juke
very active
Posts: 396
|
Post by juke on Apr 22, 2013 21:37:00 GMT
Frans Even with my limited knowledge it looks to me like you've worked miracles!
I'm sorry you got a fairly duff pair from me to work on but I'm really pleased to see such results, I wouldn't have thought it possible.
Looking forward to their arrival!
Syd
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Apr 22, 2013 21:43:56 GMT
Hi Syd,
I wil TRY to ship them on Wednesday....
|
|
juke
very active
Posts: 396
|
Post by juke on Apr 22, 2013 21:56:18 GMT
Hi Syd, I wil TRY to ship them on Wednesday.... ;D ;D ;D
|
|
|
Post by c61746961 on Mar 2, 2014 19:27:36 GMT
I have tried to 'remove' the dip (between 1kHz and 8kHz) in several ways but can't get it to go away. Luis seems to have done it but doesn't use Shure pads and think that's the trick. I think it's the pads that cause it. So I solved it electronically as it is easier to duplicate I don't think he has, but it does seem to be possible to lessen the dip, based on micmacmo's insights over at head-fi (sorry, no linky), I used a square of the lens-cleaning type of microfiber cloth as part of my damping scheme, and via a sine sweep it seems to reflect a considerable chunk of the 2 - 5 KHz energy back at the driver and to the ears. You will have to considerably tweak your damping, though, and possibly use a denser material on the ear side to tame the 10 KHz peak as well.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 2, 2014 22:07:34 GMT
The dip can be lessened by a few dB max using the backwave. On the other hand you also get cancellation on certain frequencies.
That 1 or 2 dB you can gain isn't enough to compensate for the 8 to 10dB that is missing there.
Taming the 10k peak (as it isn't a resonance) also means the adjacent bands are lowered as well. The problem there is above 10kHz there is a severe drop off already as is the case below 10kHz.
have tried a few things myself and am still amazed people keep trying to make the T50 driver do things it can't passively. It needs corrections of well over 10dB.
I can only listen to a (modified) T50RP with EQ or the correction filter. In that case there is no need to use the backwave and thus also aren't plagued by the downsides (comb filter effects) be them small anyway.
|
|
|
Post by c61746961 on Mar 2, 2014 22:33:47 GMT
I see. Do you think it would be easy to target a FR like this one or this one (sans the sub-bass rolloff) using passive filtering?. I've found out I'm fairly sensitive to treble peaks and I don't think I could stand a treble that isn't a little rolled off, my considerably peak-tamed T50's already made a fair chunk of my music close to unlistenable : ( Also, what do you think would be a good strategy for modding the T50RP while keeping the THD as low as possible?
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 2, 2014 22:48:43 GMT
Tyll's plots are a bit difficult to compare with my plots as Tyll decided a long time ago to use a 'x-field' compensation ON TOP of the compensation needed for his coupler.
So when the plot from Tyll rolls off from 2k downwards (making it seem rolled off) when measured the way I do it would not be rolling off at all and isn't as well. So when Tylls plots are 'flat' they are very hot in the treble in reality. Tyll would have to recalculate all new pdf's (the same is true for HR) which would be a lot of work and isn't needed when you only use them to compare to his other plots (like HR does in their case as well)
That peak is there on the T50 driver and usually resides between 8kHz and 10kHz (as Dan also found) depending on the driver. The ones I had here were closer to 8kHz though but may also be because of the used coupler method. In any case the peak is too high and I lowered it with the white felt. In that case the rest was lowered as well and I compensated that with the filter.
It seems to be the easy way.
I am not particularly bothered by the (relative to the expensive orthos) higher THD. Maybe the felt in front of the driver helps there too. I can't do THD measurements correctly... need to build another coupler amplifier that is less noisy to be able to see the harmonics. Now they drown in amplifier noise which is not of importance for just making plots. Accurate FR is paramount there though which is what the amp does.
|
|
|
Post by c61746961 on Mar 2, 2014 23:04:37 GMT
What do you think an Olive-Welti target compensated driver should measure like with a simple coupler like yours?, flat but with a ~6 dB boost in the lows from 20 to 200 Hz, or also with a -4 dB high shelf at 1 - 3 KHz and then onwards? I want to look into using my father's audio equipment and build a coupler like yours to measure and mod my headphones, and I want to target the Olive-Welti target and I can only see measurements taken at the eardrum, which is no use to me since I will not be getting a HATS anytime soon ; D
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 3, 2014 6:00:03 GMT
I have been meaning to 'convert' the olive welti target to a 'flat' measurement rig. The plots shown on the web are indeed to compensate a RAW measurement and do not apply to simple rigs. I suspect it will be close to a flat line with a small boost in the lows and maybe some shelving in the highs. The reason for the boost in the lows is the boost speakers have in a room and the slight loss in highs is because of lack of room reflections for higher frequencies. The OW target is simply another correction to make a headphone have the same FR as a (flat) speaker would have in a fully treated and optimised listening room with flat speakers in it. A clue can be found on Rin Choi's website. (plot taken by Rin Choi) He has OW targets for the BMF and Paradox. When you look at them you will need to apply the reciproke of the shown curve. The same goes for the Paradox which is similar to the BMF but with more damping in the lows. I suspect the slants are closer to the BMF (regarding the lows) Now have a close look at the filter curve of the T50RP filter and you will see they match closely. This means you will have to boost the sublows a few dB (check filter) boost around 10dB near 5kHz, filter the peak at 10kHz by a good 10dB but leave out the part where I extend the highs. In case of the Paradox you may notice that it follows my filter quite closely but needs some extra highs above 10kHz. Here is a quick sketch with the filter action and the resultant OW target. It only lacks a bit in the lows where OW likes to see more bass (I and others do as well) In essence all that needs to be done is turn up the bass in the filter. Have done so and makes it sound fatter and muddier and prefer it with just a tad of extra bass which would be somewhat below the OW target regarding the lows. Also the filter curve is for T50RP´s with felt in front of the driver which lowers that peak by a few dB. I personally think that an OW target removes too much 'air' though. It might also be that my older ears need a bit of a boost above 15kHz to sound 'correctly' though. The amount of correction needed for a T50RP is impossible to do with passive filtering nor with acoustic trickery. You can only gain a few dB here and there with those but will also address adjacent frequencies which proper EQ won't.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 6, 2014 7:35:58 GMT
The K812 measured by Changstar measured like THIS. If this is closest to O.W.target Harman can do than it would measure near flat on my rig with a slightly raised bass (few dB) and slightly raised highs. My rig is compensated around 16kHz and for the lows so would measure different on a flatbed coupler without any compensation. About the same as the filter corrected T50RP. In the meantime I did raise the bass a bit after this measurement was made by removing some damping wool in the cups so it now has slightly raised bass as well. Sounds pretty 'reference' to me and MUCH different from stock nor any modded T50RP. The Paradox (nor other modified) T50RP's distort nearly as much as the expensive K812 though they also show much less ringing than this flagship.
|
|
|
Post by c61746961 on Mar 7, 2014 18:02:47 GMT
Yeah, aside from the FR that looks pretty abysmal for a flagship... The ringing looks even worse than the TH-02, haha... I think I've got my T50RP fairly flat up until ~2 KHz, and I've played with a couple of parametric equalizers to simulate what your filter does. The boost in the mids-presence region is noteworthy but not night-and-day to my ears, neither is the boost past 12 KHz. More notorious is boosting the lows even by 2-3 dB and getting rid of the treble peaks. I will try to hold my efforts until I can build a rig to measure and compensate empirically, and I think I will go for a flat treble and boosted lows.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Mar 7, 2014 18:26:23 GMT
I think I will go for a flat treble and boosted lows. That's just about what I prefer tbh. A healthy bass boost does no harm imo since it makes the headphone more 'speaker like' tonally.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 20, 2014 22:48:40 GMT
|
|
|
Post by c61746961 on Mar 21, 2014 1:00:52 GMT
Wow, excellent and very in-depth, congrats and thank you for improving this hobby for us all.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Oct 19, 2015 18:34:51 GMT
I played around with Brainwavz HM5 and a few other pads on the T50RP today. Have 2 versions of the HM5 pads the leather version and a velours version. Turns out the HM5 pleather pads can be substituted for the SRH940 pads. The HM5 pads are softer and more comfortable than the SRH940 pads. Below the HM5 pleather and SRH940 pads compared. The T50RP is NOT a stock version but is modified with damping materials inside and felt in front of the driver which reduces the peak. They look very similar except for the sharp nulling dip around 7kHz. This may well be beneficial for non-modified T50RP's though. Below the difference between the velours and pleather version: The velours version almost has 10dB less bass. The differences between the pleather SRH840 and velours SRH940 is very small but these are not. While I was at it and had a few other pads around that could be made to fit I measured these too. All of them in one plot: Shure SRH 940Shure SRH 1840 (also velours and not equal to the SRH840) Philips cityscape uptown padsHM5 pleatherHM5 veloursThe HM5 pads also have been measured on the Sennheiser HD650
|
|