So what could be the problem with the impuls? this way it's useless.
What impuls you refer? The gated impulse response is very useful and below the gating frequency the room dominates anyway, this is applicable to a real listening situation too.
Sounds like a great subscription business model for lending out crossover parts to rich people with plenty of spare time. Would you like to subscribe to my monthly delivery of the newest in inductors and caps, please? You send the old back with every new.
sounds like a bad business since you already got most parts needed, tweaking is not about testing components of the same value
For the rest of us I think design freeze is a understandable concept both for manufacturing, invoicing and also for the end user wanting predictable results both setup and performance.
yes, if you are a manufacturer. but this is diy and we have the luxury to work differently, i believe most of us strive for better performance all the time, we look for new drivers and parts and new ways to build even better speakers, what we did yesterday may not be valid today
So what could be the problem with the impuls? this way it's useless.
The solution has already been mentioned.
Regards
Charles
So I tried it with the blanket anyway. This is what the setup looks like. And the measured response so far. I've put a second blanket on the inside which is a more wooly blanket, I thought that would be effective against further minimizing the higher frequency reflections of the blanket.
Aren't there drawbacks to gated measurments? And my room is super small so the time until the first reflection will be very short. I will look in to how to do gated measurments.
Try making a measurement with the speaker lying on its side on the floor.
Put the mic about 1m away on the floor, where it's centered between tweeter and mid.
Block up the back of the speaker to tilt it where it aims perpendicularly to the mic.
Here's a pict to give you an idea. Was working on a modular PA build, and it was too cold to measure outside.
![mic on floor.jpg mic on floor.jpg](https://www.diyaudio.com/community/data/attachments/837/837535-c2bcb7a2300b0a5323dbdf32706fcbec.jpg)
Put blankets on nearby objects that might reflect into the mic, but get them away from the speaker (both objects and blankets).
Personally, I don't use gating...all it does is throw away data and lower frequency resolution.... to make graphs look better, ime/imo.
Just use smoothing...and then decide what/how to EQ.
I'm trying to understand what I have to do now. I want to setup REW to do a gated measurment. But then I've got to determine the time window to the first reflection. The impuls graph should look something like the first picture. But I have this huge never ending smear. What am I doing wrong?
Attachments
Try making a measurement with the speaker lying on its side on the floor.
That is not good because side position changes diffraction behavior. Significance depends on radiators and shape of enclosure. Another disadvantage is that off-axis/directivity measurements are not possible so it's driving as blind as with other design methods with on-axis response only.
Could be okay for quick QC measurement with long time window, but not a change for design phase.
I know there is a dB and %. I've got it on the % graph, just like they do on the MiniDSP website. I've also scaled it down, but it still shows this never ending mess.
Could it be that my room is just too small to even make gated measurments because the reflections come so early? In that case I'm screwed.
Could it be that my room is just too small to even make gated measurments because the reflections come so early? In that case I'm screwed.
Attachments
That is not good because side position changes diffraction behavior. Significance depends on radiators and shape of enclosure. Another disadvantage is that off-axis/directivity measurements are not possible so it's driving as blind as with other design methods with on-axis response only.
Could be okay for quick QC measurement with long time window, but not a change for design phase.
I've found it works much better than expected for making a decent on-axis response. Probably works considerably better for basic tuning, than the majority of in-room gated measurements made at listening position.
I know this through comparisons of this technique to outdoor measurements.
Outdoors, I use both a spinorama and multiple mics on a rotating boom, off a fairly open deck. So i do have an experienced basis for comparison.
I learned this speaker-on-floor, mic-on-floor technique, from an accomplished transducer designer...it's not something i made up....
I know there is a dB and %. I've got it on the % graph, just like they do on the MiniDSP website. I've also scaled it down, but it still shows this never ending mess.
Could it be that my room is just too small to even make gated measurments because the reflections come so early? In that case I'm screwed.
1 millisecond is about a 30cm in distance, 3 millisecond makes a meter. The gated signal must be shorter than any reflected sound paths from the room boundaries or other objects in the room. Measure without the blankets/tunnel, one meter away and gate the measurement before first reflection which should be easier to spot now that you don't have reflections from the blanket. Around ~2-4ms after the highest peak could be reasonable estimate where the first reflection comes in. You can measure /calculate using simple trigonometry what is the path length from nearest boundary to the mic. VituixCad has an auxiliary tool for this. There is also good documentation how to do measurements at home, with REW, in the VituixCAD documentation. If might feel overwhelming at first but thats about the easiest way to get proper results.
Attachments
I measured everything and the first reflection is about 230 cm distance (from speaker to reflection point to microphone) so it should be at about 6,7 ms. But it shows nothing on the graph. Probably because the graph for some reason looks like a fricking mountainscap. And that's what I'm not getting. WHY THE F DOES THE IMPULS GRAPH LOOK LIKE A FRICKING MOUNTAINSCAPE WHILE IT'S SUPPOSED TO BE FLAT.
Anyways, even though it's not visible on the graph, I'll just gate at 6 ms. Then by the way, how close to the wall do they have to be doing the measurments on the MiniDSP website if it's only 4,4 ms and already 6,7 ms in my room. Weird, probably a tiny office room.
Anyways, even though it's not visible on the graph, I'll just gate at 6 ms. Then by the way, how close to the wall do they have to be doing the measurments on the MiniDSP website if it's only 4,4 ms and already 6,7 ms in my room. Weird, probably a tiny office room.
You've got to subtract the difference between the first reflection path lenght and the direct path lenght. The floor or ceiling is often the closest surface.
Why and what do you mean exactly? Distance of first reflection - distance from speaker to mic?
What does the gating doe exactly? Because the graphs look very weird.
The issue is that there is NO gating applied to the blob-like impulse response that you show us. With more than 100 ms length it includes a lot of room reverberation.
Regards
Charles
Filtering it doesn't change the look of the impuls graph. It just puts a line at the selected ms.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Loudspeakers
- Multi-Way
- In need of warmth