerry]

That's great brucek. I'll do that and maybe post the results to get some help interpreting them. Meanwhile, I'll try and hunt down a file / graph for the NT1 to maybe compare the results obtained with both the mic and the meter.

many thanks,
Mark

[Mark DuBerry]

Actually, I'm having a bit of a job finding a fre. resp. graph or a cal file for the NT1. The Rode website has one for the NT1A but it's a different design with an improved response so I don't think it's any good re. the NT1. I wonder if anyone knows where I can find one?

I will use the RS meter as brucek suggests but I would like to test a wider fre. band if possible as well without buying another mic.

Thanks, Mark

[brucek]

You can always have your microphone calibrated if you feel it's a decent mic. It can be expensive though. I suspect you could get a Galaxy meter for the same price. Sometimes stage mics don't make great measurment mics if the response is set for voice. I guess it needs external biasing, so some type of preamp is needed?

brucek

[Mark DuBerry]

I don't think it's an option to get the mic calibrated. I could buy the Behringer or the Galaxy at that rate. The NT1 is a very good mic and is regarded as fairly flat; not sure what you mean by a 'stage' mic except it is of course a recording mic and is not specifically for vocals; it works well with most applications. I imagine it would do the job well if I could get a cal file together.

I'm going to contact Rode by phone just as soon as the time zone brings them into opening hours and do my best to get a graph. If not, I may have to work within the limitations of the RS meter.

Thanks brucek. I do appreciate your help with this.

Regards,
Mark

[Wayne A. Pflughaupt]

Mark DuBerry wrote:

Actually, I'm having a bit of a job finding a fre. resp. graph or a cal file for the NT1. The Rode website has one for the NT1A but it's a different design with an improved response so I don't think it's any good re. the NT1. I wonder if anyone knows where I can find one?

You might try the Tape OP Message Board. I doubt many people here are familiar with this mic.

Regards,
Wayne

[Mark DuBerry]

Hi Wayne. I'll check out that link. Meanwhile, here's a copy of the spec page for the NT1 and the fre. resp. graph for the more recent version, the NT1A. I spoke to a guy in Rode and he said the two mics are similar enough to use this graph with reasonable accuracy. He said the main difference is in the noise level and that the frequency response is similar. Also, it is not flat but I gather from reviews like this one http://www.soundonsound.com/sos/1997...7/rodent1.html (see the section immediately under the heading ¨Rode test¨) that the enhancements are quite subtle, which seems to be reflected in the graph.

Spec sheet:



Frequency response graph:



Does it look like I could use this mic? If so, can you tell me how to create a cal file from this graph, or else point me to where I might find that out?

Thanks a lot,
Mark

[brucek]

It's very simple Mark.

It's up to you what type of frequency divisions you want to use because REW extrapolates between.

At the extremes REW will extend the last values.

If the mic reads low a negative value is entered and positive for positive.

You create a .txt file (with notepad) with frequency and decibel level. Then save the file and rename the extension to .cal

You'll have to decide from the graph what the values are.

For example here is a small sample of a cal file.

10.00 -9.23
12.50 -6.64
16.00 -4.68
19.95 -3.09
25.12 -2.08
31.62 -1.30
39.81 -0.83
50.12 -0.41
63.10 -0.14
79.43 0.00
100.00 0.15
125.89 0.22
158.49 0.22
199.53 0.26
251.19 0.26
316.23 0.26
398.11 0.22
501.19 0.20
630.96 0.14
794.33 0.07
1000.00	0.00
1258.93	-0.07
1584.89	-0.13
1995.26	-0.23
2511.89	-0.27
3162.28	-0.13
3981.07	-0.22
5011.87	-0.30
6309.57	-0.25
7943.28	-0.17
10000.00 -0.15
12589.25 -3.77
15848.93 -7.48
19952.6 -9.01

So decide what the vertical divisions on the graph are and use those values.

I would simply enter at the logarithmic values from the graph for the frequency divisions.

For example :

20 -2.00
30 -1.50
40 -0.50
50 0.00
60 (I'm sure you get what I mean):
70
80
90
100
200

create the file, and name it NT1.cal and load it into REW and see what it looks like. You'll see if it looks like your graph or not....

Questions?

brucek

[Mark DuBerry]

That's excellent thanks brucek. Very detailed and clear.

I think I've got the idea. I opened up the cal file for the RS meter in notepad. Do I copy the format there in terms of spacing between the 2 values? Also, that file goes from 7.00 to 200.00 Hz. I presume I enter in all the values between 20Hz and 20kHz in the new file, using the divisions you suggest from my graph?

Does it look to you like I should be able to use this cal file ok with my mic to go ahead and take the measurements?

Many thanks again,
Mark

P.S. The other thing the guy from Rode said was that if I wanted more accuracy, I could test the mic myself and generate my own graph. How would I do that and would it be easy enough / worth it? I know this is being picky but I just want to get it right though I imagine the graph above would be good enough to use. Thanks.

[brucek]

Do I copy the format there in terms of spacing between the 2 values?

You simply need a space between the frequency and amplitude values.

I presume I enter in all the values between 20Hz and 20kHz in the new file, using the divisions you suggest from my graph?

Yeah, just use the log values from the graph. That would be accurate enough. When you're done, and you take a look at it in REW, if you want more resolution, just add it, no problem.
At the extremes, REW will continue the last value. For example, if at 20Hz you have a -2.00, then that value is used down to 0Hz. Above the highest value it will also continue the last value - but after 20Khz, you don't really care.

Does it look to you like I should be able to use this cal file ok with my mic to go ahead and take the measurements?

Definitely.

brucek

[Mark DuBerry]

Great! I'll get down to it tomorrow, given it's 1.30 am here.

Again, I am very grateful for your help brucek.

No doubt I'll be back again with more questions when I've done the measurements!

Mark

[Mark DuBerry]

Hi brucek

I've created the cal file but when I load it into REW I can't see how to adjust the frequency and amplitude values so they look the same as my graph. Also, the 0 amplitude point seems to be fixed at 75 and I can't see how to change that. How do I change the horizontal and vertical values so they look the same as my graph? Sorry if I'm being slow here; it's just I can't see if the graph matches my own in order to confirm whether or not I've created the cal file correctly.

Thanks,
Mark

Edit: I thik I've got it working well enough now. I tried to delete this post but couldn't see how..

[brucek]

Also, the 0 amplitude point seems to be fixed at 75

And so it should. It revolves about the target level. If the target is 75, then that is zero.

how to adjust the frequency and amplitude values

I guess you figured it out. The Graph Limits icon popup can be used to set any X or Y axis....

Remember when actually taking measures and posting graphs to use a standard of 45dB-105dB vertical.

brucek

[Mark DuBerry]

It revolves about the target level. If the target is 75, then that is zero.

I don't understand that, i.e. what the 'target level' actually means but I probably just need to dig further into the programme and the help files to find out. I'm working on it tonight and in my spare time so hopefully I'll get to the actual measuring soon!

Remember when actually taking measures and posting graphs to use a standard of 45dB-105dB vertical.

Will do. I'll read through all the faqs etc. on the forum to find out the guidelines and do my best to apply them.

Thanks brucek,
Mark

[Mark DuBerry]

Hi brucek, or whoever else is reading this. I have done some measurements using both the RS 33-4050 meter and my Rode NT1 mic. Here are the results. I hope they're in the correct format; I used a wide frequency range, 20Hz - 20kHz in order to cover the full spectrum.

RS meter:




NT1 mic:



I didn't apply any smoothing. I'm not sure if that would have been a good idea or not(?)

I'm also not sure how to interpret these results so I would be grateful for any help. It looks to me like there are quite a few dips and peaks but I've no idea how good or bad a response this is. I won't be applying any eq; this is to find out how my already-treated room is responding and to make any necessary changes to that treatment. So once again, any recommendations around that would be welcome.

If I need to go back and change the format / appearance of these graphs and upload them again, please let me know. Also, would it be useful for me to post the impulse response graphs or waterfalls as well?

Many thanks,
Mark

Actually, I decided to post the waterfalls as well since the low frequency behaviour of my room is of particular interest, given I have put up quite a lot of bass traps.

RS meter:



NT1 mic:



How do these look?

[brucek]

How do these look?

Just a few comments first on the graphing itself.
You're being kind to yourself by using such a large vertical scale. When you redo them and use a vertical scale of +45dB-105dB, you'll see they actually appear a bit worse.

Then you can see that the NT1 mic is better (no typical rise at 5K). Stick to the NT1 for these full range. It's a much better mic. Don't bother with the RS meter other than to set the general 75dB target level to get started with a calibration of REW's meter.

You can see the low end response is actually meaningful from your speakers down to about 18Hz where it drops into the noise. The information below that is not useful (it's noise), so use a horizontal scale of 20Hz-20KHz.
Change the waterfall to log scale. (the horizontal axis from 20Hz-200Hz is perfect).

Now, I would say that if you didn't apply any smoothing to these plots, you have an amazingly well treated room with very little reflection. It shows almost no comb filtering, and between 60Hz-200Hz the waterfall is great. This is the area that usually responds well to treatment as you've obviously done.

From 60Hz down to the bottom end, you have some modal issues that would only respond to EQ as treatment would be too large.

You do have a pesky dip at 100Hz though.

brucek

[Mark DuBerry]

When you redo them and use a vertical scale of +45dB-105dB, you'll see they actually appear a bit worse.

I thought I'd used the default vertical and horizontal scales that load with REW and as specified in the guidelines here. I didn't notice that they had changed, though it's obvious now that you point it out. Not sure how that happened.. When you say redo them, I presume you mean change the scales on the existing graphs rather than re-doing the actual measurements?

In case you meant the former, here are the two graphs for the mic only, scaled to the specified resolution (I hope!)







Yeah, I can see much more detail now. That is a nasty dip at 100. Also, 3kHz to 20kHz looks disastrous to me. Is that to do with the room or the mic, or a limitation in measuring up that far? Anything to worry about?

Now, I would say that if you didn't apply any smoothing to these plots, you have an amazingly well treated room with very little reflection. It shows almost no comb filtering, and between 60Hz-200Hz the waterfall is great. This is the area that usually responds well to treatment as you've obviously done.

This is good news and confirms both the effectiveness of the existing treatment and (hopefully) the relative accuracy of the measurements.

From 60Hz down to the bottom end, you have some modal issues that would only respond to EQ as treatment would be too large.

You do have a pesky dip at 100Hz though.

Not so good news; not sure what to do about those two issues but I will enquire further.

This is very helpful anyway brucek and is more positive than I thought it would be.

Thanks again and apologies for the incorrect scaling of the original graphs,
Mark

[brucek]

scaled to the specified resolution (I hope!)

Perfect....

That is a nasty dip at 100. Also, 3kHz to 20kHz looks disastrous to me. Is that to do with the room or the mic, or a limitation in measuring up that far? Anything to worry about?

Usually full range measurements require smoothing (at least 1/3 octave) to reveal the underlying trend of the response. As I said, you have treated it well and an unsmoothed response shows that, but there's nothing wrong with adding some smoothing to get a better feel for it.

I suspect the mic is fine, and I suspect your speakers are fine and any problems (as usual) are attributed to the room. You could take one of your speakers into the middle of a large room (or even better outside) and do a near field response to get a better idea of your speakers response without the room influence. A foot or so away would be fine for the mic.

Are you measuring two speakers with the system in stereo, or is this just one speaker?

The dropping off after 10Khz though is likely the speakers response (don't you think?).

As this is a studio, where are you measuring from? I suppose it's somewhat near field? as opposed to a home theater that would be measured back much further on a couch. You must have the mic near the mixer chair fairly close to your monitors? Can you move the monitors around at all to try and clean up the 100Hz dip at least?

brucek

[groundie]

mark
i too have a Rode NT1 but did not use it.
can you post the cal file you created for it?
as noted, you'd expect NT1 to be a far superior mic.
i do not have BFD but their crossover product DCX2496;
anyway, after quick initial round of measurements,
i experimented with couple of filters but ended up liking
no EQ at all.

i plan to re-measure and replay with NT1 cal file to
get the full range response.
cheers ...
g

[Mark DuBerry]

Usually full range measurements require smoothing (at least 1/3 octave)

Okay. So here's the graph again with 1/3 octave smoothing:

do a near field response to get a better idea of your speakers response without the room influence.

Great idea. I'll consider that, though it would mean moving a load of gear around. I just checked the monitor specs. They are Dynaudio BM5A's in case you know them, very good quality flat near-fields. Their stated fre. resp. is 50Hz - 21kHz. Could that be the reason for the problems from 60Hz down rather than modal issues in the room (I'm forever hopeful!)? The drop off after 10kHz - well, should they not be good up to higher than that if they state 21kHz as the upper fre.? Or perhaps they do begin to drop off at the top end more gradually. Either way, yes, it would a good idea to test them independently of the room.

Are you measuring two speakers with the system in stereo, or is this just one speaker?

This is both speakers with the same signal routed to each of them. I connected the right output only from the computer's soundcard to one channel (with the pan pot centred) of my mixer which then feeds the 2 monitors in stereo, though this is obviously a mono signal. I did this because I am attempting to measure what I am actually hearing when I am mixing. Would I be better to measure with each speaker separately?

As this is a studio, where are you measuring from? I suppose it's somewhat near field?

It's very near-field actually. I have established my listening position at a point 38% into the room, which is the recommended spot. Given the room is only 10' long, that's only 116cm away from the front wall. Because of this and because of the fact that there are front corner bass traps pretty much immediately behind the speakers at an angle, I have had to position them only 80cm apart in order to have an equilateral triangle with my listening position. I placed the mic as exactly as I could at that position, forming an equi. triangle with the monitors in order to simulate the setup when I am mixing etc.

Can you move the monitors around at all to try and clean up the 100Hz dip at least?

Due to the above, I am very restricted. I can't move them further apart and back towards the front wall because of those traps being 'in the way'. If I move them forward into the room, I lose the triangle and/or the 38% position, which I have been told is very important to preserve in a room this size. So as far as I can see I'm pretty much stuck with this speaker placement. I'd love to be able to do what you suggest though so if you can see a way I could make that work then great.

Here is a photo of the front of the room taken a year ago wth my old monitors but it's pretty much the same setup now. That should give you an idea of what I'm dealing with. I have since put a plasterboard sheet up in front of that door behind the left monitor and at the same place that the wall should be in order to equalise the two sides some more and to give the bass frequencies that pass through the traps a similar surface (and same distance) to 'bounce' off:



Pretty ain't she?! Actually it's not as bad / narrow as it looks. As you can see there's a fair amount of treatment. It's similar at the rear, without the 'cloud'.

One more thing: I decided to measure my room because of the lack of satisfactory translation of my mixes to other systems. I noticed they were sounding good in my room but muddy and indistinct elsewhere. That perhaps indicated problems somewhere in the lower frequencies, which would confirm the discovery of that 100Hz dip and perhaps the issues from 60Hz down(?)

That's it for now!

Thanks,
Mark