|
Post by jsteiger on Jan 24, 2020 20:29:40 GMT -6
....The moves are reversible because the same exact filter is used in boost and cut.... Not exactly. The bridged-T notch filter that Melcor/API use is pretty accurate (as far as cut/boost frequency) up to 1kHz but starts to shift as you go upwards beyond there. Cut/boost center frequencies run off significantly by the time you get to 20kHz. Just the nature of the beast.
|
|
|
Post by matt@IAA on Jan 24, 2020 21:41:31 GMT -6
....The moves are reversible because the same exact filter is used in boost and cut.... Not exactly. The bridged-T notch filter that Melcor/API use is pretty accurate (as far as cut/boost frequency) up to 1kHz but starts to shift as you go upwards beyond there. Cut/boost center frequencies run off significantly by the time you get to 20kHz. Just the nature of the beast. I'm not sure that we're talking about the same thing. The same two caps and the same resistors to form the bridged T are used in cut and boost, just moved around. Whatever that center frequency is the same amount of boost and cut for that frequency is applied, and is reversible. There's some very HF rolloff due to caps used for stability around the buffers/amps, but we're talking <1 dB and that's there even when set flat. API manual notes the same thing. "When switching an API EQ from boost to cut, the exact same components are used, reorganized around the amplifier for the appropriate function. This guarantees 100% reciprocity. The benefit becomes apparent when a track is re-equalized to remove the previously applied equalization. The engineer can be confident that, because of the fully reciprocal design, a truly flat frequency response will be restored." Thinking about it more, I suppose due to the difference in the source and load impedance presented to the filter between the feedback loop and the buffer there will be some variance between the two positions, but this should be less than a dB.
|
|
|
Post by jsteiger on Jan 24, 2020 23:52:45 GMT -6
Not exactly. The bridged-T notch filter that Melcor/API use is pretty accurate (as far as cut/boost frequency) up to 1kHz but starts to shift as you go upwards beyond there. Cut/boost center frequencies run off significantly by the time you get to 20kHz. Just the nature of the beast. I'm not sure that we're talking about the same thing. The same two caps and the same resistors to form the bridged T are used in cut and boost, just moved around. Whatever that center frequency is the same amount of boost and cut for that frequency is applied, and is reversible. There's some very HF rolloff due to caps used for stability around the buffers/amps, but we're talking <1 dB and that's there even when set flat. API manual notes the same thing. "When switching an API EQ from boost to cut, the exact same components are used, reorganized around the amplifier for the appropriate function. This guarantees 100% reciprocity. The benefit becomes apparent when a track is re-equalized to remove the previously applied equalization. The engineer can be confident that, because of the fully reciprocal design, a truly flat frequency response will be restored." Thinking about it more, I suppose due to the difference in the source and load impedance presented to the filter between the feedback loop and the buffer there will be some variance between the two positions, but this should be less than a dB. Try running some +12/-12 sweeps set to the same frequency and you will see. I am specifically talking about the 550 style EQ's. For example, with a 550B the 20k cut/boost are WAY different. The center frequency shifts dramatically.
|
|
|
Post by Johnkenn on Jan 25, 2020 0:16:31 GMT -6
Has anyone mentioned the CAPI BT50 yet? There’s a shootout thread (here on this board) where the BT50 settled in sonically right between the Vintage API 550 and Vintage 550A. The new 550A was not very impressive in comparison.
|
|
|
Post by Johnkenn on Jan 25, 2020 0:18:53 GMT -6
Btw...I am permanently damaged from this post.
|
|
|
Post by ragan on Jan 25, 2020 0:53:34 GMT -6
I prefer Mhos, myself. But there's no accounting for taste.
|
|
|
Post by chromies on Jan 25, 2020 4:51:06 GMT -6
Has anyone mentioned the CAPI BT50 yet? There’s a shootout thread (here on this board) where the BT50 settled in sonically right between the Vintage API 550 and Vintage 550A. The new 550A was not very impressive in comparison. Ive had a bt50 sitting in my checkout cart for days now. Just taunting me (even tried to get the holiday 10% code from jeff on Instagram 🤣). Came on here because I wasn’t sure why i wanted it.
|
|
|
Post by matt@IAA on Jan 25, 2020 8:25:44 GMT -6
jsteiger ok i see what you're saying. I think there's two things going on. The first is that there is bandwidth limiting all throughout the circuit for stability, so the higher your filter frequency goes the more it starts to be carried along that limiting curve. It kinda skews what the filter looks like. But the biggest one is the small cap in the feedback loop. You feedback high frequencies to make it stable, but that is in parallel with the filter. The boost frequency response changes as you short past the filter at higher frequencies. The cap is very small though, so it only becomes apparent when the filter center frequency is very high. It effectively reduces the amount of boost at high frequencies, so when you reverse that, the cut is bigger than the boost and you see the difference. If you wanted to correct that you could strap the same value (2.2Meg // 68pF or whatever) across the passive cut point. Then you have an identical filter (bridged T parallel with the stability elements) placed both inside the feedback loop and after the buffer. Blue is the effect you're describing; green is when the second cut filter is compensated the same as the feedback loop.
|
|
|
Post by jsteiger on Jan 25, 2020 9:55:35 GMT -6
You cannot correct it. Nothing works unless cut and boost filters are different especially at the high end. I have talked to Paul about this too. Its just how it is.
You also cannot simply change 6 resistors and have a mastering version.
Paul also told me a funny story about the 550A-1. After they shipped their first batch to Sunset Sound, they got complaints that the Q's were way different than the originals. Paul did some sweeps and found they were not proportional Q like the originals. He brought it to Saul who did a 'hand to the forehead' and said they only compared at +/-12 and not the rest of the cut/boost amounts. They had no idea it was constant Q until it was too late.
|
|
|
Post by matt@IAA on Jan 25, 2020 9:58:51 GMT -6
Ok. 🙂
|
|
|
Post by EmRR on Jan 25, 2020 13:59:03 GMT -6
Measurements I took on LC53A for +/-10 @ 15kHz are about 1kHz off from each other. Slightly further apart on the TB550A. The cuts are higher in freq. Neither peak at 15K, both are 13K8. 88K2 measurement bandwidth.
|
|
|
Post by jsteiger on Jan 25, 2020 19:47:24 GMT -6
Measurements I took on LC53A for +/-10 @ 15kHz are about 1kHz off from each other. Slightly further apart on the TB550A. The cuts are higher in freq. Neither peak at 15K, both are 13K8. 88K2 measurement bandwidth. Yes exactly. When I did the LC53A I matched the actual center frequencies of a 550A, not just what is printed on the faceplate. As for the 550A, the math for the 15kHz filter is dead on but in reality...13k8. The gyrator graphics are more perfect reciprocals. Paul also told me that the proportional Q of the GME20/550 was simply a byproduct of the need to adjust the Q resistor to keep the frequency centered at the different cut/boost amounts. This is also why a mastering version is not a simple task.
|
|
|
Post by roundbadge on Jan 28, 2020 4:02:24 GMT -6
the bt50's blow away the current run API 550a's put both on kick drum with identical settings..you'll get the picture pretty quick.
|
|
|
Post by nomatic on Jan 28, 2020 6:27:04 GMT -6
What Hunter said.... The 50s slay the new APIs...
|
|
|
Post by adamjbrass on Jan 28, 2020 6:56:17 GMT -6
I like them, but having no problems living without them
|
|
ericn
Temp
Balance Engineer
Posts: 14,810
|
Post by ericn on Jan 30, 2020 22:01:35 GMT -6
Part of what makes the 550 and others like it sound so good is that it makes you work the basics. A console full of 550’s ain’t a console full of SSL EQ’s, you know you can’t just fix everything with some EQ. You have to get it right to begin with, the EQ will help you shape the tone but it’s not the do everything problem solver.
|
|
|
Post by nudwig on Nov 13, 2020 16:57:50 GMT -6
Thought I'd bump this thread instead of starting a new one. I've been trading gear with a buddy and picked up a pair of the handwired 550a's (the 2004-7 version with fewer frequencies). Was going to try the CAPI CA-0252's in them only to discover that the 2520's are soldered in. Has anybody successfully converted the solder type to sockets? I'm assuming one would simply need to add the sockets or is there something else to be wary of?
|
|
|
Post by matt@IAA on Nov 13, 2020 18:40:05 GMT -6
Dono, but the sockets are quite a bit bigger than the pins so the hole might not be large enough.
|
|
|
Post by nudwig on Nov 13, 2020 19:17:57 GMT -6
Dono, but the sockets are quite a bit bigger than the pins so the hole might not be large enough. Yeah, I'm thinking me with a handheld drill isn't the best method for the conversion. Will leave them be for now and if/when there's a failure just solder.
|
|
|
Post by Guitar on Nov 13, 2020 19:46:23 GMT -6
I can talk about plugins only, I favor the UAD API EQ's, one of UAD's best. They are a constant reach for me. The Waves are "ok" but not as good, to me. They can be noisy with the "Analog" switch turned on.
The Overloud GEM one is good in the sense that it's not stepped, so it's finely adjustable, and it has four bands rather than three. But the distortion on it, someone tested it and it's a little weird so I don't use it all the time. I reach for the UAD's anyway.
Plugin Alliance is working on a channel strip with a preamp, EQ, and comperessor I hope, but that one's not even released yet. Their other console plugs are so good that I have high hopes for it.
|
|
|
Post by johneppstein on Nov 13, 2020 20:20:34 GMT -6
I like to be able to make finer adjustments, rather than being limited to 2db steps FWIW there was a recent version of the 550A (model "550Ad") that had ~1dB increments to a max of 6dB. I assume they're special order units, but VK had several of them for sale about a year or so ago: vintageking.com/api-550ad-equalizer-reissue-5097-14-usedThat's the "Mastering version". Been around for quite a while, but maybe not alwaysd in production.
|
|
|
Post by Mister Chase on Nov 13, 2020 21:13:55 GMT -6
I only have one BT50 here but it is fantastic. I will be building more, but it is an arduous build if you ask me. I want a whole mess of them. If you can't get a mixed sound with one, your source is probably recorded wrong.
Great, great EQ. I do like the UAD 550a quite a bit, too. I just like hardware.
|
|
|
Post by EmRR on Nov 14, 2020 12:16:49 GMT -6
Dono, but the sockets are quite a bit bigger than the pins so the hole might not be large enough. Yeah, you’d certainly have to enlarge the holes AND get the centering right.
|
|
|
Post by Bob Olhsson on Nov 16, 2020 12:36:51 GMT -6
The 550 replaced the Motown/Langevin graphic eq. in my world after my first session using them at Wally Heider's. I bought a pair used a few months later and still have them.
|
|
ericn
Temp
Balance Engineer
Posts: 14,810
|
Post by ericn on Nov 16, 2020 17:09:09 GMT -6
Dono, but the sockets are quite a bit bigger than the pins so the hole might not be large enough. Yeah, you’d certainly have to enlarge the holes AND get the centering right. Or put the time and effort into some kind of daughter board for the OP AMPs. Easier to just sell and buy more traditional 550’s if this truly important. I’ll say this when I had some modern 560’s and replaced the 550’s with some red dots you should have seen the smile on my face.
|
|