Aug 6, 2013 at 8:55 PM Post #21,916 of 27,319
^ even so i would still be curious to see them especially since very little is around for this older Russian stuff.
 What a great idea to plug up those holes I never thought of it.
By the way how did you get those pads to look so good? Are they aftermarket ( toying with that idea here )
 
Inner foam had turned to gunk in mine so sliced open from the rears and added some thin foam. Might have to do it again properly.
 
 One last thing . I might be delusional but do you find after a brief period of "warming up" so to speak that these sound a bit better, I could be simply hearing things.
 
Aug 7, 2013 at 7:57 AM Post #21,917 of 27,319
^^ The pads are stock. The left ear pad deflated when I used them for only a few minutes. I will use your method of replacing the foam. This will give me a good excuse to open them. I will post pictures of the original damping scheme.
 
I have not listened to them long enough to judge the effects of "warming up."
 
Aug 7, 2013 at 8:53 PM Post #21,918 of 27,319
Quote:
Gentlemen, a fan as I am of vintage AKGs, it seems this thread has gone astray. I'm sure there are other excellent AKG threads where these matters can be discussed and also be more easy to search for those interested.
 
I haven't listened to orthos for a couple of months, been busy with my amp system and trimming my Staxes and... AKGs. Today I decided to try orthos with the system and gathered my HP50s. Turned out there were more than I remembered, even though I have picked up one now and then during the years, when they still are available. Good to have spare parts (drivers tend to fail) and be ready to mod the ultimate HP50.
 
To cut a long story short: once again the time coherence, dynamics and smooth response surprised me. Even though the sound stage cannot compete with the AKGs in size, the precision of placement in the stereo mix is astounding. This is what I call time coherence, and when this clicks the sense of air and space is just wonderful, it's so easy to identify and follow lines and instruments in the music's flow.
 
Some of the HP50s are untouched (good to have as reference), some have been modded to various degrees. What surprised me was that the first one I bought and damped actually sounded best, a battered and worn HP50S with cat scratches and all. Must have look and see what I did to that one... but basically these headphones just need a disc of felt or two.
 

And I thought having 4 KOSS 4AAAs was overkill... 
eek.gif

 
Aug 7, 2013 at 8:54 PM Post #21,919 of 27,319
Quote:
Be warned: measurements are generally frowned upon around here. Though I welcome them.

I don't understand it, but unfortunately this is true. 
 
I personally am reluctant to believe anything without being given measurements. Like how some say that the HD800 is the fastest dynamic headphone there is, even though the measurable attack is clearly inferior to many headphones far below its price range. 
 
Aug 7, 2013 at 10:19 PM Post #21,920 of 27,319
By 'here' do you mean this thread or HF in general? From what I've seen, people in this thread tend to like measurements and graphs, myself included.
 
Also in regards to the HD800 attack, do you actually mean attack, ie the time it takes from the electrical input signal to the initial physical excursion of the transducers, or do you mean the impulse response/decay, ie what happens after the initial said excitation and excursion?
 
Aug 7, 2013 at 11:13 PM Post #21,921 of 27,319
Quote:
By 'here' do you mean this thread or HF in general? From what I've seen, people in this thread tend to like measurements and graphs, myself included.
 
Also in regards to the HD800 attack, do you actually mean attack, ie the time it takes from the electrical input signal to the initial physical excursion of the transducers, or do you mean the impulse response/decay, ie what happens after the initial said excitation and excursion?

The time it takes for the driver to reach its outmost position, so yes, the attack.
 
The decay of the HD800 is unmatched, which is more of a bad thing to be honest. Know all the people saying it's the most microscopically analytical, ruthlessly revealing, amp-finicky son-of-a-bitch pair of headphones ever made? That's why. The decay is fast to the point that it's unnatural, and the sound reflects that.
 
Aug 8, 2013 at 1:07 AM Post #21,922 of 27,319
This being a huge thread dedicated almost entirely to modding, you'd imagine to find graphs galore; yet effectively no one posts measurements here. I've brought this up before, and the consensus in this thread at the time was that verbal descriptions were good enough to convey the effects of mods and that measurements would be unreliable. Ironic and baffling, the thread still remains largely measurement-free.
 
I myself came to wonder the point of putting up measurements here - not because of being in doubt of their usefulness in general but because I've seen that a graph around here has trouble overriding gut feeling, however unsubstantiated the latter may be. Specifically, I recall the recent mylar dust guard incident, that non-starter of a discussion whose counterargument to graphs was made up of 100% magical thinking.
 
Aug 8, 2013 at 7:45 AM Post #21,923 of 27,319
The problem with measurments in a DIY capacity is that they have to be viewed in the context of what they are. They may well be as accurate as Purrin's or Tyll's measurment but each individual rig is going to have variations and not only inter-headphone but inter-measurement. I do not have the equipment needed to take measurements and if I had started early with measuring, I might have been more inclined to hone that skill. I tune mods by ear and as I am generally modding for myself, I tune to what I like. If I had settled on what I perceived to be the perfect response, then measured the mod and it should some nasty dip or peak that I wasn't hearing, would I modify them again to achieve a perfect read or would I question the data as possible artifact?  The value I see in collecting data is to see what impact various techniques / materials have on the FR - this can only be achieved sequentially and after numerous measurements for you to be happy that the changes are statistically significant. ie what materials / techniques get rid of that high frequency ringing in orthos, the fostex drivers in particular? 
 
I guess what I am saying is that I am a staid old f@rt who is set in his ways and trusts his ears ( currently ) more than his ability to set up an accurate measurement rig. BUT I am not averse to measured data and in fact am quite fascinated by the FR and CSDs across the whole range of headphones. Still trying to apply what I see in graphic form to what I hear in the same model as those measured with the caveat that there are manufacturing variances and thus what I am hearing might not be what was measured. I would encourage people who are taking measurements to keep posting them as this will eventually give rise to some level of comparative data which can then be analysed. 
 
..dB
 
Aug 8, 2013 at 8:15 AM Post #21,924 of 27,319
Just a few thoughts....
Does the software that make the measurements compensate for the frequency response and other aspects of the microphone used in each measurement? 
Also there doesn't seem to be one single standard for measurement. All the prolific measurement makers of FR graphs and waterfall charts etc don't match up. They each measure the same headphones sightly differently.
 
The only true standard of measurement which is fully compensated, without flaw and contains all the information truthfully doesn't exist because that requires recording equipment without bias. 
 
 
That said, personally I welcome measurements, but they are only to be compared in the context of the graphs made by the same person who I would hope measures each headphone under extremely low tolerances. Then the subtle variations between headphones become valid and understandable against headphones the viewer is familiar with. 
 
Aug 8, 2013 at 9:44 AM Post #21,925 of 27,319
Well mines do compensate for the microphone but my graphs are very wiggly. I only get nice curves between say 60 Hz and maybe 6 kHz. I do test with sine waves afterwards to see if my ears and graph match. But then we still have different sensitivity to lows and highs; there is no set standard.

At this point I am getting lazy of setting up my measuring kit. Thinking of getting an artificial ear (or moulding mine own) and some dummy head. That way I can at least prevent variation in mic positioning in my ears. However on a dummy head it's harder to validate any seal to the ear.

In any case, I tend to mod by ears too now.
I use graphs to give me an idea of the sound and where I can change. I then use EQ to check if I would actually like that change before trying to tackle it with the measuring kit (to know if the material has any effect on the area of concern).
 
Aug 8, 2013 at 10:14 AM Post #21,927 of 27,319
It's one thing to strive for graphs that convey the sound in absolute terms, and another thing to strive for graphs that convey the sound in relative terms. As dBel points out, and in my opinion, the value of graphs in the context of this thread lies especially in the latter.
 
You measure the frequency response and it shows a dip or two you can't hear, or the treble rolls off prematurely even though you can hear it well-extended. The measured response in that case doesn't reflect your impressions and you could say it's an incorrect measurement. You measure the response again and get the same result. You apply some mod and measure the response a third time - now, in relation to the first measurement, it indicates less treble and a hump in the mids, but is otherwise the same as before. You undo the mod, measure the response, and find it the same as the very first time. You redo the mod, measure, and find the treble again lessened and a hump in the mids. The measurement is still in error in regards to the dips you couldn't hear or whichever other feature you didn't perceive, but contains the meaningful information in relative terms that the mod seemingly lessened the treble and added a hump in the mids.
 
Can you be sure in the above situation that the change in response was entirely due to the mod and with no interference from the measurement setup? No. Can you be sure that you'd correctly hear the change in response due to the mod, or that you wouldn't have a different impression of it tomorrow? No. Can you trust that someone, having done that mod, could accurately describe the effects of it via subjective impressions, omitting nothing that was there and adding nothing that wasn't there? No, you can't.
 
It's interesting that the question of standards should fall on measurements alone - even to the point of bringing up statistical significance - since logically these doubts should apply in equal force to subjective impressions. I've never seen anyone asked to increase the statistical power of their impressions, but I've seen plenty of arguments where one person says they hear this and the other person says they hear that, with no attempt to establish the baseline.
 
Aug 8, 2013 at 11:42 AM Post #21,928 of 27,319
Quote:
 I've never seen anyone asked to increase the statistical power of their impressions, but I've seen plenty of arguments where one person says they hear this and the other person says they hear that, with no attempt to establish the baseline.

So much truth right here.
 
Aug 8, 2013 at 11:44 AM Post #21,929 of 27,319
The reference to statistical significance was really in the context of variance - when you have a sense of the standard deviation between measurements which correlate to a particular modification. I would argue that subjective impressions are so riddled with bias that even trying to apply some level of scientific analysis is inherently flawed. I know this brings up the subjectivist vs objectivist argument but imo , both need to co-exist and are not mutually exclusive to our goal, but both cannot be held to the same analytical standard. 
 
all in , I think we agree with one another - measurements have their value in the process. I will use the Paradox ( T50rp mod ) from LFF as an example of how applying measurements and your own auditory experience can help create a product with great sonic attributes and enable the process to be repeated reliably. The same is true for Mr Speakers and his Mad Dogs - these just went through a few more iterations before hitting the mark, the latest sample I heard were tonally well balanced and suited my preferences. 
 
..dB
 
Aug 8, 2013 at 2:20 PM Post #21,930 of 27,319
If subjective impressions of someone else can be verified neither subjectively by others (you instead rely on good faith) nor objectively via statistical tests, the question is begged why this thread hasn't adopted a more reliable combination of subjective impressions and objective measurements.
 
The technical leap required to produce measurements might on the surface be one explanation, but I might offer that deeper down, the collective aim (or goal) of this thread isn't to increase our knowledge in a concrete way but to serve as a public diary.
 

Users who are viewing this thread

Back
Top