CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS
Lobby => Headphone Measurements => Topic started by: Marvey on June 19, 2012, 04:16:31 PM
-
Fostex TH900 Frequency Response and CSD Waterfall Plots
(http://www.changstar.com/index.php?action=dlattach;topic=351.0;attach=1373;image)
Posted on HF first since request came through HF. Just posting here for completeness.
-
Interesting. Did you cave in and buy a pair of these?
-
Interesting versus the D7000.
-
HF first? WTF! You know we have an auto request on 24/7 :p
-
Told you these were super bassy Max.
-
Thanks Purrin!
Do the graphs say that they are good? I mean... correct me if I don't see it correctly. But I don't see purple or pink, I see deep blue... safer than most cans?
-
Have you read these Hawaiian?
http://www.changstar.com/index.php/topic,120.0.html (http://www.changstar.com/index.php/topic,120.0.html)
http://www.changstar.com/index.php/topic,71.0.html (http://www.changstar.com/index.php/topic,71.0.html)
The colors are there just to make it easier to see. Look at the horizontal axis on the right, and you look for how far the ridges move out over time.
-
Oh no I haven't. I'm such a noob. Thank you! I just read through them and have a better understanding of this now. Now only to find a HP that doesn't have any spikes....
So for the TH900 they are pretty good. On the 1k is considered a spike, right?
-
bump at 1k on the FR. spike at 7k on FR with corresponding ridge on CSD.
-
Told you these were super bassy Max.
Except that they aren't. The graph shows response you normally see for IEMs with boosted bass, or Monster Beats, but the TH900 I heard sounded nothing like that.
For comparison, Monster Beats: http://graphs.headphone.com/graphCompare.php?graphType=0&graphID[]=1443
Pretty similar except the bump is moved up in the mids a bit, and lacking the extension to 20KHz.
-
I don't doubt the Beats are even bassier but I've never heard them. Using the HR measurements in comparison to Marv's is a no no. Has Marv measured those yet?
Maybe you heard a beta version of the TH900 or something.
-
These are a bit bassy to my ears, but not extremely so. It's a moderate emphasis. The D7000, LA2000, DX1000, Ed8, Ed9, and H2+ are all bassier to me. I want to say the PS1000 is bassier too.
-
They have about same bass as D5000, but with much less distortion. So it sounds much cleaner with less bass doubling an octave higher. Bass is recording dependent. Treble also balances out peak. The bass is bothersome to me sometimes. I shelve the bass down -5db with EQ.
-
I don't doubt the Beats are even bassier but I've never heard them. Using the HR measurements in comparison to Marv's is a no no. Has Marv measured those yet?
Maybe you heard a beta version of the TH900 or something.
I didn't spot his measurements for them, my only recourse was HR. It could be true about them not being the final tune. If they did change them, I'm sad, cause they did sound pretty nice, and I have warmed up to the way they sounded since I made my original comments.
-
Be wary of ye ole output Z w/ the TH900 when comparing impressions.
-
Not sure if I should be impressed or disappointed by the measurements here.
Even without factoring in the price.
The FR isn't too bad, and the CSD shows a well behaved enough decay, but.......
On a side note.. I do own the Ed8 LE.. so FARK THAT!! AMMA BUY THIS TOO!!
-
The TH900 is a much, much better headphone than the Edition 8.
-
I've gone through 2 rutheniums and 1 palladium. Somehow, the one I currently own is the only Ed8 I find pleasing.
But I... we shall see.. if there is a willing buyer for them at about 1200USD, then maybe the TH900 is plausible. I'm able to get them brand new for about 1500 from a special store :P
-
Assuming I'm understanding these correctly, it seems that there is a little ringing just under 20k? Might that be subtly be enhancing the sound in any way or is that little ringing probably doing very little to the overall sound?
-
Probably totally inaudible if you are older than 18.
-
Audiowood opened these up: http://www.head-fi.org/t/595683/fostex-th900-impressions-discussion-thread/2790#post_9006193
They have, like, dampening and shit.
-
Probably totally inaudible if you are older than 18.
Ouch, I just checked and it seems I can only hear up to about 15KHz..
-
Reviving this thread again since I'm jumping back and forth between here and HF, and I'll likely get some more brutally honest opinions about amplification here on the TH900s vs conjecture and hypotheses over there.
Finally got my PWDII back and hooked up to the X-CAN V3 that I was previously connecting directly to PC to burn in the TH900s a bit. Things definitely opened up a lot more, a bit more detail & spatial (ie. sounds more opened up than closed in) portrayal and further dialing in on top and bottom end extension at the frequency extremes. Glad to have my PWDII back (2.02 firmware).
I do get the sense that the X-CAN V3 (with X-PSU V3) is providing a more "bloated" picture, and hence making some types of music sound more "slower" than it should be, namely with electronic music. It's also a darker sounding amp with its tube stage in the circuit.
So, looking at some sub $1K options to essentially drive these headphones. May stretch a little over if need be but wondering what have you guys actually LISTENED to which gives me a little more neutrality than what I have via the amps. Of course, I'm well aware that the TH900s are colored so no need to emphasize this point. I believe that the colored nature of the headphones themselves is what I want to preserve. I just need something on the amplification side which doesn't further color the headphone's signature.
Thanks in advance.
-
If you want more neutrality, I'd EQ them. No change in amplification is going to have as large a difference in colored/neutral as its so darn boosted in the low end.
-
If you want more neutrality, I'd EQ them. No change in amplification is going to have as large a difference in colored/neutral as its so darn boosted in the low end.
Thanks for the tip, Tari. Guess what I'm asking for is about changing the subtle shade of grey to be lighter than darker. :-\
I've read purrin's review on this. EQ is an option but something I'd rather not fiddle with if I'm playing CDs through a transport directly into the PWDII as well. Not sure about PWD filter options, which I haven't played around with yet.
Also, some of my music listening will be through streaming services, such as MOG. Are there any EQ applications which sit between the streaming app and the output?
-
Could anyone share some thoughts on the 'impact' on the by comparison to other cans? Because i have been sitting swapping between a T1, HD800, LCD2r2 and a SA5000, and I realized namely the T1 barely has any impact, the Hd800 and the Sony are pretty decent on this front and the impact of the Audeze seems about the same but sort of drowns in its dark and by comparison somewhat muffled sound signature.
I noticed at a meet a bit ago that the Hifiman He-4, I beleive, has some incredible impact compared to most things I've heard (pretty much everything except rarities like R10 etc.). As far is I can tell its purely a matter of how much air the drivers are moving to reproduce a given frequency at a given volume.
I am curious how the th900 does in this regard and if anyone has suggestions of other reference cans with good impact, feel free to chime in please p;)
-
Need to try figure out my ears some time. Using sinegen, I hear treble peaks pretty much at 5KHz and 10KHz. Doesn't quite match FR graphs from anywhere I have searched. Anyone else?
http://en.goldenears.net/17034
www.innerfidelity.com/images/FostexTH900.pdf
http://personalaudio.ru/raa/otchety/naushniki/fostex-th-900/
-
Need to try figure out my ears some time. Using sinegen, I hear treble peaks pretty much at 5KHz and 10KHz. Doesn't quite match FR graphs from anywhere I have searched. Anyone else?
http://en.goldenears.net/17034
www.innerfidelity.com/images/FostexTH900.pdf
http://personalaudio.ru/raa/otchety/naushniki/fostex-th-900/
How so? Most of those show peaks close enough 5k and 10k. The Russian one is just over smoothed. It's not unusual to see a discrepancy of around 500hz-1k between various rigs and calibrations. Don't expect electron microscope or even caliper level of precision with these graphs.
As always, it's the relative analysis with context that matters, not static absolutes.
-
Need to try figure out my ears some time. Using sinegen, I hear treble peaks pretty much at 5KHz and 10KHz. Doesn't quite match FR graphs from anywhere I have searched. Anyone else?
The 5 KHz peak seems to appear in measurements that use a simulated ear. There's always a narrow peak between 9-10 KHz with any headphones for me, it's due to my ear.
-
Need to try figure out my ears some time. Using sinegen, I hear treble peaks pretty much at 5KHz and 10KHz. Doesn't quite match FR graphs from anywhere I have searched. Anyone else?
http://en.goldenears.net/17034 (http://en.goldenears.net/17034)
www.innerfidelity.com/images/FostexTH900.pdf (http://www.innerfidelity.com/images/FostexTH900.pdf)
http://personalaudio.ru/raa/otchety/naushniki/fostex-th-900/ (http://personalaudio.ru/raa/otchety/naushniki/fostex-th-900/)
CSDs are a better way to confirm treble peaks. Look at GE's CSD graph. See the ridges at 6-7k and at 12k? Exists on both of our CSDs. Note GE's CSD plots are smoothed to hell.
I've never found singen and tone sweeps reliable. Mainly because they won't necessarily pick up sharp high Q peaks and our ears have variable frequency response based on volume. 1/3 octave noise works better IMO.
-
How so? Most of those show peaks close enough 5k and 10k. The Russian one is just over smoothed. It's not unusual to see a discrepancy of around 500hz-1k between various rigs and calibrations. Don't expect electron microscope or even caliper level of precision with these graphs.
As always, it's the relative analysis with context that matters, not static absolutes.
I see. It never connected in my head that the peaks on the graph could be similar to what I was hearing but shifted higher or lower. This really helps. Thank you.
The 5 KHz peak seems to appear in measurements that use a simulated ear. There's always a narrow peak between 9-10 KHz with any headphones for me, it's due to my ear.
Thanks, I guess I need to check on that. I recall that my Sennheiser HD250 II has a huge peak around 10KHz even though the one or two graphs that I found of it didn't really show that at all. The only headphones which didn't have any annoying (sinegen) treble peaks was the Yamaha YH-1, but that one is on ear (and probably known to have a relatively rolled off treble response).
CSDs are a better way to confirm treble peaks. Look at GE's CSD graph. See the ridges at 6-7k and at 12k? Exists on both of our CSDs. Note GE's CSD plots are smoothed to hell.
I've never found singen and tone sweeps reliable. Mainly because they won't necessarily pick up sharp high Q peaks and our ears have variable frequency response based on volume. 1/3 octave noise works better IMO.
Thank you. Come to think of it, the graph here is really pretty much what I've been hearing, just shifted up a bit. I just never realized that those treble peaks were so prominent, but not necessarily painful or anything. Could also partially be because my earpads are getting rather worn and their overall shape has rounded a bit.
I was trying to figure which of my headphones was the "flattest" in the treble with sinegen. (conclusion: Yamaha YH-1, but I probably won't stick with those for long)
Is there any Windows program in particular that you would recommend for 1/3 octave noise?
-
Is there any Windows program in particular that you would recommend for 1/3 octave noise?
I have created third octave noise files from 20Hz to 16 kHz with Audacity. The effect I used is called Glame Analog Bandpassfilter. Maybe marvey can tell if these sound correct, because the filter may not be steep enough.
https://drive.google.com/file/d/0B9lFnfKnrEktZzAwalEteThLR2c/view?usp=sharing