Loudspeaker Compression || Understanding the Measurements Part 5

Рет қаралды 4,956

3 жыл бұрын

In this video I discuss how I test speakers for both dynamic and long-term compression to help you determine how well a speaker handles power and if it has limiting built-in.
____________________
If you look forward to more content, don't forget to subscribe and hit the notifications button so you will get updates when new videos are posted. Feel free to give me a thumbs up or thumbs down. Not everyone will like this but I hope it's useful for a lot of you.
My site:
www.erinsaudiocorner.com/
Facebook Page for updates:
/ erinsaudiocorner
If you'd like to support the channel by purchasing a neat t-shirt that would be cool. You can do so at the following link:
shop.spreadshirt.com/ErinsAud...
All of my tests are conducted using KLIPPEL Hardware and Software. KLIPPEL is an innovative leader in providing unique test equipment for electro-acoustical transducers and audio systems. Founded in 1997 by Dr. Wolfgang Klippel, the novel techniques developed for control and measurement systems of loudspeakers and other transducers are the result of over 30 years of fundamental research. This provides more accurate physical models of loudspeakers, micro-speakers and headphones valid for both small and large amplitudes. The focus is on revealing the root causes of signal distortion and defects and giving practical indications for improvements in design and manufacturing of audio products. For information on KLIPPEL products, please visit their site below:
www.klippel.de/company/about-u...

Пікірлер: 43

@vitaliistep 5 ай бұрын

The info on your channel helps many things to make sense finally. Thank you very much, Erin.

@jmporkbob 3 жыл бұрын

This was perhaps the most interesting video thus far.

@HeyYall398 9 ай бұрын

Well now, honeybunch, lemme break it down for ya real good. I been 'round this ol' Alabama countryside for a good long while, and I reckon I've seen my fair share of things. But that there video series on them speaker measurements? Whew, it's like a fresh breeze on a hot summer day! Ain't nothin' in these parts that's ever come close to it. The way y'all dove deep into them technical details, makin' it all clear as day? Pure genius, I tell ya! Y'all outdid yourselves, truly. So here's a big ol' tip of the hat to ya. Mighty fine work, folks!

@marcgras9064 3 жыл бұрын

Thank you. This is real usefull info. Please keep doing these test. This is one of the parameters you really know nothing about on comercial loudspeakers and yet it is indeed something you can hear.

@MRPC5 3 жыл бұрын

Thank you for this. It kind of hurts my soul that unboxing videos get more views...

@Space-O-2001 2 жыл бұрын

Thank you, that was an excellent series of videos that deserves a lot more views. Might be an idea to constantly keep them under the title of Beginners Guide in the description of all of your videos going forwards.

@IrenESorius 3 жыл бұрын

⭐⭐⭐⭐⭐/⭐⭐⭐⭐⭐,, everything,, again. First time Iw heard anyone speaking about this subject on YT. My neighbours also listens to my speakers at +90dB/20-20kHz,, loud and clear,, 😅. Cheers and thanks again, aaand again Professor E. 🍻🍻😎👍‍‍👍‍‍

@ErinsAudioCorner 3 жыл бұрын

I appreciate the kind words. No Professor E here. Just a guy helping spread the word of the real professors. (Toole and Olive) 👍

@IrenESorius 3 жыл бұрын

@@ErinsAudioCorner Ok Erin. The chance of people stumbling over this series of flicks your uploading are probably greater than stumbling over Toole and Olive,s work, and explained with curves and sheets like you do, most likely easier will hit home with the masses. With a bit of luck it will also grow a bit with the help of the "Monday Gang". Thanks again friend, and as told before, its a shame its so freaking expensive with freights, or I would send you my Genelec 8341 for a spin, to show how god a studiomonitor can be.

@ErinsAudioCorner 3 жыл бұрын

Well, I certainly appreciate it. And thanks again for the kind words! 🙂

@krom447 3 жыл бұрын

This video is good demonstration of differences between well designed product and Wall-Mart mass product.

@keepingupwiththejones2933 3 жыл бұрын

Compression matters. Thank you Erin

@matthewhilty4209 3 жыл бұрын

This is super interesting . It means we should be DSPing rooms at different volume levels and the DSP should be digitally compensating for gain.

@ErinsAudioCorner 3 жыл бұрын

I think it is simpler than that. Just purchase a speaker that exhibits low compression at the volume levels and listening distance you need.

@matthewhilty4209 3 жыл бұрын

@@ErinsAudioCorner True . I was mostly thinking about a studio mixing mastering environment or a professional concert setup. Yes of course in home use is simpler.

@ErinsAudioCorner 3 жыл бұрын

Gotcha. Yes, each venue would pose its own unique opportunities (or challenges). Heck, back in 2009 a car audio company made a product that had Dynamic EQ where you could build an EQ profile for two volume levels and the DAP would interpolate between those as you increased/decreased the output. Unfortunately, hardly anyone understood it or how to use it and it was never featured again.

@IrenESorius 3 жыл бұрын

@@matthewhilty4209 Hi Matthew. I'm working professional in recording studios since the mid 1980,s and the monitor levels extremely seldom exceeds 70-75dB. Sometimes when tracking (recording single overlay tracks in the control room) the levels goes up, to excite the body/pick-ups of electric guitars etc, but then everybody uses closed back headphones. Cheers!

@Kevin.L_ 3 жыл бұрын

This is the episode I was waiting for. Compression influences things negatively and often in my opinion. Not sure why you chose a powered monitor, as the amp is an unknown factor, but you made it work well enough. Thank you Erin. Any chance you plan on addressing how compression effects what we hear when listening to music? Specifically, comparing a typical 86db speaker to a similar quality higher sensitivity speaker. There is something about efficient speakers I've always been drawn to, even with their flaws. I understand that with enough power the outputs should be the same, but I "feel" that some details are more apparent with better efficiency . I've always suspected compression is behind what I'm hearing, but accept that it could be psycoacoustic. Would enjoy hearing your take on this someday. You have answered many questions for me through all of your videos. I love your scientific approach to finding and showing us what is really there. Thanks for all your efforts.

@davidatrakchi2707 3 жыл бұрын

Very professional and well made, thanks

@ErinsAudioCorner 3 жыл бұрын

Thank you!

@bobsagget823 2 жыл бұрын

agree

@ferdinandbardamu3945 3 жыл бұрын

Thank you.

@ReEvoluion 2 жыл бұрын

Thank you, the entire series was very informative. I have to say looking at the crazy compression graphs of the little monitors makes me wonder if it was caused by a limitation of the integrated amplifier more so than a problem with the drivers or enclosure. With the Klipsch I have experienced a similar phenomena with a car audio sub on cold mornings where there was a noticeable lull in output for the first several minutes as if the surround and/or spider needed to loosen up. Maybe after a tough day I am suffering from compression in the morning as well. :-P

@iliascharis5668 3 жыл бұрын

Great work, thank u. I m looking forward for your future measurements regarding dipole speakers .open baffle.......magnepans ......line arrays ....and so on .....generally measurements of those speakers always troubled me...

@ErinsAudioCorner 3 жыл бұрын

I hope to one day!

@JerryRutten 2 жыл бұрын

This is interesting stuff!! Thermal Effects * Some numbers about thermal compression… Sound Power * 1 watt sound power gives about 112 dB Sensitivity and efficiency * 102 dB @ 1W ≈ 10 % * 90 % into heating * 92 dB @ 1W ≈ 1 % * 99 % into heating * 82 dB @ 1W ≈ 1 ‰ * 99.9 % into heating Ultimate case * Just before or just over the edge of meltdown * 6 dB power compression * Re doubled * 250 ºC raised (0.4 % per ºC for copper and aluminium) * From 20 ºC to 270 ºC Impedance curve & Power compression * Zn * 8 Ω * Re * At room temperature * 5.9 Ω * At meltdown * 11.8 Ω (+ 5.9 Ω) * Zmin * At room temperature * 6.8 Ω * At meltdown * 12.7 Ω (+ 5.9 Ω) * Power compression * -5.43 dB (20 * log 6.8/12.7) * Zo (at fs) * At room temperature * 62.7 Ω * At meltdown * 68.6 Ω (+ 5.9 Ω) * Power compression * -0.78 dB (20 * log 62.7/68.6) * Numbers from data sheet of Scan-Speak 22W/8534G00 * Power compression at mid frequencies the highest * Power compression at resonance the lowest * Power compression at high end of driver low *usually the impedance at high frequencies not as high as at resonance) Electromagnetic damping * Qms = 4.14 * At room temperature * Qes = 0.43 * Qts = (4.14 x 0.43) / (4.14 + 0.43) = 0.39 * At meltdown * Qes = 2 x 0.43 * Qts = (4.14 x 2 x 0.43) / (4.14 + 2 x 0.43) = 0,71 * Thermal effects at resonance: Higher Q, so decompression! Passive crossover filters * Voice coil impedance (let’s say) doubled * Crossover frequency halved (when first order network, second order?) * Difference between drivers * Power into heating / heated mass * Time factor (reaction time for heating and cooling) * When playing music * Changes in crossover frequency * Differences in amount per driver * Differences in speed * Resulting in peeks or dips * Depending on drivers and on heating or cooling That’s why I designed and build my own audio system * Current-drive: no thermal compression * No electromagnetic damping: stable Q’s * No passive crossovers: stable crossover frequencies In the Klippel Nonlinearity Poster (I think I saw one in your background) they make a difference between distortion in sound pressure and distortion in current (current distortion). I think that’s an interesting subject!

@heathjones1980 2 жыл бұрын

Have you got any more info on your design? I'd love to know more!

@JerryRutten 2 жыл бұрын

@@heathjones1980 I can tell you a lot... The question is what do you want? To start I can tell my experience and what I did... First of all, the challenges with current-drive are: * There is almost no common knowledge around the concept. * You have to rethink all the things you already know. The things of the world of voltage-drive are literally turned upside down compared to current-drive. * There are a lot of misconceptions about current-drive. There are even people claiming it cannot work, it will explode and that kind of nonsense… Or that you will need a servo-system… People see problems with current-drive, but those problems are bigger with voltage-drive. * You cannot exchange one component at a time, you have to redesign the loudspeakers, the amplifier and the crossovers all together. I think that is the reason why some people who tried it, have bad experiences. They change the amplifier and use an existing loudspeaker with (voltage-drive) passive crossovers. And that doesn't sounds good. (But actually not as bad as I expected, you immediately hear there is potential…) I have a rather simple and minimalistic system: * Two-way sealed loudspeakers without crossovers. * A four channel current-drive amplifier. * A miniDSP DDRC-24 and a measurement microphone. What I did * I read the book of Esa Marilänen, see www.current-drive.info. Actually several times. * I designed and build a power supply with a toroidal transformer from scratch, based on information and heuristics from internet. * I bought an inrush-current limiter on internet. * I designed a current-drive amplifier, based on an existing (voltage-drive) design, where I changed the feedback and the stability network. * I designed a PCB for the amplifier, and let it produce, four times. I assembled the boards and total amplifier. * I designed and build the loudspeakers. I use 4 Ω drivers, looked for drivers with a low fs, big displacement and moderate Qms. And I looked for drivers with non-conducting voice coil formers, to eliminate also the eddy current electromagnetic damping. I didn’t find (reasonable priced) tweeters with non-conducting voice coil formers. And I looked for drivers without ferro-fluids, because with current-drive you have much lower distortion in you drivers and the noise of the ferro-fluids gets observable. I use Scan-Speak drivers. I think that Scan-Speak Discovery drivers are perfect for current-drive: big displacement, light, low distortion, sensitive, available in 4 Ω and woofers and midranges are available with non-conducting voice coil formers. * I designed the crossovers myself, based on Linkwitz-transformers, to get synchronous crossovers and to use the loudspeakers full range (from 18 Hz). * I use Dirac Live with a target room curve which I designed myself (I build a spreadsheet to adjust and to generate the target room curve). Along the way I made a mistake by using protecting capacitors for the tweeters. It sounded terrible. After three days I realised that protecting capacitors will not work with current-drive. I took them out and it sounds beautiful. And that’s an example that the usual common sense doesn’t work with current-drive. Another example is that, when you use a current-drive amplifier with nothing connected, it will clip.

@ralmslb 9 ай бұрын

I ask myself if the SPL targets should be based on the max rated by the speaker. Like if the speaker rated max spl is 95dB, that would be the purple line, than do the other 2 levels by subtracting 10dB on each. This I feel would be more fair. 102dB is a lot for a lot of speakers and I assume that would be out of spec for the many speakers.

@SDX9000 3 жыл бұрын

With these tests I feel like it would be very interesting to see your review of Neumann KH120 monitors. Those speakers have all sorts of manufacturer data provided, but still not this level of data :)

@ErinsAudioCorner 3 жыл бұрын

Maybe one day!

@NakeanWickliff 3 жыл бұрын

Thanks for doing this. You touched on it briefly but is this what you would attribute to dynamic range? It would Make the most sense. Input in = input out. If there is compression then you have less dynamic range. I’m confused why people attribute sensitivity to dynamic range. Is there any correlation? Have you tested Klipsch RP600m. Would be amazing to see klipsch RP600m with GR Research upgrade.

@ErinsAudioCorner 3 жыл бұрын

It is, at least to a degree. I think people also confuse dynamic range of a speaker with that of the system. I discussed this in this video: kzfaq.info/get/bejne/p6ile9CFzaeRZ6M.html

@ErinsAudioCorner 3 жыл бұрын

And nope, I haven’t tested the RP600. I would be interested to test one with Danny’s mod and without as he uses measurements to verify his process, though, the measurements are gated and therefore smoothed to a high degree. It would be nice to see the higher resolution results as I would expect them to show the differences even better.

@NakeanWickliff 3 жыл бұрын

@@ErinsAudioCorner Yes, his measurements are smoothed to 1/3 octave which is pretty low resolution. I was still left with a tilted treble response that I fixed with mini DSP. I think an interesting video you could do would be to explain how deviations in certain frequencies change the perception at other frequencies. The 6db fix at 1.8khz was a welcome change but the speaker still sounded small and called attention to itself. I toed them waaaay in (I'm talking like +55 degrees) and also shelved response to match harmon curve and the speakers now call far less attention to themselves and sound so much more full, due to the reduced output in the treble region.

@NakeanWickliff 3 жыл бұрын

@@ErinsAudioCorner Know of anyone in the PNW that has a Klippel machine? :)

@__-fm5qv 3 жыл бұрын

I'm always suprised that people can tolerated to listen to stuff at high output. But then again I'm used to around 75 dB at 1.5 ft/0.45m or so. I probably never stress speakers enough to notice compression as a result tbh.