I am now DSP-tuning a full active 5 way system. And now it sounds top notch. Balanced. Sweet. Powerful, hard AND juicy, controlled and musical! All Front stage.
Well. Do you like flat-flat?
No one does? So why is flat still an issue and just not a very bad old joke, like it IRL is? And this is equally true in all home High End systems?
Pictures 1: 3 way front stage, closed boxes midbass. Not in picture: Twin 6,5 inch subwoofers front, closed box. One 8 inch subwoofer reae, closed box. Picture 2: NOT FLAT =) REW RTA measurement.
Well. Do you like flat-flat?
No one does? So why is flat still an issue and just not a very bad old joke, like it IRL is? And this is equally true in all home High End systems?
Pictures 1: 3 way front stage, closed boxes midbass. Not in picture: Twin 6,5 inch subwoofers front, closed box. One 8 inch subwoofer reae, closed box. Picture 2: NOT FLAT =) REW RTA measurement.
Attachments
It's likely a matter of taste/purpose. I prefer a boost on the low end (100Hz and below) and a relatively flat response through the rest of the spectrum.
Having a system that can produce a flat response to the highest required volume level, is a good thing.
Some may listen to recordings that are thought to be very near prefect. For those recordings, a perfectly flat system may be preferred.
Having a system that can produce a flat response to the highest required volume level, is a good thing.
Some may listen to recordings that are thought to be very near prefect. For those recordings, a perfectly flat system may be preferred.
I suppose that if a song was mixed by the producer using a set of monitors with flat response, and you listened to it using a pair of speakers with flat response, hearing that music as the producer intended you to hear it would not be such a bad thing.
I have "tuned" speaker systems for everyone from Skrillex to Neil Young, Capitol Studios to Tommy Lee, scoring engineers and movie composers etc. Virtually no one is creating, mixing or mastering on speaker/room combinations that are "flat".
A rising response in the low end is typical. Usually starting lower than the OP's but typically with a 4 to 6dB difference between say 60 Hz and 10 kHz. What is rarely discussed is how much the time domain response of the room affects our perception of the frequency response, so much so that in our studio design efforts we prioritize that aspect over frequency response.
You're also in a car which changes things a little as not only are the speakers near boundaries so are you. You really want to get that tweeter closer to the midrange though.
It's fairly well established that we like a ~5dB downwards slope for averaged room SPL across the audio spectrum.
You're almost 15dB which could be exaggerated by the boundary conditions of measurements and positions of speakers but certainly wouldn't be anything out of the ordinary for a car. Usually you need the bass turned up quite a bit for it to have a similar level of impact with engine/road noise vs when the car is stationary and turned off. Not only that most car users tend to prefer having the bass exaggerated beyond that anyway.
I did a car system for someone once and finely balanced the bass/mid/treble using the cars DSP and it sounded great when stationary. But then we went onto the road and the bass vanished. Clicked the bass boost button and everything was right again. As soon as we stopped, and turned off the engine, the bass was overblown. The owner of said car loved it with the bass cranked up all the time. Go figure.
It's fairly well established that we like a ~5dB downwards slope for averaged room SPL across the audio spectrum.
You're almost 15dB which could be exaggerated by the boundary conditions of measurements and positions of speakers but certainly wouldn't be anything out of the ordinary for a car. Usually you need the bass turned up quite a bit for it to have a similar level of impact with engine/road noise vs when the car is stationary and turned off. Not only that most car users tend to prefer having the bass exaggerated beyond that anyway.
I did a car system for someone once and finely balanced the bass/mid/treble using the cars DSP and it sounded great when stationary. But then we went onto the road and the bass vanished. Clicked the bass boost button and everything was right again. As soon as we stopped, and turned off the engine, the bass was overblown. The owner of said car loved it with the bass cranked up all the time. Go figure.
My experience was that sitting with the engine off versus driving was a totally different set of requirements for playback quality and equalization. The tires, engine, and wind tended to make it more difficult to hear correctly without tilting the response in favor of bass and low midrange.
The mostly flat response works well at home, but its completely up to the listener what curve suits their taste and needs. I do like a bit more bass and less treble sometimes, especially with newer albums that I find can be very aggressive. There are also a few, rare albums were I want a little less bass.
If I'm not reading that graph wrong, I am a little surprised that your bass appears ~30dB higher in amplitude than your treble. Was the mic at ear height and position?
Cheers.
The mostly flat response works well at home, but its completely up to the listener what curve suits their taste and needs. I do like a bit more bass and less treble sometimes, especially with newer albums that I find can be very aggressive. There are also a few, rare albums were I want a little less bass.
If I'm not reading that graph wrong, I am a little surprised that your bass appears ~30dB higher in amplitude than your treble. Was the mic at ear height and position?
Cheers.
Yes, i was a little surprised myself where it all ended up, a very nice tilt on the graph🙂👍
And yes i am using a calibrated mic and it is placed where the drivers head are expected to be.
So i am sligtly surprised myself. But it sounds really musical, full bodied, crisp, clear and strong. And not bloated or fatiguing.
Just like real musical instruments and voices sounds in real life. Lush, full bodied, present and never electrically harch and thin.
Well, im just thinking out loud at this point. And flat-flat sounds paper thin an un-musical, and un-natural. In my opinion🎺🤗🎸
And yes as many of you have written, its ultimately a question if taste 🙂
My sound ideal is natural sounding, high resolution musicality with some extra sugar and a little, little extra power and pepp in the step👍
Do you mean flat as in a flat measured sound graph, or flat as in no added EQ, bass or treble?
I think of "flat" as a starting point. Use that as a way to make sure there aren't any significant holes in the system. Then adjust to taste.
Yea okey🙂
No bass&treble EQ can, and often do sound top notch awesome, if the whole system is matched and level tuned right.
I am talking about the thin sound of a flat measured sound graph.
In this car and measurement i have not yet added or extracted anyting with the EQ.
No bass&treble EQ can, and often do sound top notch awesome, if the whole system is matched and level tuned right.
I am talking about the thin sound of a flat measured sound graph.
In this car and measurement i have not yet added or extracted anyting with the EQ.
If all your calibration is correct, I would not trust a measurement with a level that low (30 dB SPL at higher frequencies). If your desired signal is not significantly above background noise, an accurate reading is unlikely. Something in the 80-90 dB SPL range is more typical for tuning a car. Wear hearing protection if you need to be in the car to do this. Without protection it gets annoying fast and will make it hard to judge what you hear after.
If you are using a correlated pink noise signal to measure both channels simultaneously with RTA or are using an impulse-based measurement, it's very easy to introduce errors when measuring multiple speakers simultaneously (e.g., left and right channels at the same time). Very small differences in the distance from the microphone to drivers will create cancellation. This affects high frequencies more, since the wavelengths are smaller. I can't tell from your description if this applies. If it does apply, switch to uncorrelated/random/unsynchronized pink noise. Sorry for the multiple descriptors, but people call it different things.
It's also pretty common to tune the left and right sides or individual amp channels in isolation to get things flat, then start combining things to apply crossovers and global EQ.
Steady-state measurement (RTA/pink noise) vs. direct sound (impulse-based measurement) affects this greatly. The two methods often produce different results, with steady-state showing in-room downward-sloping response with increasing frequency while a direct sound/impulse measurement shows flatter response. The following full paper is available free and deals with this topic. It also includes some data on typical car curves (section 4.2 and Figure 15).
The Measurement and Calibration of Sound Reproducing Systems, by Toole
https://www.aes.org/e-lib/browse.cfm?elib=17839
Last edited:
IMO, flat (measured, not just dial markings) is a necessary target for designers to start with. We'd be lost without it. OTOH, it doesn't usually sound the best in real world listening situations. There shouldn't be an stigma to using tone controls or full equalization, though I'd add one caveat. Traditional equalization can't fix anything other than response, and many problems have to do with reflections and room treatment. If you need to fix the room, you need to fix the room!
Simply top notch advise. I always use hearing protection when uncorrolated pink noice is playing.
Save your ears for: MUSIC.
STFU, and just do.
🙂
Well, besides all above. Still: You will need a landslide from the bass to the treble, to make things real, natural and strong.
But i think you are thinking right: Noice versus real signal. Could be fooling us. So. I will play the pink noice louder. Uncorrolated. And post the measured curve.
🙂🤚
I was always under the impression that flat reproduction was a goal to avoid coloring the recording. Ie. leaving how bassy a drum sounds or how sharply a high hat peaks up to the guy doing the mixdown.
That said, the brick wall limiting loudness wars have pretty much killed dynamic range IMHO. So I say tune it how ya like it.
That said, the brick wall limiting loudness wars have pretty much killed dynamic range IMHO. So I say tune it how ya like it.
And just to be more than clear clear:
The initial posted REW uncorrolated pink noice curve is done AFTER i have run the Helix automated time/distance based adjustment between all single speaker elements, to avoid cancellations.
🙂🤚
The initial posted REW uncorrolated pink noice curve is done AFTER i have run the Helix automated time/distance based adjustment between all single speaker elements, to avoid cancellations.
🙂🤚