Thanks Mark! I value your opinions/thoughts very much. If you have any suggestions or development ideas, I'm all ears. Hopefully we will see a certain dual 15" sub in our next session. :yes:Mark Seaton said:I want to give Illka proper credit for what I'm certain was a MAJOR undertaking. There is LOTS of very interesting data there. Definitely more information I'll add to my online "library."
bobgpsr said:Well Ilkka I do have a question. What happened to reverse sweeps? Do these test results just use the TrueRTA built-in (increasing freq) quick sweep for max SPL tests?
Reverse sweeps are exactly what I used also this time. I also did Quick Sweep, but its results weren't that much higher. I need to work something else for peak output tests.Max output level and power compression
Maximum long term output level was measured using a 30 second sine sweep from 100 Hz to 10 Hz. First sweep was level matched at 90 dB at 50 Hz. Drive level was rised by 5 dB after each sweep. There was around 20 s delay between each sweep. Sweeps were taken up to the point where the output level stopped rising anymore (5 dB or more of compression). I also tried 2 dB as the last step, but usually it didn’t help. This is a very demanding test and the results should not be compared to other tests with faster sweeps etc. Power compression graph is showing the relative compression to the 90 dB sweep.
Yes, this would probably be the best option, but it's PITA to measure. I mean it takes a lot more time than a 30 second sweep. It also requires much more attention since you have to carefully listen at the subs. If I could measure only a few subs in a day...but 27 in two days?Mark Seaton said:Ilkka,
By far the most appropriate way to do peak output tests is with the 1/3rd octave tone-bursts that Don Keele and Sigfried Linkwitz first put to use years ago. There is plenty of information on the web about the signals, but they are basically an enveloped sine wave of various filter types, typically 5-6 cycles long. I believe this is also what the new CEA testing is adopting for its stimulus. So far as actual limits, I would say both the limits defined by the CEA and simple audible thresholds would be quite useful. With these tones it is very easy to hear the onset of distortion and any resultant noises. If you look at the process Keith Yates used, he wrote down various observations (what he heard) alongside the numerical data.
It was a reverse linear sweep. Log-sweep didn't work with the THD plotting, that's why linear. And of course I looped it, that's essential with TrueRTA. And actually it needs a trick or two to even get it flat.To clarify about the reverse sweep signal, I am assuming it was a reverse log-sweep? The measurements appear to track what others have measured, but I am guessing you did loopback tests to confirm TrueRTA was reading the signal as flat, correct? I'm more asking to clarify for anyone reading along.
I do not know why I did not read reverse sweep from the 100 Hz to 10 Hz text. :blush:Ilkka said:Maximum long term output level was measured using a 30 second sine sweep from 100 Hz to 10 Hz.
Reverse sweeps are exactly what I used also this time. I also did Quick Sweep, but its results weren't that much higher. I need to work something else for peak output tests.
At least someone can appreciate what Keith Yates went through in his Way Down Deep report. Now remember that there were 2x that many subs since he also did a similar style report for Home Theater Magazine. Of course Keith's measurements were from 5 sets of measurements for each subwoofer on different days. Oh, and he did extensive listening tests as well! I remember him grumbling a lot and mentioning something about a perfect job for a college intern.Ilkka said:Yes, this would probably be the best option, but it's PITA to measure. I mean it takes a lot more time than a 30 second sweep. It also requires much more attention since you have to carefully listen at the subs. If I could measure only a few subs in a day...but 27 in two days?
The TEF measurements Keith Yates and I use are in fact a linear sweept that can be set to sweep either direction. For subwoofers it is usually better to start high and sweep low. We should remember there are slight differences that will impact compression somewhat so far as how much time of the sweep is spent over various frequency ranges. I would expect that the shaping of the magnitude you had to do actually balances this out closer to that of a log-sweep for the TrueRTA to read flat.Ilkka said:It was a reverse linear sweep. Log-sweep didn't work with the THD plotting, that's why linear. And of course I looped it, that's essential with TrueRTA. And actually it needs a trick or two to even get it flat.
Indeed, the data looks to largely follow the sort of performance I would expect from the various devices tested. Again, I can appreciate how much a PITA such an undertaking is. Great Job.Ilkka said:And I believe my absolute SPL was pretty accurate too. If you look at the Genelec 7073A results, it scored ~106 dB at 20 Hz with ~2 dB of compression, exactly the same what Yates and AV Talk measured.
I nominate Illka.I would think that someone would be able to write script that could look for a distortion threshold such as described by either the CEA standard or a 12dB/octave approximation of it.
Yeah, my second visit to AVS was pretty short... Now my both nicks are banned for forever. :rolleyesno:Jack Gilvey said:Looks like any reference to this work over on another forum has been nuked...typical. I nominate this forum as the place for learned discussion since the mods will actually grasp what's being discussed.
I nominate Illka.
Edit...one thread is still there...for now.
I apologise my wording. It was not needed. I have edited my post accordingly.Tom Vodhanel said:Hi Ilkka,
I think you'll find most forums frown upon "linking" to other competing forums. You may get away with it more in the states because they(forum owners) probably won't consider a finnish forum much of a "threat" to steal members/viewers. A forum in the UK could think very differently though. How long would it take for you to start another thread at AVforums with all of your data...20-30 minutes? They have already invited you to post ALL of your data on their forum(you just can't post it on a RIVAL forum and use AVForums to generate more traffic for a competing forum.....so why even introduce the idea that there are "hidden motives" at play here? There is so much drama in the subwoofer world as it is...I don't see any need to invent it when it isn't there...