Dear owners of Koss Porta Pros/KSC75s: Please test this!
Oct 5, 2007 at 1:09 AM Post #31 of 50
Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
I invested a few hours with the CurveEQ and could significantly improve the sound. In head-to-head comparison to how it's now, the regular Koss seems to sound more often than not like some kind of toilet bowl with built-in speakers. :)


Could you post the new filter separately or in the filters archive (as opposed to the big archive containing the executables)?
 
Oct 5, 2007 at 1:28 AM Post #32 of 50
Quote:

Originally Posted by pfloding /img/forum/go_quote.gif
Keep in mind that the ears slowly adjust to the sound you are hearing, and any sudden switch may cause the new sound signature to sound alien.


Yes, I am aware of that. But with this specific song, I don't really think it matters if you are adjusted to the new signature or not. It really stands out... it actually sounds holophonic.
 
Oct 5, 2007 at 2:01 AM Post #33 of 50
There was one last thing I wanted to address about post #15 (I'm not going to address everything), regarding DH's limitations:
Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
What we cannot get rid of are the subtle differences between the ear types, so it will always sound bad for some people.


You brought it up several times. Yeah, this might be part of the reason why I think DH mangles a lot of recordings but it seems to me that this is also a problem for what you propose to do: if you try to make an EQ by ear, you will also end up correcting the FR issues that come from the differences between your ears and Dolby's, right? Worse still, you will also correct the issues that come from the interaction between your ears and the headphone (the "small room"). So the EQ might sound good to you (and possibly to many) but, to others, your EQ might conceivably sounds worse than straight DH. This would be a problem for a collaborative process to build an EQ. Of course, all of this assumes that such individual differences are actually a substantial problem in practice.
 
Oct 5, 2007 at 2:45 AM Post #34 of 50
i tried the new version and it gives me this error:

Unrecoverable playback error: KS output error: unable to find filter corresponding to selected device

replacing it with normal quality makes the seek bar just go back and forth with no sound and nothing gets played.

onboard sound here.

whihc filters are what? is there a graph for each one?

on the sound for the first version, i felt like the mids or maybe treble was too pronounced, and the stereo effects were the coolest for the channel test. this needs to be applied for a closed/isolating headphone
 
Oct 5, 2007 at 3:07 AM Post #35 of 50
Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
whihc filters are what? is there a graph for each one?


Check filters.zip (attachment or link above).

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
this needs to be applied for a closed/isolating headphone


Well, you could use the matlab scripts to build a filter yourself (they're well-commented) but it would certainly be easier for sound_man to provide you with one. If he's open to requests, I might sneak in a couple as well. :wink:
 
Oct 5, 2007 at 4:23 AM Post #36 of 50
it seems like only a handful of ppl tried this, ill be interested in this soon enough!
 
Oct 5, 2007 at 5:52 AM Post #37 of 50
Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Check filters.zip (attachment or link above).


Well, you could use the matlab scripts to build a filter yourself (they're well-commented) but it would certainly be easier for sound_man to provide you with one. If he's open to requests, I might sneak in a couple as well. :wink:



i was referring to the so called newer treble filters in the new zip, of which there are 2 more than in the filters.zip.

i will look into the matlab scripts, i am somewhat familiar with it, and am taking a class in signals and systems with convolution, fourier transforms, frequency response, phase shifts, and all that jazz. i guess i would take a graph from headroom or something? is there a way i can get that graph in matlab?

so assuming he gets the rest of the chain working perfectly, all we have to do is match the FR of our headphones to what DH expects? and i can turn in some 0 soundstage IEMs into full sized phones?

the new version seems to have 1 or 2 more DSP's in the chain, what do they do? what is new (ie how can i get the old version back because the link is gone and new one doesnt work for me)
 
Oct 5, 2007 at 12:48 PM Post #38 of 50
Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
so assuming he gets the rest of the chain working perfectly, all we have to do is match the FR of our headphones to what DH expects?


So far as I know, he's not working on the rest and we don't know what DH expects so this is really about developing an EQ that sounds good.

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
and i can turn in some 0 soundstage IEMs into full sized phones?


That would be interesting, but I don't see it happening.

The idea is to turn your headphones into Dolby's rooms... in other words, to make your headphones sound like speakers. Not that it's going to work IMO but so what? It's still a nice effect.

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
the new version seems to have 1 or 2 more DSP's in the chain, what do they do? what is new (ie how can i get the old version back because the link is gone and new one doesnt work for me)


I don't know what DSPs you're talking about? I'm aware of FreeSurround, Dolby Headphones and the convolver. Are there new ones in the new version?
 
Oct 5, 2007 at 5:17 PM Post #39 of 50
This is one very interesting thread. I missed it before. (most likely because I don't own a Koss Headphone.

I didn't read everything yet, but I read a lot already and I think people who do own a KSC75 should definitely try this.

I am going to read the rest of this when I've got the time to do so.
 
Oct 5, 2007 at 9:22 PM Post #40 of 50
FYI, I've just bought a pair of '75s and I'm burning them in now, but I'll try this setup out either tomorrow or Monday, and I'll let you know my thoughts. Thanks for your hard work!
 
Oct 6, 2007 at 4:45 AM Post #41 of 50
thanks for the work, i'll give this a try.
smily_headphones1.gif
 
Oct 6, 2007 at 6:21 AM Post #42 of 50
Some of you may have wondered why there's another DSP in the chain of the updated foobar package and why I haven't just posted the changed filters. This is because I took the Koss Filter 3 as a baseline and EQ'd some of its issues with the CurveEQ (with the help of the K701 and a few movie soundtracks). Because the changes are rather complicated and very subtle in places, I haven't yet incorporated them into the source plot, but instead directly captured what the CurveEQ produces. The result is in delta.wav, which therefore contains the changes relative to Koss Filter 3. To use it, one has to add a second convolver after the first one and tell it to use delta.wav. But since you cannot use a DSP plugin twice, I had to use the Stereo Convolver for this job. And that's why I just uploaded the folder instead of writing a guide about how to do this by hand. Of course this is only temporary.

So the situation is as follows:
We have a filter which sounds pretty good with stuff where my K701 also sounds good. And this is mostly classical, soundtracks, and a few better albums. But there's a lot of music which is too bright with that setting (ignoring that it may be possible to adapt to it), so I added a few variants with less treble. One can probably stick with this setting for audiophile music. But the next step is to carefully bring the groove back into the sound signature, to get another setting which sounds natural in the upper frequencies but also makes your foot tap like you'd expect with the Porta Pro :) I've already added two settings which go into this direction, namely Koss Filter 3d_groove and Koss Filter 3d_cleanbass -- please tell me what you think.
The Filters can also be found here.


Quote:

Originally Posted by HFat /img/forum/go_quote.gif
If I had to make a preliminary generalization, close-miked sounds alright and recordings which already have some room ambiance built in sound wrong (the instrument influences that obviously). Largish resonant instruments in general seem to have more issues. One random track I tried which sounded really bad was partly, I suspect, a recording of speakers (live amplified performance). Overall, I'd say many acoustic recordings tend to sound reverby and wrong while electric guitars, electronic music and such tends to sound good to me. Binaural also has weird reverbs (I know it's not supposed to sound right).

Do you really think it's possible to solve those reverbs/room issues with EQ? how?



You may be right in that the reverb of DH is a problem.
Theoretically, we could get rid of that with a bit of work (but EQ won't help there) but I'm not sure if this would kill the experience for the other recordings (those which sound ok). We'll see.

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Do you know about this program's built-in EQ: http://www.cockos.com/wiki/index.php..._Documentation
I'd rather use that because VST in foobar seems like a PITA. Not that I'm any good at EQing anyway.



No, I did't. Indeed, foobar's VST host is not exactly ready for prime-time yet. But it's still ok if you have only a single VST plugin to manage. The cool thing about that EQ is that you can zoom in and build any eq shape you want (because it's based on a spline editor).

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Yes. First, it's often said that one shouldn't compare IEMs with headphones when it comes to Headroom's graphs. Second, look at how the curves evolve, not at some kind of disjointed freq-by-freq average.
I think you've already gone in the right direction by replicating the bass roll-off but you don't have the other part of the "dome" which would form the mid-bass hump and what you have above 1kHz is a smooth roll-off totally unlike the jagged curve that pretty much all headphones have... look at the big hump around 6-8KHz for example. There's a reason the curves look like that (ask the pros for details... I know next to nothing about this stuff).
It might make more sense to use supra-aural headphones as models for the target FR rather than cirucum-aurals if you're going to apply the filter to Portapros by the way.
That said, maybe the FR Dolby assumes is nothing like those headphones and looks more like the ones you imagined. Do we have any documentation, interviews or whatever to give us a clue?



You have a point with the supra-aural phones. Do you know of any good ones which can be used for reference? Because I'd guess that the Koss phones are already among the best of this type...
Btw: There seems to be no documentation around. Probably it's NDA'd.

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
If you can make your Portapros sound like the K701 you borrowed, congrats! I'm not talking about sound quality or some such: just try to get a similar perception of where the sound comes from.


Actually, I was surprised to find that the soundstage of the K701 is not so different from how I perceive DH. As a person who is not yet very experienced with different soundstages I'd say it feels like 2m wide and 50cm deep. What's cool about it is that the perceived sound altitude is pretty much on eye level.
With DH, it feels like 2m wide and 2m deep, but the frontal sounds seem to come from slightly more elevated positions. But I'm not sure that I get the full "K701 experience". This may be due to my soundcard or due to my habituation to the Koss. The problem is that it sounds very bright with a lot of music. The only things that sounds really enjoyable are movie soundtracks, classical music and some 20-30% of the other music.


Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Here's the FR plot for a headphone that's bundled with DH hardware:
http://www.head-fi.org/forums/showpo...0&postcount=38

What about using that as a target? It's more likely the DH unit was tailored to the FR of the headphone than the other way around but you never know.



Thanks, I'll try that. Hopefully the thing which is diffuse-field corrected there is the sound of the headphone and not the plot itself.
... and if you want to sneak in some other plots, please go ahead! :wink:

Quote:

Originally Posted by PeterDLai /img/forum/go_quote.gif
Here you go, they actually have it on their own website: "Such Great Heights" by The Postal Service (192kbps)

I still haven't found anything that comes even close to having this much of an effect.



Thanks, it's nice (you mean the stereo effects, do you?). But I think I cannot fully experience it atm, b/c I don't have your phones. I think yours are somewhat brighter.


Quote:

Originally Posted by HFat /img/forum/go_quote.gif
There was one last thing I wanted to address about post #15 (I'm not going to address everything), regarding DH's limitations:

You brought it up several times. Yeah, this might be part of the reason why I think DH mangles a lot of recordings but it seems to me that this is also a problem for what you propose to do: if you try to make an EQ by ear, you will also end up correcting the FR issues that come from the differences between your ears and Dolby's, right? Worse still, you will also correct the issues that come from the interaction between your ears and the headphone (the "small room"). So the EQ might sound good to you (and possibly to many) but, to others, your EQ might conceivably sounds worse than straight DH. This would be a problem for a collaborative process to build an EQ. Of course, all of this assumes that such individual differences are actually a substantial problem in practice.



Well, if these differences would outweigh the differences between the Koss and the ideal phone for this setup, it would make the attempt futile. But I think they are not that large. If they were, they would also outweigh the diffences between various headphones and any general recommendations such as "buy a DT880, it sounds good" would also be futile. Because I want to create more than just my personal EQ, but something that I could also give to friends etc., it makes sense for me to have a few people vote how a certain filter sounds to them or even contribute by doing their own changes.

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
i tried the new version and it gives me this error:

Unrecoverable playback error: KS output error: unable to find filter corresponding to selected device

replacing it with normal quality makes the seek bar just go back and forth with no sound and nothing gets played.

onboard sound here.

whihc filters are what? is there a graph for each one?

on the sound for the first version, i felt like the mids or maybe treble was too pronounced, and the stereo effects were the coolest for the channel test. this needs to be applied for a closed/isolating headphone



Oh, I've made a mistake here (in the output section of the preferences, I had "DS: SB Audigy 2" selected, instead of the generic "DS: Primary Audio driver".
Its fixed now. Plus, I have added all the other new settings as explained above. So you're on the safe side when you redownload once again (sorry).
The filter descriptions can now also be found in the current filters.zip

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif
i was referring to the so called newer treble filters in the new zip, of which there are 2 more than in the filters.zip.

i will look into the matlab scripts, i am somewhat familiar with it, and am taking a class in signals and systems with convolution, fourier transforms, frequency response, phase shifts, and all that jazz. i guess i would take a graph from headroom or something? is there a way i can get that graph in matlab?

so assuming he gets the rest of the chain working perfectly, all we have to do is match the FR of our headphones to what DH expects? and i can turn in some 0 soundstage IEMs into full sized phones?

the new version seems to have 1 or 2 more DSP's in the chain, what do they do? what is new (ie how can i get the old version back because the link is gone and new one doesnt work for me)



You just type mimic_filter('your_actual_headphone.png', 'your_wanted_headphone.png', 'the_result.wav',44100,2000);
But you won't get that far with the Headroom plots, because they are sometimes far from the actual response. So you'll have to tweak manually.
The additional DSP is a second convolver which adds my recent corrections on top of the sound. It will go away once I have incorporated the changes into the base filter.

Quote:

Originally Posted by baglunch /img/forum/go_quote.gif
FYI, I've just bought a pair of '75s and I'm burning them in now, but I'll try this setup out either tomorrow or Monday, and I'll let you know my thoughts. Thanks for your hard work!


So it seems about time for me to get some KSCs myself!

... time to go to bed right now
smily_headphones1.gif
 
Oct 6, 2007 at 4:13 PM Post #43 of 50
Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
Indeed, foobar's VST host is not exactly ready for prime-time yet. But it's still ok if you have only a single VST plugin to manage. The cool thing about that EQ is that you can zoom in and build any eq shape you want (because it's based on a spline editor).


Have you got a link for this stuff and maybe a howto or some such? No need to write it yourself... we can research it ourselves unless you happen to know where to find a writeup.

Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
You have a point with the supra-aural phones. Do you know of any good ones which can be used for reference? Because I'd guess that the Koss phones are already among the best of this type...


I like the HD25-1 but they're coloured and they're closed anyway. The SR-60 is significantly better than Portapros IMO but, again, colouration ahoy! I'm not really into supra-aurals anyway...

Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
Actually, I was surprised to find that the soundstage of the K701 is not so different from how I perceive DH.
...



You're talking about the potentials areas in which the imaginary sound sources could be placed I assume. Surely you'd agree that the placement is very different for a given recording: with DH it's usually more frontal, distant and less widely spread out for example, right?
I've never heard a K701 but what I found striking about the K501 compared to other headphones is how many recordings were layered in depth (some stuff sounding more distant than the rest). DH doesn't do that for me. Not that it's something I find important... on the contrary, I found the K501's soundstage a bit weird. Still, that's an in interesting effect.

Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
The problem is that it sounds very bright with a lot of music. The only things that sounds really enjoyable are movie soundtracks, classical music and some 20-30% of the other music.


Well, according to Headroom's FR graph, the K701's has exaggerated bass. Maybe what you're missing may be a certain kind of bass such as impact or a boomy effect perhaps as opposed to the actual low-frequency SPL.

IMO, the basic EQ should be done with acoustic recordings if you want fidelity. A fun EQ for the other stuff could them be derived from it.
It's not really my place to give advice about EQing but still, I think it's fair to say that realism is more of an issue with acoustic stuff. So, if that's your goal, those recordings should be your benchmark.

Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
... and if you want to sneak in some other plots, please go ahead! :wink:


For starters, my headphone of choice is the DT880, if only for comfort (Portapros quickly become annoying for me). So it would be cool if you could make a filter with that as a source. I'm not sure how accurate Headroom's plot is however, especially since I've got the older version.

Something else that I think would be neat to test your approach to EQ (as well your scripts and such) would be to make filters that turn two similar headphones into each other. Stuff like HD555<>HD595, K601<>K701, DT880<>DT990, SR-80<>SR-225 and so on... people owning both could then tell us how accurate they think the simulation is. I don't think that this would work very well with headphones that differ significantly in design but, since I've never tried it, I might of course be surprised.

Quote:

Originally Posted by sound_man /img/forum/go_quote.gif
But I think they are not that large. If they were, they would also outweigh the diffences between various headphones and any general recommendations such as "buy a DT880, it sounds good" would also be futile.


I agree, which is why I don't think that the problems many experience with DH can be fully explained by variations in ear shapes and such.

I also happen to think that these general recommendations are indeed futile by the way. There are too many differences, not only between ears, but also between tastes...
 
Oct 6, 2007 at 9:31 PM Post #44 of 50
I made a mistake when I uploaded the delta-corrected Filters/foobar last time.
You cannot use a mono wav for stereo convolver (you need two different stereo files), otherwise the stereo separation is halved and the changes are applied two times (I didn't notice that b/c I was in a hurry with uploading and the treble boost was not fully captured because of a clipping problem). Now it should be alright.

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Have you got a link for this stuff and maybe a howto or some such? No need to write it yourself... we can research it ourselves unless you happen to know where to find a writeup.


The installer can be downloaded here.
To get foobar working with it, one has to activate the "VST brigde" DSP and configure it to use the VoxengoCurveEQ.dll. Then you can start fiddling with it. The little "?" button leads to a menu containing save/load preset, set to default (required when you want to switch DSP chains) and a help screen.
The control area contains a number of red control points just like the sliders of other EQs. But you can right-click each so you can drag it in 2 dimensions, double-click it to remove it, double-click the curve to add a new point, draw a box around multiple points to move them all at once and use the mouse wheel to change the overall strength (distance from 0db) of the currently selected points.

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
Well, according to Headroom's FR graph, the K701's has exaggerated bass. Maybe what you're missing may be a certain kind of bass such as impact or a boomy effect perhaps as opposed to the actual low-frequency SPL.


Yeah, there's enough deep bass (25-40Hz) I would say. But the part from 50 to maybe 700Hz is somewhat recessed, compared to what I'm used to.

Quote:

Originally Posted by HFat /img/forum/go_quote.gif
IMO, the basic EQ should be done with acoustic recordings if you want fidelity. A fun EQ for the other stuff could them be derived from it.
It's not really my place to give advice about EQing but still, I think it's fair to say that realism is more of an issue with acoustic stuff. So, if that's your goal, those recordings should be your benchmark.



Yes, that's pretty much my approach. First fix the issues based on acoustic recordings and then work back to get the rock music etc. sound good, too.
I do most of the EQing with soundtracks such as Gladiator, Pirates Of The Caribbean (I), Once Upon A Time In Mexico, Mononoke Hime, Old Boy, etc. which are all acoustic. And for cross-checking, I use music from Bach, Keiko Matsui, Jon Lord etc. And then comes the shock when I play the other music and my ears get shattered by the exagerrated treble. :wink:


Quote:

Originally Posted by HFat /img/forum/go_quote.gif
For starters, my headphone of choice is the DT880, if only for comfort (Portapros quickly become annoying for me). So it would be cool if you could make a filter with that as a source. I'm not sure how accurate Headroom's plot is however, especially since I've got the older version.

Something else that I think would be neat to test your approach to EQ (as well your scripts and such) would be to make filters that turn two similar headphones into each other. Stuff like HD555<>HD595, K601<>K701, DT880<>DT990, SR-80<>SR-225 and so on... people owning both could then tell us how accurate they think the simulation is. I don't think that this would work very well with headphones that differ significantly in design but, since I've never tried it, I might of course be surprised.



Yeah, that might be a cool thing (and could also help us to systematically assess the correctness of these plots). In the best case, Headroom will be asked by the manufacturers to take them offline :) But I belive there are some systematic problems in the plots (or maybe in my load routine). I'm not so sure whether they are diffuse-field corrected and whether the scale is in fact correct...
 
Apr 2, 2008 at 11:10 PM Post #45 of 50
Dear Friend, I tried your special foobar with my beloved KSC35s and was in ecstasy for a while, it was hard even to change tracks because I don't wanted to miss a millisecond of listening. I just adjusted the level of the whole thing increasing the level only in the stereo convolver. I like to use a very clean chain, just resampler to 48 and the default EQ to my personal taste, ah! and Kernel Streaming. But this thing is so great:
- Much better soundstage:not longer inside my head but around my nose first but after my brain got it moved almost to my computer desk, just like near field monitors woooow. A funny story: The ksc35 are very comfortable for me so I forgot about them, was very late at night and my family is next to the pc room and suddenly I thought: Hey, I'm being too loud here! thinking it was my nice sounding tweaked pc speaker.
- Soundwise, the midrange is sweeter than normal which is a good thing.
- I felt the highs slightly recessed but just a little, in combination with the new midrange sounds like an analogic tube EQ or something.
- The bass is the same, I like it loud so I'm using the Groove file, but works for Classical, Rock, Pop and Jazz.
- Jack Johnson's newest is born for this machine.
- ¿Something bad for you to correct? Let me try it more. Well, I wasn't able to use it in a newer Foobar. Some compatibilty issues.
For now, that's it. Big thanks to you, is very difficult for me to buy new cans now and you gave a great gift!
 

Users who are viewing this thread

Back
Top