Dips around 9k and 17k

[mwcm]

Received an email detailing two issues with Pitcher, the first is that pitcher's output dips lower around 9k and 17k where the original SP-1200 does not, it seems like overall our peaks are a bit lower as well, except for the peaks close to the 9k and 17k dips, which are more emphasized

Original SP-1200

Pitcher

[MFSAKA]

I hope you sort this where it from, brother.
If you need assistance - you know where to find me.

[mwcm]

spent some time on this a little while ago, wasn't able to reproduce the findings in the images though - what I was seeing was a far more pronounced general roll-off of the highs past 10k due to the output filter:

I'm surprised the original images sent to me have so much high end content if they were created using pitcher w/ default settings, would expect gradual rolloff beyond 10k

if the original screenshots were created on audio generated by pitcher (maybe with output filtering turned off?) my best guess would be that either the input filter or the ZOH are at fault

the general shape of the input filter matches the dips above, but the values don't really line up

it's also possible that a ZOH with the right N value could produce a curve with dips at 9k and 17k, tested this a little bit, seemed plausible but was unable to verify this was the case

[MFSAKA]

I use no filter a lot, for 7-8 outputs unfiltered sound. Signal path is as far as i know: - Sampling Input Filtering - ADC 26.04 Hz - 12 Bit Quantization - 8 bit DAC that controls VCA only - ZOH - Output Filtering (for channels 1-6) ср, 4 июн. 2025 г. в 18:27, morgan ***@***.***>:

…

*mwcm* left a comment (mwcm/pitcher#8) <#8 (comment)> spent some time on this a little while ago, wasn't able to reproduce the findings in the images though - what I was seeing was more like a general roll-off of the highs past 10k due to the output filter: image.png (view on web) <https://github.com/user-attachments/assets/551ef679-667e-4c97-adac-f0ac694423e9> I'm surprised the original images sent to me have so much high end content if they were created using pitcher w/ default settings, would expect gradual rolloff beyond 10k if the original screenshots were created on audio generated by pitcher (maybe with output filtering turned off?) my best guess would be that either the input filter or the ZOH are at fault the general shape of the input filter matches the dips above, but the values don't really line up image.png (view on web) <https://github.com/user-attachments/assets/08dd11fb-26a4-4dbf-aca6-dc5134ca1828> it's also possible that a ZOH with the right N value could produce a curve with dips at 9k and 17k, tested this a little bit, seemed plausible but was unable to verify this was the case — Reply to this email directly, view it on GitHub <#8 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIBZ2WJPMQMWYXFHDNYBKM33B4F7HAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNBQGQ3DIMRYGU> . You are receiving this because you commented.Message ID: ***@***.***>

[mwcm]

Ahhhhh gotcha, your freq charts above makes a lot more sense if output filtering was disabled

[MFSAKA]

Try zero crossing processing. In theory it will sound better. Try it. ср, 4 июн. 2025 г., 19:54 morgan ***@***.***>:

…

*mwcm* left a comment (mwcm/pitcher#8) <#8 (comment)> Ahhhhh gotcha, the graphs above makes a lot more sense if output filtering was disabled — Reply to this email directly, view it on GitHub <#8 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIBZ2WI6TLAKRB4KYUVDLQ33B4QEHAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNBQG42DMOJRHE> . You are receiving this because you commented.Message ID: ***@***.***>

[mwcm]

will take a deeper dive into this when I get some free time, might be a while though, have some other projects on the go

if anyone else wants to take a look I think a lack of precision in either the ZOH or input filter definition is a likely culprit here

[mwcm]

Try zero crossing processing. In theory it will sound better. Try it.

ср, 4 июн. 2025 г., 19:54 morgan @.***>:

mwcm left a comment (#8)

#8 (comment)

Ahhhhh gotcha, the graphs above makes a lot more sense if output filtering

was disabled

—

Reply to this email directly, view it on GitHub

#8 (comment), or

unsubscribe

https://github.com/notifications/unsubscribe-auth/AIBZ2WI6TLAKRB4KYUVDLQ33B4QEHAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNBQG42DMOJRHE

.

You are receiving this because you commented.Message ID:

@.***>

Not really sure how zero crossing would be relevant here? Not an signal processing expert or anything so maybe I'm missing something

[mwcm]

Have done some tests vs sp950 over the past few days and am having a lot of trouble reproducing this issue. Haven't managed to create anything with the dips at those frequencies, everything I've made is looking within reasonable range, I see some little dips around 13k and 15k, but other than that it looks pretty reasonable to me

Will try a few more tests against files from the machine itself when I have some time

pitcher

sp950

[MFSAKA]

Hi, compare only with RX1200, brother. SP950 is not legit to original. Inphonik one is the closest model. If you doesn't have a license there's a cracked version out there, or i can ask William a license for you, I'm in alpha tester team. вс, 15 июн. 2025 г., 04:50 morgan ***@***.***>:

…

*mwcm* left a comment (mwcm/pitcher#8) <#8 (comment)> Have done some tests vs sp950 over the past few days and am having a lot of trouble reproducing this issue. Haven't managed to create anything with the dips at those frequencies, everything I've made is looking within reasonable range pitcher_vs_reference_volume_matched.png (view on web) <https://github.com/user-attachments/assets/1ec69e7b-f234-482d-a89b-cffd8725593c> pitcher pitcher.png (view on web) <https://github.com/user-attachments/assets/17ac1f99-e89f-4c0d-8625-9f5091ee25b7> sp950 sp950.png (view on web) <https://github.com/user-attachments/assets/05be14c9-dce1-4a59-b793-088a8cc97602> — Reply to this email directly, view it on GitHub <#8 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIBZ2WJIPJCJA3X7727DVVD3DTGNPAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNZTGQZDGMRYHE> . You are receiving this because you commented.Message ID: ***@***.***>

[MFSAKA]

-1 in SPitcher [image: изображение.png]

…

-1 in RX1200 [image: изображение.png] I speed up sample with Reaper. Here's wav files Spitcher settings [image: изображение.png] Maine it's something related to SRC? I notice that SPitcher make 48K 16 bit wav files only I work in 44.1k on my system вс, 15 июн. 2025 г. в 12:10, MEMFIS AKA ***@***.***>:

Hi, compare only with RX1200, brother. SP950 is not legit to original. Inphonik one is the closest model. If you doesn't have a license there's a cracked version out there, or i can ask William a license for you, I'm in alpha tester team. вс, 15 июн. 2025 г., 04:50 morgan ***@***.***>: > *mwcm* left a comment (mwcm/pitcher#8) > <#8 (comment)> > > Have done some tests vs sp950 over the past few days and am having a lot > of trouble reproducing this issue. Haven't managed to create anything with > the dips at those frequencies, everything I've made is looking within > reasonable range > > pitcher_vs_reference_volume_matched.png (view on web) > <https://github.com/user-attachments/assets/1ec69e7b-f234-482d-a89b-cffd8725593c> > > pitcher > pitcher.png (view on web) > <https://github.com/user-attachments/assets/17ac1f99-e89f-4c0d-8625-9f5091ee25b7> > > sp950 > sp950.png (view on web) > <https://github.com/user-attachments/assets/05be14c9-dce1-4a59-b793-088a8cc97602> > > — > Reply to this email directly, view it on GitHub > <#8 (comment)>, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AIBZ2WJIPJCJA3X7727DVVD3DTGNPAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNZTGQZDGMRYHE> > . > You are receiving this because you commented.Message ID: > ***@***.***> >

[MFSAKA]

*i just check 48k. Run everything (OS, Reaper, SPitcher) after i reboot and set my RME to 48k clock. Same problem with frequency response. вс, 15 июн. 2025 г. в 14:02, MEMFIS AKA ***@***.***>:

…

-1 in SPitcher [image: изображение.png] -1 in RX1200 [image: изображение.png] I speed up sample with Reaper. Here's wav files Spitcher settings [image: изображение.png] Maine it's something related to SRC? I notice that SPitcher make 48K 16 bit wav files only I work in 44.1k on my system вс, 15 июн. 2025 г. в 12:10, MEMFIS AKA ***@***.***>: > Hi, compare only with RX1200, brother. > SP950 is not legit to original. > Inphonik one is the closest model. > > If you doesn't have a license there's a cracked version out there, or i > can ask William a license for you, I'm in alpha tester team. > > вс, 15 июн. 2025 г., 04:50 morgan ***@***.***>: > >> *mwcm* left a comment (mwcm/pitcher#8) >> <#8 (comment)> >> >> Have done some tests vs sp950 over the past few days and am having a lot >> of trouble reproducing this issue. Haven't managed to create anything with >> the dips at those frequencies, everything I've made is looking within >> reasonable range >> >> pitcher_vs_reference_volume_matched.png (view on web) >> <https://github.com/user-attachments/assets/1ec69e7b-f234-482d-a89b-cffd8725593c> >> >> pitcher >> pitcher.png (view on web) >> <https://github.com/user-attachments/assets/17ac1f99-e89f-4c0d-8625-9f5091ee25b7> >> >> sp950 >> sp950.png (view on web) >> <https://github.com/user-attachments/assets/05be14c9-dce1-4a59-b793-088a8cc97602> >> >> — >> Reply to this email directly, view it on GitHub >> <#8 (comment)>, or >> unsubscribe >> <https://github.com/notifications/unsubscribe-auth/AIBZ2WJIPJCJA3X7727DVVD3DTGNPAVCNFSM6AAAAAB2BZ72MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSNZTGQZDGMRYHE> >> . >> You are receiving this because you commented.Message ID: >> ***@***.***> >> >

[mwcm]

-1 in SPitcher
[image: изображение.png]
…

image isn't there

[mwcm]

rx 1200 -2

pitcher again

looks about the same to me, maybe slight differences around 11 and 15k