Thanks feedback, maybe explanation for what happening in "OB-1_test" correction is in place then, overall its same as original the one with 25x IIR filters plus two more normal IIR (minimum phase) and a third FIR filter manipulating phase.
1st filter is IIR low shelve 100Hz / Q0,7071 / -3,5dB, idea behind is because Harman curve is supposed to replicate good speaker in good room and that curve they scientifically researched also and is to see as red curve below. It is very hot in lows and my thinking for dial down the worst of it is because my favorite curve for linear phase speaker is the black one below and head phones should be linear phase in pass band because there is no XO point happening, where i can think of why they prefer so hot a curve is because real time speakers have XO slopes in lower region so a more hot curve would probably help mask a bit and speed up phase to have little less the group delay that develop using XO points. Blue curve below is the B&K research from past.
2nd filter is sensed by listening over and over and become a broad IIR high shelve 5kHz / Q0,4 / -0,5dB to help tip bright tonality a bit. Can't document any objective curves for this correction but in its broad with Q0,4 it should be non resonance neutral.
3rd filter is FIR that slow phase a bit from mids up towards VHF, its also broad and technical its 2nd order / Q0,5 / 30kHz. Idea and thinking behind it is all devices in recording and replay chains are bandwidth limited at highs and when summed they all push to final phase curve from mids and up will tilt heavy in speedier direction than lows, so its based guessing/prediction that to get closer to phase of original instrument that sat in front of microphone at recording studio some broad correction for all these phase distorting devices is maybe help into better realism, and of course listening over and over again was reason for it was added after trials of different values in filters settings.
There is many ways to get Windows stream thru JRiver DSP and think there is good help over their web site, as long as using only IIR filter correction lip sync is no problem but using FIR filters will need JRiver to do all the media stuff and help over there is probably needed. For your case using the 25x IIR filters set in PEQ container, a fast way is to turn on JRiver virtual sound card and set it as default sound card in Windows sound settings plus make shure JRiver is always open. Go to "Tools/Options/General/Features" and tick on "WDM Driver" and restart JRiver, now go to Windows sound system control panel and set JRiver virtual sound card as default sound card and preferred sample rate then you should be there and as long filters is IIR have no movie lip sync problem.
Attachments
Last edited:
Byrtt,
I turned off that "Normalize Filter (Recommended)" and WOW, made all the difference. Sounds fantastic now. Balanced with distinct powerful bass and clear smooth mids and airy highs. The OB-1's now rock! That little radio checkbox option is the dumbest thing - why would that be recommended?
I can immediately tell the sound is something special now. Stunningly good.
I have only listened to it with balanced drive (capable of double the voltage swing and 4x the power) and it can definitely get way louder than I can take if volume knob moved too high.
Thank you, Byrtt!
I turned off that "Normalize Filter (Recommended)" and WOW, made all the difference. Sounds fantastic now. Balanced with distinct powerful bass and clear smooth mids and airy highs. The OB-1's now rock! That little radio checkbox option is the dumbest thing - why would that be recommended?
I can immediately tell the sound is something special now. Stunningly good.
I have only listened to it with balanced drive (capable of double the voltage swing and 4x the power) and it can definitely get way louder than I can take if volume knob moved too high.
Thank you, Byrtt!
Last edited:
Great X it works better now, also that we closer in agreement about sound quality : )
Think why that is expressed as recommend and ticked on as default is if filter with dangerous boost levels is turned on by accident or non experienced user on his thousand dollars system with volume set to twelve o'clock or more his cones can end flying thru the room and amps break fuses.
If you look below finale correction curve for OB-1 convolution file on a +/-10dB scale its balanced hovering around zero dB line so as even down at lows where boost is higher (+4,57dB @20Hz) it shouldn't be problem in tracks have no energy down there other than bit of DC thumbs that probably create and add bit of realism.
Think why that is expressed as recommend and ticked on as default is if filter with dangerous boost levels is turned on by accident or non experienced user on his thousand dollars system with volume set to twelve o'clock or more his cones can end flying thru the room and amps break fuses.
If you look below finale correction curve for OB-1 convolution file on a +/-10dB scale its balanced hovering around zero dB line so as even down at lows where boost is higher (+4,57dB @20Hz) it shouldn't be problem in tracks have no energy down there other than bit of DC thumbs that probably create and add bit of realism.
Attachments
Last edited:
BYRTT
Thank you for a detailed explanation of the added filters. I am not sure that I understand all the technicalities, but I believe you. Also thank you for the info on streaming through JRiver using a virtual sound card. Very interesting, and I guess I will have to purchase JRiver now – my copy will expire soon.
X
Yes, it is recommended (by BYRTT) that both the “Normalize filter volume (recommended)” and “Automatically switch filter based on input format (recommended)” options must be switched OFF. Glad you discovered that. As I mentioned previously, you can switch the Convolution option on or off while the music plays for AB comparison. There is a slight difference in overall volume though, but easy to adjust quickly with the amp volume.
I do agree with you X that on some tracks the mid bass could be a bit more pronounced. If we know which frequency range it is I assume one could experiment and adjust one or two of the filters?
Thank you for a detailed explanation of the added filters. I am not sure that I understand all the technicalities, but I believe you. Also thank you for the info on streaming through JRiver using a virtual sound card. Very interesting, and I guess I will have to purchase JRiver now – my copy will expire soon.
X
Yes, it is recommended (by BYRTT) that both the “Normalize filter volume (recommended)” and “Automatically switch filter based on input format (recommended)” options must be switched OFF. Glad you discovered that.
As I mentioned previously, you can switch the Convolution option on or off while the music plays for AB comparison. There is a slight difference in overall volume though, but easy to adjust quickly with the amp volume.I do agree with you X that on some tracks the mid bass could be a bit more pronounced. If we know which frequency range it is I assume one could experiment and adjust one or two of the filters?
Listening more, really enjoying the new sound from the OB-1's. However, the DT880's without correction also sound quite nice. They still seem to have a tad more midbass presence. The convolution seems to lift a fog of mid-bass mudiness when applied on the OB-1's. It's like the instruments and vocals all separate out and are more distinct.
When you have time Byrtt, if you can make a DT880-250 convolution file for me that would be great. No rush. Would like to see what improvements, if any, can be had on those and played single ended through my top amp, the Silicon Harmony.
When you have time Byrtt, if you can make a DT880-250 convolution file for me that would be great. No rush. Would like to see what improvements, if any, can be had on those and played single ended through my top amp, the Silicon Harmony.
Hi X,
Regarding DT-880 correction look post 953 that have either all the filters data to set in PEQ container or use the delayed convolution file.
About OB-1 myself hope get either a little new replacement driver or they allow me to buy one as a spare part so i can repair and listen them in same condition as you and twocents.
Regarding DT-880 correction look post 953 that have either all the filters data to set in PEQ container or use the delayed convolution file.
About OB-1 myself hope get either a little new replacement driver or they allow me to buy one as a spare part so i can repair and listen them in same condition as you and twocents.
Last edited:
Hi Byrtt,
I have been testing both the OB-1 and the DT880 with their respective convolver. Not surprisingly, they both sound excellent and similar. The OB-1 in balanced drive can get way louder thouigh. What's interesting is that the convolver engine can process the OB-1 wav file at 24x while the DT880 is slower and is 9x only. There is a larger noticeable delay for the DT880 one to start working once it is invoked. I wonder if there are more FIR processes going on with the DT880 file?
I have been testing both the OB-1 and the DT880 with their respective convolver. Not surprisingly, they both sound excellent and similar. The OB-1 in balanced drive can get way louder thouigh. What's interesting is that the convolver engine can process the OB-1 wav file at 24x while the DT880 is slower and is 9x only. There is a larger noticeable delay for the DT880 one to start working once it is invoked. I wonder if there are more FIR processes going on with the DT880 file?
I get mine from Aliexpress, but I have seen the parts listed for either Digkey or Mouser. I seem to recall having a hard time finding it on Mouser.
Its good to know saying they sound excellent and also similar, thanks for that feedback.
The different process time was as said because I'm not good in optimize numbers of taps/FFT length for IIR filter, in past mostly used FIR filters and being lazy for user input simplicity simply allocate more taps than is needed because we can (computer) and that cost added delay, and the difference you hear in delay is 8192 verse 262144 taps.
Find below at 44,1kHz rate OB-1/DT-880/OB-1_TEST set to same 8192 taps so delay is minimal and the same, only one using FIR is OB-1_test the others is pure IIR filtering.
Last edited:
It depends : ) and for detail whats different to normal OB-1 correction look post 982 where all three added filters is described, honestly think analytic we have to think about our setups chains all included are far from being same except for PCA and OB-1.
In general myself is happy both versions for OB-1 but the one with "test" name take subjective level higher and twocents comments is quoted below and seems a bit into same direction. That said my OB-1 left channel was lacking performance in low end because of defective driver unit so its hard to say if test version is so much better in reality because the three added filters are relative broad and gentle. By the way even i told about intensions of modification before executed to Status Audio and later told about the defective impedance curve for left channel driver unit, they mailed today they shipped me a new driver unit and is very happy for that service because in first place touching anything at all was prepared to be on my own regarding warranty.
For technically interested how driver unit into OB-1 looks there is good pictures on below aliexpress link.
42mm speaker unit diy headset unit HD650 unit 54ohms(for open earphone housing)-in Earphone Accessories from Consumer Electronics on Aliexpress.com | Alibaba Group
Last edited:
I was preparing to post about this very phenomenon when I saw you two had beat me to it.
After trying several different output cap combos in both my PCA builds, my new goal for this second build is to achieve the best possible sound with no bypass at all. Here are the opinions I've formed through my listening tests while rolling caps with this amp:1) Small value film bypasses (on the order of 1/100 main value) make a negligible contribution to the sound of a good electrolytic. Sometimes I think I can hear a little extra smoothness up top with a film bypass, but they fail to add substantial air or detail.
2) Larger value electrolytic bypasses (around 1/10 value) make a definite difference. For instance, bypassing a KA or OSCON with a KZ or Silmic definitely opens up the sound and brings in some additional treble information and bass texture.
3) Now the caveat: Every bypass arrangement I've tried, to some extent, has sacrificed the precise timing of the music. I opted for a Silmic bypass for the KA I used in my first build, because it ameliorated the somewhat stuffy mid-centric response of the KA, giving an overall improvement. But even in this case, I noticed a subtle loss of focus.
"Smearing" is a good word for it. To me, listening to two different caps in parallel is something like having very slightly blurred vision. Transient attacks lose their sharpness, so that the music loses the immediacy of its overall presentation. A single cap sounds more perfectly focused to me. To date, my favorite output cap for this amp is the 25V 1000uF Panasonic FR. I've tried several bypasses with it, and nothing so far sounds as clear and precise as the Panny standing alone!
I just did a shootout between this Panasonic and pretty much every audio grade Nichicon cap (sans KZ) in the 10x20mm size, including the Fine Gold. Panny slayed them all. With the Pannys in place, I played the amp today for my buddy who has a degree in audio engineering. He kept repeating the words "unbelievable!" and "crisp!" He also commented on the superb instrument separation and his jaw dropped when he heard the bass drop
I am still waiting for a few more challengers from PartsConnexion--namely Elna Silmic and Cerafine, and one Nichicon KZ--which is too big to fit my current build, but I want to hear it anyway. Will report back. By the way, PartsConnexion is running a 25% off sale on almost all items this month.