YouTube Loudspeaker Demo Videos - Yes No Maybe?

[DanielT]

Judge this for yourself, what do you think?

"Sound reinforcement worker
4 months ago
You certainly can not hear the sound of the speakers in the video.
It is the original sound without any room influence.
I have made a video with the track over the speakers"

Edit:
So therein lies another problem with speaker videos, fake or not?
Rather odynamic video, pictorially. Only a still image. It was perhaps an extreme example, but you get the point. The problem with speaker videos on Youtube.

It should be added that we do not know the purpose of the video. Is it to show the sound in some speakers or is it just a picture of some speakers and a song being played.

 

[antcollinet]

Here's one of the videos I watched. They generally recorded from seats right in front of each pair of speakers:

Well that was interesting (at least the first 10 minutes I listened to - I don't have time - or motivation - for the full hour ).

My opening impression was that while I could hear (I think - but based on different music) sonic differences - all were lacking bass, sounding thin and reedy on my system. I assumed the microphone/recording system/youtube was unable to deliver the bass. Then at 8 minutes (the big black speakers) - massive, boomy (and to my ears on my system unpleasant/uncontrolled) bass.

So who knows. What I would say is if I believed that was how the speakers really sounded, I wouldn't buy any of them at any price. And while it may be possible to detect sonic differences, those differences are overwhelmed by the "coloring" (for want of a better word) added by the sound delivery channel into my room. I am absolutely convinced I am not hearing those speakers as they sounded to the person doing the recording - and that alone makes the process useless for buying decisions.

 

[Rednaxela]

One might as well take a set of anechoic data and use it to generate a convolution for the desired speaker. That would at least eliminate room modes and reflections from the demo.

Its usefulness would still be rather limited (particularly in the case of non-uniform directivity) but the listener would no longer be hearing two rooms simultaneously.

I like this idea!

Struggling to fully understand it though.

Assuming you'd take out one room by using a good pair of headphones or IEM's, how would this approach then work exactly? Genuine question, just trying to wrap my head around this.

 

[antcollinet]

I like this idea!

Struggling to fully understand it though.

Assuming you'd take out one room by using a good pair of headphones or IEM's, how would this approach then work exactly? Genuine question, just trying to wrap my head around this.

The theory is if you record an impulse response from a speaker (or generate one from a measurement sweep) in an anechoic chamber, you can generate a convolution filter from that that when applied to any music recording should sound as it would when played through that speaker in an anechoic chamber. Then when it emerges from your speaker only your own room acoustics (and your own speakers characteristics) are applied.

Another application - REW can generate an impulse response from a measurement sweep of your room. Then when used in a convolution filter, it acts as room equalisation. How I was applying REW EQ until my Flex arrived recently. Theoretically you can use an impulse response from any room, and apply to to music to make it sound as if you are listening in that room.

The Royal Albert Hall in your living room, anyone?

 

[fineMen]

For me, that would be too negative a world view.
I believe that dishonesty and not meeting a set of quality standards are two different things.

You are defending the naive? I didn't accuse anybody to deliberately fool others. Only half, one could say. Even a youtuber should be willing to scrutinise the own competence in such a tech topic. But to not do so has become pardonable. You know "unboxing" videos? I see those "reviews" on this level. No bad feelings.

Otherwise, a forum participant already posted I think four dummy head recordings of speakers in-room. Made by really qualified personnel. That's an other dimension. It was fun to discriminate the models--listening over real (expensive and still good) headphones of course. I'm afraid youtube doesn't deliver.

 

[audiomaestro]

I found a Tannoy vs Genelec vs Rogers LS3/5a comparison done in separate videos. The uploader appears to have done a reasonably good job with mic position.

Tannoy

Genelec

Rogers

Original source track for one of the songs

The differences are night and day. When compared to the source track using my HD650 headphone, I found the Rodgers to be easy on the ears yet dull, the Genelec neutral and clear, and the Tannoy suffering from muddy nasal coloration.

 

[fineMen]

I found a Tannoy vs Genelec vs Rogers LS3/5a comparison done in separate videos. The uploader appears to have done a reasonably good job with mic position.

Nope, sorry

A microphone does not hear in any way like a human being. That is why we have dummy heads.

Why are dummy heads then not used for every recording?

Because a recording for replay over speakers needs tons of post-processing anyway, and the characteristics of the dummy head would make the task of p/p even harder.

You must not think that a microphone detects "the real thing". You must not think that speakers replay "the real thing". This myth shall be debunked. It was the industry to spread the word of "high fidelity" while, gracefully, making You forget the truth of necessary manipulations to the "signal" so that You can enjoy the "information" by understanding the meaning ...

It it so harsh to ask an audio enthusiast to buy, steal or loan a set of microphones to understand the tech of the valuable stereo a tiny bit?

 

[audiomaestro]

Nope, sorry

A microphone does not hear in any way like a human being. That is why we have dummy heads.

Why are dummy heads then not used for every recording?

Because a recording for replay over speakers needs tons of post-processing anyway, and the characteristics of the dummy head would make the task of p/p even harder.

You must not think that a microphone detects "the real thing". You must not think that speakers replay "the real thing". This myth shall be debunked. It was the industry to spread the word of "high fidelity" while, gracefully, making You forget the truth of necessary manipulations to the "signal" so that You can enjoy the "information" by understanding the meaning ...

It it so harsh to ask an audio enthusiast to buy, steal or loan a set of microphones to understand the tech of the valuable stereo a tiny bit?

That's rather amazing then, that despite all the variables mentioned it was still not enough of a hindrance to mask the tonality of each speaker. If you read around, you'll find the objective data and many subjective opinions for each of those speakers is coinciding with exactly how their tonality sounded in the videos. (when compared against the source track)

 

[fineMen]

... not enough of a hindrance to mask the tonality of each speaker ...

Again, that is a misconception. The sonic impression for a human can be very different. The method is invalid. As many many said before.

I actually don't get why so many people try so hard to circumvent doing measurements on their own?! As if all audio knowledge comes from theorising in (cost-free) audio forums, propelled by (cost-free) online "reviews". It would be way more effective, time-wise, to just learn by--doing. Cost again? My phalanx of measurement equipment was significantly below 70$, about 5 of Linkwitz type Panasonics plus USB interface with phantom power and digitising. The 20m cables were the most considerable factor.

What is it about to expect insight from youtube videos? Any explanation? If there is something, please tell what it is and how come. Thank You, I'm lost!

 

[Jim Shaw]

Come on @Jim Shaw, you can’t call others argumentative and then bow out with an argumentative remark. That’s unfair.

Feel free to have the last word, but at the risk of coming across as argumentative myself, I want to have said that I only see a constructive and civil conversation here. What I see is people doing their best to express themselves with care, stay on topic, and back up their views with helpful examples. Going through today's What's New list, I think there are quite a number of less refreshing and enjoyable threads to take part in. This is thanks to everybody’s contribution here, including yours.

So thank you for having taken the time to share your views, and should you be stepping out because of something I wrote - my sincere apologies.

Apology accepted. [Full stop]

 

[My adventures in stereo]

I started recording samples , to share with my friends
Youtube presented an easy way to do so
I am using all copyrighted material, so I dont make a dime from this
Try to pick good quality source materiel from my personal collection

Started with doing mobile phone recordings, then switched to USB mikes
Found them lacking, so currently using a dual mic set up with a Moto interface & audacity
Have found incremental , improvement in sound quality

A complete novice to recording, hence enjoying the learning process
Would love feedback, thanks in advance
My adventures in stereo youtube channel

 

[HarmonicTHD]

I started recording samples , to share with my friends
Youtube presented an easy way to do so
I am using all copyrighted material, so I dont make a dime from this
Try to pick good quality source materiel from my personal collection

Started with doing mobile phone recordings, then switched to USB mikes
Found them lacking, so currently using a dual mic set up with a Moto interface & audacity
Have found incremental , improvement in sound quality

A complete novice to recording, hence enjoying the learning process
Would love feedback, thanks in advance
My adventures in stereo youtube channel

Aren’t you just a tad bit off topic. The thread was if one can judge the speaker sound quality by watching YouTube reviews. (Which I personally think is bogus for the many reasons already mentioned by others).

Maybe start a separate thread so we can discuss recording quality in home studios. So it gets the attention it deserves. Just proposing. Or am I not seeing something here?

 

[DanielT]

This is not so strange. Recording signal into speakers, intercepting sound through speakers (inverted, ie microphone) into new speakers, or headphones.Then listening. It will be what it will be, as so many others have pointed out depending on .... the microphone, the acoustics in the listening/recording room, on ..and so on ..

Have you thought about this? These videos on speakers, has EQ been involved? Absolutely nothing wrong with that IF it is reported that this is the case. Can you read about it in these videos? Is it reported? Nop.So what are you listening to?
(it may have already been pointed out in the thread, then I missed it)

 

[Rednaxela]

The theory is if you record an impulse response from a speaker (or generate one from a measurement sweep) in an anechoic chamber, you can generate a convolution filter from that that when applied to any music recording should sound as it would when played through that speaker in an anechoic chamber. Then when it emerges from your speaker only your own room acoustics (and your own speakers characteristics) are applied.

Another application - REW can generate an impulse response from a measurement sweep of your room. Then when used in a convolution filter, it acts as room equalisation. How I was applying REW EQ until my Flex arrived recently. Theoretically you can use an impulse response from any room, and apply to to music to make it sound as if you are listening in that room.

The Royal Albert Hall in your living room, anyone?

I see. Thank you for clarifying!

So would that also mean one could play some program material through a speaker, record that anechoically, and then apply ones own room’s impulse response to it, play it back over headphones, and this way have a virtual home demo?

Just thinking out loud.



Edit: fixed a little typo.

 

[Rednaxela]

You are defending the naive? I didn't accuse anybody to deliberately fool others. Only half, one could say. Even a youtuber should be willing to scrutinise the own competence in such a tech topic. But to not do so has become pardonable.

I guess I commented on something you did not write indeed. Sorry for that. As to defending the naive - not sure, but will have to think about it some more.

You know "unboxing" videos? I see those "reviews" on this level. No bad feelings.

Understood.

Otherwise, a forum participant already posted I think four dummy head recordings of speakers in-room. Made by really qualified personnel. That's an other dimension. It was fun to discriminate the models--listening over real (expensive and still good) headphones of course. I'm afraid youtube doesn't deliver.

Would you happen to have a link? Would love to give these a listen.

 

[Rednaxela]

So therein lies another problem with speaker videos, fake or not?
Rather odynamic video, pictorially. Only a still image. It was perhaps an extreme example, but you get the point. The problem with speaker videos on Youtube.

A very good point indeed, thank you.

Even if you hear some room acoustics, that can be fake too.

 

[phoenixdogfan]

Oh hell no. The audio analog of looking at a cellphone pic of Xerox of a mineographed (for those of us old enough to remember) copy of the Mona Lisa.

 

[Rednaxela]

Oh hell no. The audio analog of looking at a cellphone pic of Xerox of a mineographed (for those of us old enough to remember) copy of the Mona Lisa.

I love it.

If you had made one of Marilyn Monroe and printed it, you’d be a millionaire now.

 

[antcollinet]

I see. Thank you for clarifying!

So would that also mean one could play some program material through a speaker, record that anechoically, and then apply ones own room’s impulse response to it, play it back over headphones, and this way have a virtual home demo?

Just thinking out loud.



Edit: fixed a little typo.

I guess (I'm a long way from an expert here, just done enough background reading to eq my own system with REW) in theory - yes. But then only with that one recorded song.

Whereas if you created the convolution for the speaker, and followed it up for the one for your own room, you could demo any music you like.


In any case it will never be a perfect simulation - for one thing it will only apply to the exact microphone position for each measurement.

 

[Kvalsvoll]

You're not going to be able to compare many of those things from a YouTube video comparison, but you will absolutely be able to discern whether the tonality is on point or not. There appears be confirmation bias in your perception, because I think most people will admit the general tonality is very well matched.

Quality of the youtube stream is not an issue here, for 2 reasons.

First, the recording is compared to the original streaming through the same service with same lossy compression. If the original from yt sounds better, or at least different, it does not matter if the quality of both is limited by lossy compression.

Second, you are in for a surprise if you compare the original on yt with a lossless version of the exact same piece of music. Do a ABX in foobar, see if you can easily hear the difference between lossless and the AAC 128K that yt uses.

The recording sounds different. In several ways. But that is expected, and really does not tell much about the speakers in use. But it is fun and entertaining.

Is it possible to make such a recording, so that it actually sounds like the original? Yes, and it is not difficult to define the requirements for such a set-up. But will this be a system that is representative for how music sounds through a good sound system, a sound that is enjoyable to listen to? I think not.

So the recording of such a scenario with music played back through speakers in a room will always be different, and those differences are attributed not only to the speakers in use, but also room acoustics, placement and type of mics.

If anything, it will tell something about the system used for monitoring - the system used to listen to the samples. It will be easier to spot the differences in a system that is more revealing and does not compromize sound quality in itself.