I finally bought a measurment microphone to be able to guess a bit less when building speakers. To use what's offered I also did room measurements that seem off. First the gear:
Dr. Jordan microphon (IEC60651 Class 2)
RME Fireface UCX interface (set to 96kHz)
FuzzMeasure 4 software, loopback function enabled
Class D amplifier board, flat from 20 to 20k
The microphone was placed pointing downwards onto a carpet, room center distance to carpet 1.5 meters.
Plots:
I hear a distinct slap echo after the sweep is done and always was aware of a very distict slap echo when clapping my hands. When doing the measurements I heard the sweep ring for about two seconds after the sweep. Neither the waterfall plot nor the RT30 plot seem to show any of that, right?
Am I misinterpreting the graphs or just expecting too much?
I would be happy if you could could give me some input on that matter. If you need more details, let me know.
*Dear moderators, I didn't know where else to post this, if there's a more suitable subforum please move it. (hoping this doesn't turn into one of these that end up in the Lounge 😉)*
Dr. Jordan microphon (IEC60651 Class 2)
RME Fireface UCX interface (set to 96kHz)
FuzzMeasure 4 software, loopback function enabled
Class D amplifier board, flat from 20 to 20k
The microphone was placed pointing downwards onto a carpet, room center distance to carpet 1.5 meters.
Plots:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I hear a distinct slap echo after the sweep is done and always was aware of a very distict slap echo when clapping my hands. When doing the measurements I heard the sweep ring for about two seconds after the sweep. Neither the waterfall plot nor the RT30 plot seem to show any of that, right?
Am I misinterpreting the graphs or just expecting too much?
I would be happy if you could could give me some input on that matter. If you need more details, let me know.
*Dear moderators, I didn't know where else to post this, if there's a more suitable subforum please move it. (hoping this doesn't turn into one of these that end up in the Lounge 😉)*
An externally hosted image should be here but it was not working when we last tested it.
Yes It's shame when that happens. I don't think they realise the value of some of the threads that are placed In the Lounge.
BTW - Can't see any of your pictures.
These are too small for me to read, but a slap echo you will see a dip and re-appearance. I do see that towards the lower part of the range, just below the second numbers I can't read.
The first signal has a strong bass mode. See how it vanishes and comes back a couple of times?
Best,
E
The first signal has a strong bass mode. See how it vanishes and comes back a couple of times?
Best,
E
Sorry, look at the initial plot. On the left is the highest point, that's the bass mode I see. now follow the Z axis towards you and you see it go away and return.
I see theLF mode, the slap echo is more in the 1kHz range.
I tried three image hosters and with two nothing shows up and with the third the thumbnails don't get bigger when clicked...
I tried three image hosters and with two nothing shows up and with the third the thumbnails don't get bigger when clicked...
So the slap echo doesn't necessarily show as a long extension in the z axis, but more in the fact that it decays and comes back again?
Yep. If you can hear it as a distinct set of "bam bam bam"
DIYAudio does it's own image hosting. Use big pictures, and click "Go Advanced" and upload your pics. Then once uploaded you'll see a list of pics. Right click, copy link, and then you can use the image icon.
Best,
Erik
DIYAudio does it's own image hosting. Use big pictures, and click "Go Advanced" and upload your pics. Then once uploaded you'll see a list of pics. Right click, copy link, and then you can use the image icon.
Best,
Erik
Unless you have your room or speaker calibrated when you measure a speaker in a room it is a combination of the two.
If you want to see how the room affects the speaker the ETC (Energy Time Curve) or spectrogram graphs are very useful. In the ETC you will see the impulse and then later in time the reflections. The spectrogram will give a different representation.
The length of the sweep you use will have a big impact on what you hear when listening to it. If you use a short sweep at a high level a very large amount of energy is output in a very small time, in a room that sounds really loud and is likely to excite the echo. Use a longer sweep and it won't sound the same when you listen to it.
Without seeing the scale of the graph it's hard to judge but the RT60 seems quite even and if it is the range of 400ms then it is a fairly normal room without much treatment. Your room won't be ringing for 2 seconds unless you live in a concert hall 🙂
A waterfall or CSD graph can be very tricky to interpret because it depends very much on the window settings you have chose. In all the graphs there is a trade off between frequency and time accuracy. When one is more accurate the other is less.
REW Room EQ Wizard is free measuring software that has good graphs and is familiar to a lot of people on this forum. It is free, try it out.
Is your microphone calibrated and do you have a 90 degree file? If your mic is calibrated for 0 degree (pointed straight) and you use it at 90 degrees there will be quite a difference in the range above 1kHz, if it isn't calibrated at all that adds another unknown into the mix.
If you want to see how the room affects the speaker the ETC (Energy Time Curve) or spectrogram graphs are very useful. In the ETC you will see the impulse and then later in time the reflections. The spectrogram will give a different representation.
The length of the sweep you use will have a big impact on what you hear when listening to it. If you use a short sweep at a high level a very large amount of energy is output in a very small time, in a room that sounds really loud and is likely to excite the echo. Use a longer sweep and it won't sound the same when you listen to it.
Without seeing the scale of the graph it's hard to judge but the RT60 seems quite even and if it is the range of 400ms then it is a fairly normal room without much treatment. Your room won't be ringing for 2 seconds unless you live in a concert hall 🙂
A waterfall or CSD graph can be very tricky to interpret because it depends very much on the window settings you have chose. In all the graphs there is a trade off between frequency and time accuracy. When one is more accurate the other is less.
REW Room EQ Wizard is free measuring software that has good graphs and is familiar to a lot of people on this forum. It is free, try it out.
Is your microphone calibrated and do you have a 90 degree file? If your mic is calibrated for 0 degree (pointed straight) and you use it at 90 degrees there will be quite a difference in the range above 1kHz, if it isn't calibrated at all that adds another unknown into the mix.
I tried another time to attach images. Looks like it worked finally. Thanks for the heads up, Erik!
@fluid I'll try a longer sweep, maybe 10 seconds? Might doing sweeps with a limited frequency range be helpful? Like sweep 20Hz to 200 Hz in 5 seconds?
Another thing: when I wrote that I can hear the ringing of the room for two seconds after the sweep is done I only assumed that there is an issue with my measurements. Is it possible that my ears are ringing and not the room?
@fluid I'll try a longer sweep, maybe 10 seconds? Might doing sweeps with a limited frequency range be helpful? Like sweep 20Hz to 200 Hz in 5 seconds?
Another thing: when I wrote that I can hear the ringing of the room for two seconds after the sweep is done I only assumed that there is an issue with my measurements. Is it possible that my ears are ringing and not the room?
Attachments
Ha ha, yes possibly your ears. 🙂 To be honest, I don't see the 1K echo in your graphs. But I know the sound well, it's typical of bare rooms with gypsum board walls.
Do try REW, it will run on the Mac and is a common platform here. Plus you can upload your REW files for others to look at and analyze.
Do try REW, it will run on the Mac and is a common platform here. Plus you can upload your REW files for others to look at and analyze.
I don't see very convincing evidence for a slap echo except as I mentioned, just below 100 Hz is a little odd. It may also be noise we're looking at.
The graphs look pretty normal. Using longer sweeps is unlikely to change the graphs, it was more to demonstrate that the sound you hear is very different due to the time the energy is released over. I don't recommend listening to the sweeps unless you have to, particularly if they are loud. I set REW to start after a few seconds delay and walk out of the room. Limiting the frequency range is fine if you are only interested in that range but it won't change the result.
What is it that you are trying to achieve with measurements?
Pano brings up a valid point, lets focus on that room for a bit, what is it like?
Let me describe two scenario's: I'm living in an old house, dating back to 1927. It has some curtains and a wooden floor. A couch made of fabric, a carpet in the center. If I clap my hands walking trough the room I don't perceive a clear echo. It became even better after hanging 3 large damping panels in it.
When filled with more people talking I have no problem separating the conversations.
My sister lives in a more modern house. No curtains, leather couches, hard materials everywhere, concrete floors... just having a conversation at a birthday party while more people are speaking makes it obvious it's acoustics are largely different. I didn't clap my hands, walking around in that room, but to get good audio there I know it would need more work than my room.
Aside from the room, there's still a possibility the echo comes from within the measurement chain. But lets talk about the room first. Do you recognize something of the two scenario's I gave here?
- Status
- Not open for further replies.