Hello! 24:00 thank's for mentioning! At the moment I'm working on a video about UI and capabilities of the decoder. I've might add that the capture itself is not required to be very high quality. It will work even with crappy USB dongles and low-quality decks with unstable drum speed or low-bandwidth signal path. The main issue is, as you mentioned at 24:40, most of video capture devices can not capture inactive region and ~20 lines (per field) of PCM at the top of the frame will be lost. But it is not as bad as you can imagine. STC-007/PCM-F1 have error correction codes built-in and allow up to 16 (PCM-F1) or 32 (STC-007) consequtive lines to be damaged/lost in a span of 128 lines and then it will be fully recovered into original samples. It means that NTSC recordings should be fully recoverable and PAL STC-007 recordings should also be fully recoverable. Only PAL PCM-F1 ones will have problems with too many missing lines. For others if tape and VTR are in good working order then decoding should be errorless. Also, there is some capture hardware that can capture inactive region as well. Some old Magewell and Matrox cards should do that I think. So "full-frame" capture is possible and software decoder will not have a disadvantage of missing lines of data. PCM-1600/1610/1630 and PCM-1 decoding now is also available but that is still in testing mode because I don't have many captures to test my decoder on. One thing that software decoder provides (what hardware one does not) is counters for how many samples were actually lost/damaged/interpolated during the process. So you can be aware if there were any issues during the capture.
13:10 - I don't know PCM-1630 internal structure and schematics, but based on my knowledge of STC-007 decoders and PCM-1600 format specs I think that this board converts from analog video to the digital audio bitstream. It seems like top right part of the board works with video and binarizes it into stream of 1's and 0's. After that bottom half of the board captures words like "left audio sample, "right audio sample", "parity word", "CRC word" from the stream, stores those into single-frame memory buffer and then deinterleaves data and corrects any errors that it had found. Output should be just audio bitstream that could be routed to SDIF/SPDIF/DAC/whatever. So for you purposes this one board is really the one you really DO NEED. PCM-16x0 doesn't use any visual markers (as STC-007 does) to align data horizontally, so it should rely on precision of a VTR and some internal fixed timing that tells this board when to sample each of 193 bits in the video line. And my thought are that that timing is what is off on this board (can those trimmers on the PCB be for that?..), so entire line virtually shifts to the side and data gets screwed. Or, maybe, frame buffer RAM gone bad and damages data before it gets deinterleaved.
Update. I've checked manual on PCM-1630 and yes DEC-15 board does exactly what I though - it converts analog video into digital audio bitstream. Those trimmers are in the video part but those have nothing to do with timing, those set video sync and data separation levels for binarization. Probably should be verified with an oscilloscope anyway. There also SW3 dip switches that should be set to ON-ON-OFF-OFF (1...4) and manual says that those are for factory testing and should not be altered. Big square ICs are CX23071 (Decoder-A) and CX23073A (Data separator). Also there are DIPs CX23072 (Decoder-B), CX23074 (Sync separator) and HM6148 (Static RAM). Sync and data separators handle video input and produce timing signals for bit extraction. Decoder-A checks for CRCs and interfaces with SRAM for the frame buffer, Decoder-B handles parity checks, error correction and muting.
This is probably one of my favorite videos. I was around in radio studios when CD's arrived so I was aware of what it took to make them, but never had the chance to see these amazing machines, I knew they existed. Thank you for that delve in to the chips. I quite like the wire mod, it worked even tho it didn't meet with your approval :) they did what they had to do... Thanks for this, most enjoyable.
The PCM 100 (and also the PCM 10) uses the STC007 standard, so NTSC 44.056 kHz sampling rate. I have a betamax tape with a recording made on a PCM 10 and it is fully compatible with the F1/701/601/501, and of course 14 bit and NTSC formatted. Interesting subject!
@@VCD-Channel Hello VCD, I have seen your very interesting website with the development of a PCM processor based on the PCM 10. I hope that one day I can order such a processor from you :). I still have the PCM 10, although it does not function properly. I had the PCM 100, but sold it again, as did the 1610 and 1630. I didn't have enough room for it and I'm still sorry I sold it. In a corner is a Sony PCM 85, which I can't find anything about on the internet. Maybe someone here knows more about it. My interest in PCM dates back to 1987, when I bought my first modified 501. Many followed. The modified ones are especially interesting, such as those from Audio+Design and RTW. There seem to have been more variants. Sorry for straying from this topic. Kind regards, Lennard
As you know, the pcm u-matic format was used to make master tapes to send to the pressing plant to make a glass master. It’s possible the odd box you have, came from a pressing plant perhaps. My dad worked extensively with this format, and dropouts were a huge problem, triggering the CRC. There is no dropout compensation, so this is where the RAR functionality is useful. The DMR-2000 was primarily used to make recordings and the DMR-4000 to playback One thing the DMR-4000 has, which the standard BVU-800 doesn’t, is an automatic head cleaner, which helps reduce clogging. I also have a PCM-1630/DMR-4000 setup and the AES option, which I’ve never been able to get to work (a rainy day project). I also have the course notes for the DMR-4000, as well as a rare product brochure, which I must get scanned.
I remember a rack mounted Studer/ReVox CD player that had separate digital outputs for left and right and I'm sure it also had a word clock output, so maybe that used this SDIF format too. Most professional equipment I've come across have the standard AES/EBU XLR (or BNC) sockets which carry a single stereo pair, so this CD player was odd having independent left and right channels.
Actually there were some early digital audio systems that used minicomputers (think PDP-11) to actually store digital audio data. However that was _really_ expensive and thus only used for editing. Recording it onto conventional cheap U-Matic or even Beta tapes with cheap recorders was much more practical. Remember that U-Matic recorders were as cheap as cars even back in the 1970s.
Brilliant thank you for posting - the Digital Domain is a law unto itself it would seem - when I moved from the Analogue path on 24 - 16 - 8 - 2 Track tape machines to the Adat Black Faced recorders in 1990 it was another world altogether with many trials and errors - I still have my 2x8 Adats plus the BRC both with very low, around 800 hours, used for backing up the tape machines apart from the 24 2" recorders - now we're into different territory with handheld and location units going through the Sony pro units to todays Tascam, Roland R-26 - R-07 Sound Devices MixPre6ii and the Tascam HD-P2 plus the HHB Minidisc - all for me acceptable then and now although using the higher 96 & 192 take up that much more space... does it actually match up the the wide frequency of analogue yet? Perhaps the human perception destroyed by early CDs with the filters taking out second and third harmonics that we are somehow aware of leaving that brittle sound of those early CD's - who knows but thank you so much for posting your brilliant videos and so instructive. Michael OAK77uk
For a while I owned the 4000 model umatic, which actually came from a mastering house called The Exchange. Unfortunately I didn't have the decoders and couldn't find any, so I sold the unit in the end.
Noise reduction akin to Dolby on analogue tape. Of course it's pointless on a format that has 96 dB DR to begin with and causes compatibility issues, but that was the idea.
It will be interesting to see a software recovery in a computer using one of the later models uMatic machines with built in TBC card, Software methods came a long way and they are more flexible in terms of options and settings, and don't worry no one is going to put you out of business, Very few people own and know how to repair uMatic machines or any vintage of piece of gear for that matters. For capture cards, I think for PAL is not an issue, For NTSC capture cards capable of doing 486 lines should capture the entire PCM frame. I will ask the vhs-decode team to see if they can work on a PCM plugin since the hardware is capable of capturing full frames 625/525 lines.
Maybe you can borrow a heat camera and point it at the bad pcb then the good pcb to see if one i.c is consuming more power than it should and getting slightly warmer than the otherpcb. I would of blamed the supply capacitors to, noise free supplies are a must. Sorry i know bugger all about decoders.
I have 9 hits on the subject of 88.2 Khz at Qwant. If there is 4 channel software for 88.2 kHz (SRC+Mix), you can convert the stereo tracks to 88.2 kHz and make a copy. Move tracks 1+2 one sample back, leave the copied stereo tracks 3+4, and downmix tracks 1+4 to 44.1 Khz stereo. Software may search someone with better knowledge of English.
hi you are right the cost of keep this gear running is very high why in the past i have keep away i only buy some gear you keep going on about I HAVE I HAVE all the time i have said in my old channel just ask do i have this formats i am getting gov backing so i could start up again love your video's space and money 13gb a hour is very high i been using 4gb a hour for 12 years can't go to HD the monitor i have plays up i have to cut corners
Yes, very similar mechanism. It has the same design fault too: If a tape is sticky and grabs hold of one of the guide posts during unlace, then the mechanism for the post placement can get on the wrong side of another lever and jams up the loading ring properly.
Hello!
24:00 thank's for mentioning!
At the moment I'm working on a video about UI and capabilities of the decoder.
I've might add that the capture itself is not required to be very high quality. It will work even with crappy USB dongles and low-quality decks with unstable drum speed or low-bandwidth signal path.
The main issue is, as you mentioned at 24:40, most of video capture devices can not capture inactive region and ~20 lines (per field) of PCM at the top of the frame will be lost.
But it is not as bad as you can imagine. STC-007/PCM-F1 have error correction codes built-in and allow up to 16 (PCM-F1) or 32 (STC-007) consequtive lines to be damaged/lost in a span of 128 lines and then it will be fully recovered into original samples.
It means that NTSC recordings should be fully recoverable and PAL STC-007 recordings should also be fully recoverable. Only PAL PCM-F1 ones will have problems with too many missing lines. For others if tape and VTR are in good working order then decoding should be errorless.
Also, there is some capture hardware that can capture inactive region as well. Some old Magewell and Matrox cards should do that I think. So "full-frame" capture is possible and software decoder will not have a disadvantage of missing lines of data.
PCM-1600/1610/1630 and PCM-1 decoding now is also available but that is still in testing mode because I don't have many captures to test my decoder on.
One thing that software decoder provides (what hardware one does not) is counters for how many samples were actually lost/damaged/interpolated during the process. So you can be aware if there were any issues during the capture.
Digital audio on video tape has always been one of my favorite topics. Thanks for posting!
13:10 - I don't know PCM-1630 internal structure and schematics, but based on my knowledge of STC-007 decoders and PCM-1600 format specs I think that this board converts from analog video to the digital audio bitstream. It seems like top right part of the board works with video and binarizes it into stream of 1's and 0's. After that bottom half of the board captures words like "left audio sample, "right audio sample", "parity word", "CRC word" from the stream, stores those into single-frame memory buffer and then deinterleaves data and corrects any errors that it had found. Output should be just audio bitstream that could be routed to SDIF/SPDIF/DAC/whatever. So for you purposes this one board is really the one you really DO NEED.
PCM-16x0 doesn't use any visual markers (as STC-007 does) to align data horizontally, so it should rely on precision of a VTR and some internal fixed timing that tells this board when to sample each of 193 bits in the video line. And my thought are that that timing is what is off on this board (can those trimmers on the PCB be for that?..), so entire line virtually shifts to the side and data gets screwed. Or, maybe, frame buffer RAM gone bad and damages data before it gets deinterleaved.
Update.
I've checked manual on PCM-1630 and yes DEC-15 board does exactly what I though - it converts analog video into digital audio bitstream.
Those trimmers are in the video part but those have nothing to do with timing, those set video sync and data separation levels for binarization. Probably should be verified with an oscilloscope anyway.
There also SW3 dip switches that should be set to ON-ON-OFF-OFF (1...4) and manual says that those are for factory testing and should not be altered.
Big square ICs are CX23071 (Decoder-A) and CX23073A (Data separator). Also there are DIPs CX23072 (Decoder-B), CX23074 (Sync separator) and HM6148 (Static RAM).
Sync and data separators handle video input and produce timing signals for bit extraction. Decoder-A checks for CRCs and interfaces with SRAM for the frame buffer, Decoder-B handles parity checks, error correction and muting.
This is probably one of my favorite videos.
I was around in radio studios when CD's arrived so I was aware of what it took to make them, but never had the chance to see these amazing machines, I knew they existed.
Thank you for that delve in to the chips.
I quite like the wire mod, it worked even tho it didn't meet with your approval :) they did what they had to do...
Thanks for this, most enjoyable.
Fun fact: In the video for Dire Straits' Money for Nothing, there's PCM audio off a U-matic tape used as static on the video wall.
Here it is:
ua-cam.com/video/lAD6Obi7Cag/v-deo.html
The PCM 100 (and also the PCM 10) uses the STC007 standard, so NTSC 44.056 kHz sampling rate. I have a betamax tape with a recording made on a PCM 10 and it is fully compatible with the F1/701/601/501, and of course 14 bit and NTSC formatted. Interesting subject!
I have a PCM-10. Сonfirm your words.
@@VCD-Channel Hello VCD, I have seen your very interesting website with the development of a PCM processor based on the PCM 10. I hope that one day I can order such a processor from you :).
I still have the PCM 10, although it does not function properly. I had the PCM 100, but sold it again, as did the 1610 and 1630. I didn't have enough room for it and I'm still sorry I sold it.
In a corner is a Sony PCM 85, which I can't find anything about on the internet. Maybe someone here knows more about it.
My interest in PCM dates back to 1987, when I bought my first modified 501. Many followed. The modified ones are especially interesting, such as those from Audio+Design and RTW. There seem to have been more variants.
Sorry for straying from this topic.
Kind regards, Lennard
As you know, the pcm u-matic format was used to make master tapes to send to the pressing plant to make a glass master. It’s possible the odd box you have, came from a pressing plant perhaps.
My dad worked extensively with this format, and dropouts were a huge problem, triggering the CRC. There is no dropout compensation, so this is where the RAR functionality is useful. The DMR-2000 was primarily used to make recordings and the DMR-4000 to playback
One thing the DMR-4000 has, which the standard BVU-800 doesn’t, is an automatic head cleaner, which helps reduce clogging.
I also have a PCM-1630/DMR-4000 setup and the AES option, which I’ve never been able to get to work (a rainy day project).
I also have the course notes for the DMR-4000, as well as a rare product brochure, which I must get scanned.
I remember a rack mounted Studer/ReVox CD player that had separate digital outputs for left and right and I'm sure it also had a word clock output, so maybe that used this SDIF format too. Most professional equipment I've come across have the standard AES/EBU XLR (or BNC) sockets which carry a single stereo pair, so this CD player was odd having independent left and right channels.
Yes it mighty have been SDIF.
Awesome machines and as usual great video. Keep up the great work
Awesome machines and as usual great video. Keep up the great work 👍👍
I'm only a few minutes into this video, but I know it is going to be a very, very, very interesting video!! Just my piece of cake.
Here is how you prounounce Mr. Fagear's name. It's Maksim "Kryuckov".
Been working on my own version using dsPIC33.
Actually there were some early digital audio systems that used minicomputers (think PDP-11) to actually store digital audio data. However that was _really_ expensive and thus only used for editing. Recording it onto conventional cheap U-Matic or even Beta tapes with cheap recorders was much more practical. Remember that U-Matic recorders were as cheap as cars even back in the 1970s.
Brilliant thank you for posting - the Digital Domain is a law unto itself it would seem - when I moved from the Analogue path on 24 - 16 - 8 - 2 Track tape machines to the Adat Black Faced recorders in 1990 it was another world altogether with many trials and errors - I still have my 2x8 Adats plus the BRC both with very low, around 800 hours, used for backing up the tape machines apart from the 24 2" recorders - now we're into different territory with handheld and location units going through the Sony pro units to todays Tascam, Roland R-26 - R-07 Sound Devices MixPre6ii and the Tascam HD-P2 plus the HHB Minidisc - all for me acceptable then and now although using the higher 96 & 192 take up that much more space... does it actually match up the the wide frequency of analogue yet? Perhaps the human perception destroyed by early CDs with the filters taking out second and third harmonics that we are somehow aware of leaving that brittle sound of those early CD's - who knows but thank you so much for posting your brilliant videos and so instructive. Michael OAK77uk
It looked like when you had the original decoder board in, both A and B lights were on, but when you changed to the Known Good, only A was on again
Yet RAR was on, so it must have been working on both channels. Strange.
Is there any value in a 1630/dmr4000/2000/dta setup?...asking for a friend. thanks
Oh yes, these systems in good working order are valuable. Hard to put a price on it, but a good system might be worth over £500.
Yet another very interesting video - thank you! I am curious to know what the extra little "daughterboard" is actually for on that decoder board.
Really don't know. It will be documented somewhere with Sony.
For a while I owned the 4000 model umatic, which actually came from a mastering house called The Exchange. Unfortunately I didn't have the decoders and couldn't find any, so I sold the unit in the end.
What was the idea of using "Emphasis" on digital formats?
Noise reduction akin to Dolby on analogue tape. Of course it's pointless on a format that has 96 dB DR to begin with and causes compatibility issues, but that was the idea.
Sony Umatic as "Digital" in the BVU series. I haven't seen that before. But a nice setup you have and exciting to follow your channel
So, they added that extra chip as a sort of latch to make the VU meters hold the last value? I guess someone wanted that feature for some reason
You think so? I suspect it was a fault.
Hi, I wanna shop a 1610 or 1630, could you tell me how I can find them ?
Very hard to find, especially the 1630 with all the options installed. I have a 1610 that I'm not using, but not sure I really want to part with it.
It will be interesting to see a software recovery in a computer using one of the later models uMatic machines with built in TBC card, Software methods came a long way and they are more flexible in terms of options and settings, and don't worry no one is going to put you out of business, Very few people own and know how to repair uMatic machines or any vintage of piece of gear for that matters. For capture cards, I think for PAL is not an issue, For NTSC capture cards capable of doing 486 lines should capture the entire PCM frame. I will ask the vhs-decode team to see if they can work on a PCM plugin since the hardware is capable of capturing full frames 625/525 lines.
Maybe you can borrow a heat camera and point it at the bad pcb then the good pcb to see if one i.c is consuming more power than it should and getting slightly warmer than the otherpcb.
I would of blamed the supply capacitors to, noise free supplies are a must.
Sorry i know bugger all about decoders.
I asked Santa for a FLIR but he said they are too expensive...
I have 9 hits on the subject of 88.2 Khz at Qwant. If there is 4 channel software for 88.2 kHz (SRC+Mix), you can convert the stereo tracks to 88.2 kHz and make a copy. Move tracks 1+2 one sample back, leave the copied stereo tracks 3+4, and downmix tracks 1+4 to 44.1 Khz stereo. Software may search someone with better knowledge of English.
hi you are right the cost of keep this gear running is very high why in the past i have keep away
i only buy some gear you keep going on about I HAVE I HAVE all the time i have said in my old channel just ask
do i have this formats
i am getting gov backing so i could start up again love your video's space and money
13gb a hour is very high i been using 4gb a hour for 12 years can't go to HD the monitor i have plays up
i have to cut corners
It's definitely a modified BVU800
Yes, very similar mechanism. It has the same design fault too: If a tape is sticky and grabs hold of one of the guide posts during unlace, then the mechanism for the post placement can get on the wrong side of another lever and jams up the loading ring properly.