Frequency response

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
P

Pygmy

Member
Joined 2016
As I'm fairly new to diy speaker building I'm not very good at reading the graphs yet.

I'm looking at the following graph, and to me this looks pretty good - fairly flat frequency response on and off axis is what I make out of it.

Is that correct?
What properties can you experienced people read from this graph? I'm eager to learn from you.

lS3MRn7.jpg
 
P

Pygmy

Member
Joined 2016
And what I'm wondering is - do these graphs show the "quality" of the speaker?
Can an educated diy'er tell from graphs like these whether it's a good (expensive? Non-diy?) speaker or a cheap one?

I've seen expensive speakers with graphs not very flat, and cheap(ish) speakers with fairly flat response.

What factors determine the real life sound quality in the end?
 
Last edited:
P

Pygmy

Member
Joined 2016
bwaslo said:
Sounds like you're reading it correctly -- provided you also know what angles the off-axis curves are measured at! (if those are 10degree steps, pretty good, but if they're 1degree.....)
Click to expand...
These were measured at 0 / 15 / 30..
Therefore this graph seemed pretty good to me - I'd like to be able to diy something that could approach something like this I think?
 
Last edited:
Ron E

Ron E

Member
Joined 2002
There is nothing immediately wrong with either graph; when comparing the impedance with the frequency response, they don't seem to match up well. Looks like the box is tuned to ~35Hz and the bass starts rolling off at 100? Probably a bad splice and/or scaling job on the LF measurements.

Read about FLoyd Toole's research if you want to know what listeners prefer. Flat frequency response and smoothly narrowing directivity. IT also must be linear enough that it doesn't fall apart at high volumes.

I have some small Sony's (with just a cap on the tweeter) that measure fairly flat on-axis and sound good at low volumes, but fall apart above 85-90dB... I sometimes use them as demos to show what can be done with a $40/pr speaker. I replaced the caps with mylar, and added some foam damping material to the walls... No chance I could make a better speaker for that money.
 
Last edited:
Y

YSDR

Member
Joined 2013
This is the measurements of the Omnes Audio Exclusive 3/8, if i'm not wrong. The wrinkles around 6-7kHz maybe add a little metallic taste to the sound but i think it's a reasonable design overall. Distortions are below 1% from 150Hz and up, even at 95dB, the waterfall graph is relative clean too according to the published measurements. The phase angle of the impedance curve is more severe at some point but no problem for a good amp i think.
 
Y

YSDR

Member
Joined 2013
Pygmy said:
Spot on, it's the Omnes Audio Exclusive 3/8!
It bugs me that I can't read the graphs very well yet - I have no idea what the impedance graph means and how it relates to the frequency response graph.
Click to expand...

If you use solid state amp (low output impedance=high damping factor), you don't have to worry about the impedance response too much unless it dips too low, but if you use tube amp to power the speakers the impedance curve need to be fairly flat because of higher output impedance of the amp. With tube amps, impedance correction is often necessary.
In this case the impedance response is relative flat and not dips too low. I think it works well with SS amp and maybe with tube amp too with minor deviations to the frequency response compared to SS.
What type of amp you plan to use?
 
Last edited:
P

Pygmy

Member
Joined 2016
denibeni said:
If you use solid state amp (low output impedance=high damping factor), you don't have to worry about the impedance response too much unless it dips too low, but if you use tube amp to power the speakers the impedance curve need to be fairly flat because of higher output impedance of the amp. With tube amps, impedance correction is often necessary.
In this case the impedance response is relative flat and not dips too low. I think it works well with SS amp and maybe with tube amp too with minor deviations to the frequency response compared to SS.
What type of amp you plan to use?
Click to expand...

Probably either a diy TPA3255 based amp or a spare Onkyo TX NR636.
 
P

Pygmy

Member
Joined 2016
So say I could build this for about 300 euros per speaker using crossover components comparable to the high quality version sold here:

Strassacker, Komponenten: Lautsprecher, Frequenzweichen, Bauelemente

(I can get some of the parts quite a bit cheaper through a friend)

how would they fair against "commercial" speakers?
Aside from the fact I enjoy building stuff, would it make sense sound quality wise to build these over buying f.e. a set of Kef floorstanders?
 
Last edited:
Y

YSDR

Member
Joined 2013
Usually a successful DIY speaker costs fraction of a comparable commercial stuff sound quality wise. If you know what you are doing then you have a great chance to never going back to commercial.

It's hard to say how to compare this speaker to a KEF floorstander, since I have not heard any of them but i think the rule of thumb applies here too.
 
Juhazi

Juhazi

Member
Joined 2007
Ron E said:
There is nothing immediately wrong with either graph; when comparing the impedance with the frequency response, they don't seem to match up well. Looks like the box is tuned to ~35Hz and the bass starts rolling off at 100? Probably a bad splice and/or scaling job on the LF measurements.
---
Click to expand...

The nearfield response of the woofer goes usually up to 3-400Hz, in a presentation which is a combination like we most often see (Stereophile etc.). The box has a slot vent and it's nearfield measurement is missing here! The slight "boost around 100Hz is because there is some baffle step compensation used in the xo.

It is also possible that the measurement was made in a a large anechoic chamber, and it might be farfield from the lowest frequencies up. In a normal listening room low bass will get help from the floor and front wall. This kind of measurements are used in SoundStage!, measurements done in NRC, Canada. Explanation here and they look like this (a bass reflex speaker, at 0-15 degrees horizontally)
fr_on1530.gif
 
Last edited:
bwaslo

bwaslo

Member
Joined 2006
libinst.com
Pygmy said:
I've seen expensive speakers with graphs not very flat, and cheap(ish) speakers with fairly flat response.

What factors determine the real life sound quality in the end?
Click to expand...

Well, in the hifi world there's not necessarily a strong correlation between "expensive" and "quality"! It's a market of perception and things that are expensive are assumed by the target market to be of higher quality -- at too low a price they might not sell as well. Weird, but true. But a high quality speaker will typically have higher costs to manufacture (and total lowest possible cost speakers wil pretty much be low quality, but those don't really get considered here often).
 
P

Pygmy

Member
Joined 2016
bwaslo said:
Well, in the hifi world there's not necessarily a strong correlation between "expensive" and "quality"! It's a market of perception and things that are expensive are assumed by the target market to be of higher quality -- at too low a price they might not sell as well. Weird, but true. But a high quality speaker will typically have higher costs to manufacture (and total lowest possible cost speakers wil pretty much be low quality, but those don't really get considered here often).
Click to expand...

That's what I understood as well, the thing is that just when I think Im reading the graphs right I find measurements of diy speakers costing ≥1000 euros that in my opinion read worse or just as good as some diy speaker kits of say 300 - 500 euros, with the expensive ones getting high praise in reviews.

So that's when I get the feeling I'm either not reading the graphs right, or Im missing some crucial info explaining why the expensive ones are somehow better even if they're less linear or have worse off-axis response.

Edit:
I've paid 17 euros to get access to the archive of German diy speaker magazine Klang&Ton to get my hands on measurements of all kinds of speaker kits, as well as speaker drivers. I've considered all kinds of speakers from fast to open baffle to mtm to transmission line to tqwt and everything in between :)
It's just hard to find out what's the best option given a certain budget and room size / speaker positioning :)
 
Last edited:
P

peterbrorsson

Member
Joined 2006
Pygmy said:
Edit:
I've paid 17 euros to get access to the archive of German diy speaker magazine Klang&Ton to get my hands on measurements of all kinds of speaker kits, as well as speaker drivers. I've considered all kinds of speakers from fast to open baffle to mtm to transmission line to tqwt and everything in between :)
It's just hard to find out what's the best option given a certain budget and room size / speaker positioning :)
Click to expand...

Hi Pygmy,
I prefer HobbyHifi! The process of coming to final result is much more detailed in HobbyHifi than Klang+Ton.
I.e. one learn more in HobbyHifi...

Peter
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom