Attempt to use Sony PCM-2500A/2500B DAT recorder to convert PCM1630 SDIF-2 to S/PDIF. Part 1.
HTML-код
- Опубликовано: 20 июл 2024
- We bought in a Sony PCM-2500A DAT recorder with PCM-2500B interface unit. The plan was to use this to convert the obscure and obsolete SDIF-2 digital data stream from a Sony PCM-1630 Umatic digital audio recorder set, to a more modern S/PDIF, AES/EBU or Toslink signal.
Though huge progress was made from an initially dead system, in the end there are good reasons that this was not the best equipment for the job.
However, there's a part 2: • Part 2, some success! ...
Then part 3: • REC button presser. Pa...
00:00 Intro
02:18 Look inside PCM-2500A
05:12 Look inside PCM-2500B
06:07 First power-up
07:54 Start fault-finding
23:15 First light
30:04 S/PDIF on a 'scope
33:09 Searching for S/PDIF inside PCM-2500B
34:47 Trying a Tascam DR100 MKII recorder
38:23 A "Wow!" moment
40:28 SDIF-2 gives us music!
42:00 Conclusion
Audio and video transfers: www.video99.co.uk/
Please support us on Patreon / video99couk
or Paypal to colin@video99.co.uk
Music “Let It Run” with permission, copyright Cristie/MacFarlane.
Sorry I do not offer an audio or video equipment repair service. 
Наука
There's a part 2: ruclips.net/video/WOYMCUdlLUY/видео.html
Highly recommended!
this is like Archeology, when we see the pyramids, we think of ancient civilizations , their artists, their scientists , their hopes and dreams , that is how I feel right now learning about this ''artifact'' , amazing engineering, passion, long hours at work, a-ha moments.... wow
I absolutely admire your work ❤
A converter sounds like a fun microcontroller project. S/PDIF is well documented, though I'm less sure about SDIF-2.
SDIF-2 is documented too. But we're talking a bigger job that I could tackle.
What a very interesting video about the PCM 2500A / B. Thank you for that. I immediately threw everything I was working on aside to watch this video.
I have also tried this from 2500B to PCM-100 but it didn't work because both units don't have WC-in. There was only a flaky sound.
I myself have the RTW-AD3 digital audio format converter which converts 701 / F1 format to SDIF2. You can connect the copy output of a PCM-601esd to the video input of the RTW so that you can then go from S/PDIF to SDIF2, and then it works, at least with 1610. But the RTW box only has WC out. So it only works one way.
And Audio + Design has had one in the ProBox series, ProBox 3.
A Sony K1183 or audio & design probox 3, will also do SDIF to AES conversion, but these are equally scarce. I have the Sony DABK-1631 for the PCM-1630, but could never get it to work. It’s a rainy day project but talking to another engineer, we agreed that the dip switch configuration is probably wrong. Early days of AES.
You know who to sell it to if you ever give up on it! Do you have the installation manual? It's not that hard to find or I can send it via wetransfer.com if you need it.
What a journey, too bad it didn't really go anywhere in the end...
The chassis of the AD/DA converter looks like a DAT mechanism would fit into the blank space. Maybe that was their plan until they found that one case was just not going to be enough to fit everything...
I have come across so many bad ELNA capacitors over the years, I don't consider them one of the good brands anymore. They are on the list of suspicious components.
Is there really no way to access the power supply and mechanism from the bottom? Maybe there is a way to take the chassis apart somehow to gain access, even if there is no "official" bottom access panel? I know too well some Sony equipment was designed without serviceability in mind, but from what I can see in the video, I can't imagine how they could have assembled everything in the factory without bottom access...
There's another PCB underneath. Access is appalling. However this may be a prototype unit, so things probably improved later in production. I could pull it all apart again to work on the mechanism. I may get some hints by looking in the DTS-1000ES. But it is worth it? The machine won't do what I want it to do anyway.
@@video99couk I agree there is probably no point repairing the DAT, it looks like it has been through hell and back, so who knows what it would take to make it work.
I saw there is a circuit board on the underside, however maybe that is a design similar to the old VCRs where all or most of the connectors to the board are along one side, so that it would be possible to fold the bottom circuit board out of the way...
@@DrCassette Yes, it could swing out nicely. It would be good if it did! Still there is such a mass of cabling from the deck that access will be dreadful.
Perhaps you also know the Sony PCM-85?
This is a feedback about using the PCM decoder PC software, I have good news, I tried it with a capture device hooked up to my VCR using S-Video out, the capture device has SDI out (BrightEye75), then from SDI I used the BM UltraStudio SDI to USB to connect to Win10 machine, Worked like a charm. I tried SDI out just because you told me your VCR output is SDI and I wanted to test it, my tape is 16/44.1 though so not sure about Umatic. Also I didn’t have to save video, just wav file, although I could if I wanted to by using the copy out button.
Interesting that you've got that to work. Was this a Umatic or Beta sourced PCM format?
video99.co.uk VHS format
@@Capturing-Memories Oh right. VHS wasn't so often used for PCM audio. I think what I was trying to say then, is what PCM audio format is it, 1610/1630, PCM-F1, or some other format?
video99.co.uk It’s PCM 16/44.1 made by a Sony PCM adapter.
@@Capturing-Memories Sorry to be pedantic, do you know the model number of that adapter? I'm just trying to work out which formats work with that software.
S/PDIF is designed to stay "high" and "low" for equal amounts of time so that on average, the DC voltage evens out at 0. It's too bad you didn't know that, because you would have recognized at a glance that the signals at 33:09 and following, are definitely not even close to S/PDIF.
I have no experience with SDIF2 but a quick Google seems to indicate that SDIF2 may simply be serial bit streams (sort of like a serial port, only much faster and therefore with strict requirements about things like impedance). The channels are separated and there's a word clock that probably just gets used to recognize the start of each sample. It could be similar to I2S but simpler, but I don't know.
In S/PDIF, each bit is encoded as either 1 transitions from positive to negative or vice versa (representing a 0 bit), or 2 transitions (representing a 1 bit). This is called biphase encoding. Instead of a separate line for word sync, the S/PDIF signal simply violates its own biphase encoding at the beginning of each sample: Each sample (known as a subframe) is preceded by a preamble that indicates whether the following subframe is for the left channel or not, and whether it's the start of a block. There are 32 bits in a subframe (including the preamble) and 24 of those bits can be used for audio. The remaining bits are for parity and validity, and for two subchannels that only transfer one bit per subframe or one bit per frame. See my article at hackaday.io/project/24911-propeller-spdif-receiver/details for some more information, based on the ISO-60958 standard.
As I said, I don't know much about SDIF2 but it seems to me that a fast modern microcontroller should be able to convert an SDIF2 signal into an S/PDIF signal and vice versa; the main problem is probably that no-one needs it so almost no-one makes a converter and they can ask for premium prices.
Good luck!
Well I worked out pretty quickly that the signal I had didn't look right so I didn't expect it to work. It's interesting to have all that extra information on S/PDIF. Now there are S/PDIF and AES/UBU outputs in this system, but what data they are fed with we don't know, and it's probably only from tape so not related to SDIF-2. So this system is just not going to do it. I'm going to be looking for one of the alternative solutions, but this was still an interesting exercise.
I think I can decode it from a PC captured monochrome analog signal with an homebrew made software.
It would be a great project to do.
The specs of the PCM1630 says it encodes 6 samples per line of the monochrome analog video signal.
Tell me if you're interested. I will do it for free.
Some sites says it is 14bits per sample plus two for error correction (or simply error detection). But the samplerate could be 44100 or 44056 depending if the framerate is 30 or 25Hz.
It doesn't mention reed solomon FEC or anything like that. I guess the error bits are only for parity.
Edit: I found the full spec on the manual of the PCM1630 after 1-10 (signal format).
I had thought about this a while back, then recently some PCM decoding software appeared out there (can't quite find the link right now). I tried it with some PCM-F1 format digital audio, which is quite similar to the PCM-1610/1630 format. The results were poor in my experience, there were too many errors. I think the real hardware handles tape imperfections and timing jitter better than you can do with a captured video. It sounds like a really good idea but in practice it may not work quite so well.
Yeah a video capture device is not going to work well with PCM on a video tape. The problem is that the audio is encoded as full-brightness and black, on a line-by-line basis, and all video encoding methods are going to look at multiple lines of PCM data and go: "Yeah this 8x8 pixel block looks like noise so I'm going to encode it to something that kinda looks the same based on cosine gradients and stuff". And it will compress all the useful data out. You would have to have a video grabber that processes the incoming analog video in uncompressed format, which is surprisingly difficult because of the required bandwidth.
@@JacGoudsmit Is the DV codec compressed in luma? I'm not sure it is. But still it didn't work very well when I tried it with some PCM decoder software on PCM-F1 format.
@@video99couk I'm not sure if I understand the question. Digital Video hardware in itself is probably perfectly capable of creating a digital version of a video tape with PCM encoded audio on it, as long as the resolution of the digitizer is sufficient. PCM only uses the luminance signal (No color) and video grabbers separate luma from chroma before they do their work. The luma digitizer has higher resolution than the chroma digitizer but for PCM tapes there's no chroma so it doesn't matter.
If you're talking about the format of the data on e.g. a miniDV camcorder: yes that's compressed: unless I'm mistaken they use MJPEG which is basically a stream of JPEG pictures. Just like almost all video (file) formats they use "macro blocks" of 8x8 pixels to represent the video. And to reduce the data bandwidth, they use lossy compression to represent each macro block; that's "good enough" for normal video but won't work for video that contains pcm audio data. Compression algorithms usually represent the luma and chroma information in some variation of YUV encoding so the chroma information is included whether you encode a color picture or not. Depending on the encoding, the compression algorithm may assign more bits to the luma signal if there's no chroma information but I'm not familiar enough with video codecs to know the details about this.
@@video99couk I would try a live feed to the capture card than an actual capture that way you don't have to worry about the video codec used. The decoder has some settings that you may have to tinker with to get it error free by clicking the setting button, Here is a thread I made:
www.tapeheads.net/showthread.php?t=82905
I have the feeling that Sony was making devices limited on purpose, so you have to open your wallet yet again!.
And if you purchased another manufacturers device, there would be a special part of the sony digital signal to create an error, so wallet open again.
I don't trust companys for doing " The right thing.", money money money..
I hope you do find a solution, there has to be a way to block buster ha ha off track there :-D
In this case, it was just early days for digital audio. The standards hadn't been set yet. S/PDIF and the professional AES/EBU came along just because the SDIF-2 protocol required too many cables.
he is a mini-you lol