2 things. Whatever you want to call it, I'm so glad to see that I'm not the only who uses this type of subtractive comparison to detect all kinds of distortion, even MP3 distortion. I've had arguments with people who can't grasp the concept, and think I'm doing something weird and pointless when I'm checking for differences between 2 amplifier channels. This can be a very powerful tool to dispel tons of audiophile myths, too! 2nd thing is that when you do full rectification on 60Hz, you get 120Hz ripple, not 60Hz.
Your joy in demonstrating makes these videos great to watch. Your ability to explain what is happening makes watching worthwhile. Thanks for making us all a little better informed. And let's all thank technology that we aren't still lifting racks of heavy iron amps every day.
Love the interaction of you two. Interesting topic. Love it! & Oh how you would have a field day in my shed. Keep it coming Dave, appreciated. Stephen Shivas, New Zealand.
I think we actually still have several of those in a rack case buried somewhere in our warehouse. I remember lugging those suckers around and cursing when they removed a caster wheel off the bottom when rolled over a door threshold or crack in uneven concrete… I remember naming one of the amp racks, “tripod” because it constantly had only three functional caster wheels due to the sheer mass of the stack of Crest 4801’s inside! Modern tech is constantly improving the chances of me being able to still walk upright in my autumn years, thankfully. Thanks for posting the wisdom and reminding me how damn old I actually am, Dave. Cheers - Eric in Sacto
I've been binging these videos. The free content alone is absolutely amazing. Thank you for doing what you do. Subscription coming soon from Down Under 👍
So it's your daughter ? It seems to me to be a beautiful and healthy relation there, it's refreshing to watch ! By the way, the video content is perfect as usual, really interesting and educating ! Thanks a lot Dave, you're the man !
You're getting very close to a phenomena that has eluded me in discussions on recording and mixing forever - the phase coefficient involved in RC filter use on similar instruments which are blended or summed, basically in any Q above 1.0 at any db/octave or decade (normally at +/-3/6/12 or 1st, 2nd or 3rd order), increasingly becoming incoherent as their signal levels become more proportionate to what an amp sees as unity. I have a theory that I've never been able to proof, nor have I seen or heard anyone else even allude to it, that goes like this: when a bass drum thumping at say 120hz is overlaid with a bass guitar thumping perfectly in time at that exact same frequency, if that bass guitar signal was recorded at -3db relative to the drum and then boosted by +3db with a 1st, 2nd or 3rd order filter having a Q of 3.0 or higher the phase to power ratios inherent in the resultant slopes on either side of the peak of 120hz will absolutely butcher any perceivable differences between the signals from both the bass drum and the bass guitar due to their moving to Q peak at different rates apart from ''peak power''. It also stands to reason that the same is true if the roles are reversed, ie bass drum is boosted to match the bass guitar, or even when either is cut given mirrored settings. This whole theory MOSTLY only works for non synthesized instruments which naturally tend to waver at pitch, as synthesized tones are more pitch stable, but from an amplifier's perspective though there is little difference - different amps will respond to such differing phase relations sometimes equally badly. Another way of looking at this is the steeper the slope of a boosted eq (whether bell or shelf) any pitch that wavers from 100hz to 120hz in regards to volume, raw power, and back (let's just say 20%+/- to or from peak Q) is likely moving through over 45 degrees of phase (ie in/out of 0 phase by 45 degrees at said Q's peak voltage). In mixes where I stuck to the rule of NO EQING WITH A Q OVER 1.0 I've seen much better power response and signal integrity on subs and mains (not so much on tweets and, really, who cares what's happening at 3-20k as long as no one starts bleeding, am I right?). It seems to me that the difference between modern mixes with several plugins where instruments are ''blended'' together by dozens of high Q filters and those done 50 years ago using a handful of outboard Pultec type filters, which were chosen specifically in increments which have the least phase incoherence, has everything to do with why even horrible recordings from mobile trucks (think The Rolling Stones, Led Zeppelin, The Who, etc) have a more pleasing low end to them than nearly anything today does - even before they were remastered and re re re re re released by Jimmy Page 5,000,000 times so he can afford to buy more teenagers' blood to infuse with his own.
Awesome video! Love watching things like this! I actually came across you from the new type of subwoofers you use in your enclosures! I compete in Spl car audio nation wide but I like the best of both worlds and love clean crisp concert loud music in my small cabin area of a suv ! Would love to get a couple of those into my demo vehicle! Think it could be the next big thing! I could see 2 in a 4th order enclosure!! Currently I’m running 4 15s in a 4th order on 15k watts ! Man it would be cool to just hang out and chat and pick your brain! Keep up these videos man! It’s awesome!
Thank you and very cool. The spl competitions are crazy. Maybe incorporating one or two those M-force motors in a car would give you an edge up. They are monsters
@@DaveRat that is exactly what I was thinking! I just don’t quite understand how they work! I am very familiar with coils and cones like we all run now but this is a whole new world! I am into really low frequencys and I play below 18htz daily on some “ rebassed music “ ! Actually have some decent guys who make clean music but unstack the bass track and replace it with 808s ! Would be cool to give you a demo! 156db at 27htz is fun! If I play my peak at 46htz witch is my cabin frequency I can almost touch a 160db but I’m on low power for the subs I am running at the moment! Anyway sorry for rambling but it’s nice to talk with someone who I consider somewhat of an audio master! I’ve been learning a lot watching your video’s!
Great stuff. The M-force has an extremely high magnetic force so it can generate more output in a smaller enclosure than a cone speaker. Like 10x the magnetic force of a professional 18". The issue is it will blow apart boxes. My email is drat@aol.com I'd be up to assist in getting an M-force into a car!
Hmmm, I will ponder that, I think we all have seen what amp clipping looks like in a scope, trying to focus on things that are outside of what is common.
Awesome video! And Sam is really picking up on info that is really not easy stuff. I hope the other amps tested will be public videos. The outro part of the videos, with the sub tones, do you have a stand alone video with those? Would be nice for soundcheck. Thanks for great videos!
We have actual dyno’s for car audio! Have a few 20k plus amps and there are a bunch of different variables to choose from! It’s fun finding out what actually puts out real power and what are just false advertising!
Great demonstration! But I believe what you are hearing is subtracted from the signal… not “added in” as you mentioned. The clean channel contains the whole signal and the clipped signal contains a lowered signal. The difference between these signals is what is missing from the clipped channel. Also IGM (current limiting) and ACL (voltage limiting) are derived using different circuits and may produce different sounding results depending how it is clipping (I or V). I also believe what you are hearing is the HF bits riding on the LF crests that are being ‘clipped’.
Hmmm, pretty sure what we are hearing is the somewhat isolated distortion component. I am driving double the voltage on one channel and monitoring that channel Midway between 2 identical loads in series, so monitoring 1/2 of double the voltage. And then the other channel is driven at the not double voltage so the two voltages that the listening speaker is attached between should be identical when the output signals are identical. Any differential between the two channel outputs is what is heard. So if one channel compresses, what is heard gets louder. If one channel clips, what is heard is clean minus not clean removing the original signal. I do a similar test comparing mic preamp distortions by running 2 mic preamps out of polarity with each other and driving one 6db harder than the other and then summing the out of polarity outputs and turning down the louder one 6db. This results in hearing just the differences. Any compression, clipping and non linearities are clearly heard. I've done this test in one or more of the digital to anaog compare vids and also some newer vids comparing x32 to m32 that are not public yet. And thank you!
@@DaveRat Hey Dave, your videos are great! I have been watching them for years. Thanks for responding. I agree. We are hearing the difference (subtraction) between the two channels. However, I think what we are hearing is missing (subtracted) from the clipped channel and not "added in" as you mentioned in the video. People might think that the amp "adds in" this difference when clipping but, in fact, it is subtracted.
Hmmm, we may be saying the same thing. By monitoring between the two outputs we are listening to the differences. Yes the one channel is clipping and losing part of it's waveform and when that happens, a bunch of harmonics are created which is what we are hearing. One way to look at it is that those harmonics are not in the clean signal and they are on the distorted signal so we are subtracting the clean and hearing what is added. Another way to look at is that all signals have all frequencies except all frequencies except what we hear is cancelling out. The clipping is just some of those frequencies not cancelling out. And the monitoring of the two outputs we the extra frequencies in the clean waveform that are not being cancelled. But that angle has a bit of the 'plannets revolve around the Earth' complexity to it and while things can be calculated that way, it makes things more complex.
@@DaveRat Dave, thanks for keeping this dialog going. This is great stuff. Yes, I believe we are saying the same thing. Yes, clipping adds distortion. But it is transient. Don’t forget about the amps limiter. Most of what you are hearing could be the difference from the amps protection limiter engaging. As the clipped channels limiter engages further, you hear more difference. I believe this (difference from limiter) is most of what you are hearing. And this portion (difference from limiter) is subtracted from the clean channel. i.e. What you are hearing could be a little distortion mixed with what is removed by the limiter.
What you say is true and when I release the video of a the powersoft amp driven hard and we hear that difference, you will hear the impacts of the limiter from an amp that actually has a limiter. The crest amp does not have a limiter so the odds that we are hearing the impact of the limiter are quite low and highly improbable.
Poor Crest its only a baby compared to the others. One of my first power amps was a Crest 601 that is before I moved onto Crown and then and especially QSC. In comparison the LabG and Powersoft x4 are absolute beasts there monsters and come with a price tag to suit but you want an amp to rock the house wow. Still use my QSC amps to this day to drive Marshall cabs/PA rigs, in racks on wheels thankfully, they weigh an absolute ton but only once has one failed on me, look after your amps they look after you. Interesting comment below about valve amps and their beautiful warm harmonic distortion, but reliability? questionable. This is more a question about using a valve preamp but similar harmonic tone responses can now be achieved with modellers and plugins (amp emulation saturation and sims). Yes my Marshall amps are loud but they do not compare and compete with QSC especially when driving multiple cabinets chained together. how close do modellers/power amps get to the real deal, so close you could not tell them apart, nostalgia for sure but sadly that is the future for valve amps. Yes I still sometimes put "Marshalls amps" on top the stacks but the cabs are actually powered by QSC, poor cabs look lost lonely and missing something without the amps there but in truth, I now use empty amp cases, they just look right but lack real power when you need it else you need 6 Marshalls amps for each QSC. More recently we have significantly reduced the on stage volume and rely more on the venue/show PA, gone are the days needing a row of cabs on stage, just enough to hear what your playing is more than enough !! We even went through a period using empty cabs. but that was just a phase, now we have one stack each just for monitoring purposes but I know many musicians who now only use in ears so there is almost no need for any on stage monitoring, how times have changed over the decades.... gawd I must be gettin old. Meant to add, Crest amps have built up quite a good rep as home Hi-Fi system power amps, for a good reason, there really good amps !! oh yeah forgot, there now called/badged Peavey oh well...
I feel like you know this, but maybe didn’t articulate it clearly in the video. You keep saying that we’re hearing what the clipping is “adding”, but actually we’re hearing what it’s taking away. It’s the peaks which wanted to go past the rails but couldn’t. We do hear some distortion and “added” harmonics, but that is the crossover distortion caused by ripping out the middle of the waves. It’s very much like a gate with too fast time constants, and in fact sounds a lot like a gate especially in the impulse and music tests. This may seem pedantic, but I think if you actually found a way to cancel out the fundamental in the clipped signal so that you only heard the harmonics generated, it would sound significantly different, and if you did that same procedure with this kind of crossover distortion, it would be different again still. IDK how to accomplish that without some FFT voodoo, but I’d be interested to see it.
So when clipping flattens the top of a waveform and in order to reproduce that newly flattened waveform a bunch of extra high frequencies that were not present in the original signal are now generated, that does not equate to adding??
Honestly, I argue not. You can look at the scope and see that nothing is being added. It's only being taken away. What we are hearing in your test is exactly that difference of what is being taken away. If you measure the integrated voltage or power, you'll find that the output wave is in fact less than the original. The output wave does look (and sound) a lot like what would happen if we added a bunch of harmonics of the original signal, but that's almost just because that's what that new waveform sounds like, you can only ever get an approximation by means of addition. Nothing is actually being added, and the perception that harmonics are being added or generated is mostly an artifact of the analysis (fft or earbrain). But I haven't convinced anybody with that argument yet. :) But honestly, that's beside my original point. Whether harmonics are added or not, there is a change in the perceived harmonic series, but that's not really what we're hearing in your video. We're basically just hearing the part of the quieter signal which not being cancelled by the clipped signal. That's perfectly valid and instructive in it's own right, and I really didn't mean to say it wasn't. It just feels to me like you're telling us it's something slightly different which would also be quite interesting if we could figure out how to really do it.
Nothing visually or electrically is being added but in order to acoustically re create the signal we hear additional frequencies the did not exist in the original. If the signal was just lower, we could slightly adjust the level and cancel the lower level. Which is what I am doing, what we hear is what did not exist in the original. A 100 Hz pure tone sends nothing to the tweeter crossed over as 1000 Hz But a clipped 100hz signal does. So from the perspective of a listener, speaker, amplifier equalivlzers and so on, there are additional frequencies added. those seem to be the perspectives that apply to the humans.
Hmmm, it remains a matter of perspective. I understand the viewpoint that chopping the tops leaves less, so subtractions not adding. But measuring the signal with an analyzer will show added HF, so there is adding from that perspective, the perspective of what we hear, what a speaker sees and what all the electronics see as the signal. It's like a building from 5he perspective of "how many square feet on the ground does it occupy?" Before it's built, the materials, if you add up the sq feet occupied on the ground, is much larger than the finished structure. If you remove some supports and the building falls over, you have removed material yet the square footage occupied has increased. This is not unlike a signal, removing part of the signal can increase the bandwidth. Or looked at another way, if the only thing that occured is the removal of part of the signal, then if you put it back, the signal will return to its original state. Do you believe that if the removed part of the signal was added back in to the chopped signal, it would return to being the original signal? If A minus B = C then C plus B = A Right? Unless something else was added or created, right?
